DETERMINANTS OF NUTRITION-RELATED RICKETS IN CHINESE
CHILDREN AND ASSOCIATED HEALTH OUTCOMES
Mark Allan Strand
B.A., Luther College, 1985
M.S., University of Minnesota, 1991
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Health and Behavioral Sciences
! A i
@2004 by Mark Allan Strand
All rights reserved.
This thesis for the Doctor of Philosophy
Mark Allan Strand
Strand, Mark Allan (Ph.D., Health and Behavioral Sciences)
Determinants of Nutrition-related Rickets in Chinese Children and Associated Health
Thesis directed by Associate Professor David P. Tracer
Combining the methods of epidemiology and the behavioral sciences, this
study was designed to determine the prevalence of rickets in rural Chinese children
and to ascertain the determinants of rickets in these children. Additionally, the
perceptions of mothers and Chinese health workers of this disease were determined.
Subjects were 250 children age 12-24 months in rural Shanxi Province, China
and their mothers. Children were examined physically for the presence of rickets,
blood was drawn for laboratory analysis and x-rays were taken of each childs wrists.
Mothers were administered a 33-question survey of the nutritional, behavioral and
environmental influences on the child.
Relying on an integrated method of clinical signs, x-ray and alkaline
phosphatase levels, the prevalence of rickets was found to be 3.7%. Furthermore, it
was demonstrated that only five clinical signs reflect active rickets wide wrists,
frontal bossing, pearling, Harrisons sulcus, and O-shaped legs. These results have
important implications for the method of diagnosing rickets in China because using
the standard Chinese method of clinical signs alone determined a rickets prevalence
Children with rickets were found to have increased rates of diarrhea and upper
respiratory tract infection. Vitamin D-deficiency in April was found among 65.3% of
children. Low calcium intake was found to be correlated with low 25-OH-D. Iron
and protein intake were also negatively associated with rickets. This reflects the
multifactorial nature of rickets, and children with rickets suffer from a host of other
Maternal knowledge of how to prevent rickets is not associated with
preventing rickets in her child and mothers do not consider rickets to be a serious
disease. The intensity with which the Chinese practice of zuoyuezi (cloistering of
mother-infant pair for >100 days) is practiced contributes to rickets. Intensity of
zuoyuezi is unrelated to family SES, maternal education, or the role of po-po, so it is a
practice embraced by the young mothers themselves and not related to being
uneducated or poor. Children whose fathers spend more time away from home have
higher rates of rickets and poorer health.
This abstract accurately represents the content of the
David P. Tracer
This dissertation is dedicated to my wife Rene, and our three children Bjorn,
Marit and Anders. The process of doing the research and writing this dissertation has
been a heavy burden for all of us. Thank you for your support and prayers on my
behalf. I love you.
I also want to dedicate this work to my mother. Mom, you taught me early to
love reading and writing, and this passion has never left me. Along the way you gave
your life so selflessly to us, your children, that you had no time or energy for interests
of your own. For this sacrifice, I am forever indebted to you. Thank you and I love
This research relied on the support of many people. Dr. David Tracer of the
University of Colorado at Denver program in Health and Behavioral Science guided
the design of this project, and saw me through to the end. Dr. Zhang Peiying, Shanxi
Women and Childrens Hospital, has been involved in rickets prevention work for
many years, and since 1997, has served as Chief Advisor for the Well Baby Program
of Shanxi Evergreen Service. Dr. Li Sihan of the Shanxi Medical University has
done research on vitamin deficiencies in infants and children in north China for many
years, and selflessly guided this project as well. It is a privilege to have benefited
from the expertise of these mentors. Research colleagues from the Shanxi Province
Women and Childrens Hospital, Shanxi Medical University, the Yuci Maternal and
Child Health Station, and all the workers in the township clinics were such a blessing
to me, and worked so hard to successfully collect the data, even in the face of SARS.
In addition to David Tracer, the other members of my dissertation committee -
- Craig Janes, Nancy Krebs, Stanley Xu and Phil Fischer also showed much
kindness to me and took personal interest in the success of this project. Thank you so
Dr. Peng Guanxi of Yangqu County first introduced me to the problem of
rickets back in 1996. It has been a privilege to work together all these years. May the
children of Shanxi be the better for our efforts.
Thank you Evergreen colleagues for your support and for the privilege of
working together in Shanxi Province. It is my hope that our efforts are a blessing to
the children and others whom we serve. Then what does the Lord require of us? But
to do justice, to love mercy, and to walk humbly with God. Micah 6:8
This research was partially funded by a generous grant from the Nestle
List of Figures
List of Tables.
1. INTRODUCTION AND BACKGROUND...................................1
Purpose of the Study.......................................1
Rationale for this Research................................2
Main Research Questions.............................2
Rationale for these Research Questions..............3
Conceptual Framework of this Study........................10
Social and Cultural Epidemiology...................10
Social Cognitive Theory............................11
Cultural Explanatory Model of Illness..............13
Arrangement of the Dissertation...........................15
2. THEORETICAL AND EMPIRICAL BACKGROUND.........................16
The Biology of Rickets....................................16
Basic Introduction to Rickets......................16
Vitamins, Minerals and Bone Metabolism.............17
The Physiology of Rickets..........................20
Epidemiology of Rickets...................................24
Rickets Increasing in the USA......................24
Rickets in China.....................!.............25
Dietary Deficiencies Among Children in Rural China.27
Social and Cultural Factors in Rickets Epidemiology.......29
Decision-making and Authority in the Home............31
Traditional Health Beliefs and Practices...................32
History of TCM......................................33
Source of Illness in TCM.............................33
The Concept of Human in the Universe.................35
Rickets and Traditional Health Beliefs.....................35
Ancient History of Rickets in China..................35
Analysis of Rickets in Traditional Chinese Medicine..36
3. MATERIALS AND METHODS..........................................39
Characteristics of Subjects Involved.......................39
Research Location: Yuci, Shanxi Province, China......39
Process of Collaboration.................................. 43
Experimental Design and Methodology...................... 46
Composite Scores.................................... 47
Physical Examination and Biochemical Measures........51
Qualitative Data Collection.............................. 60
4. RESULTS I: QUANTITATIVE.......................................64
Children and Mothers.............................. 64
Blood Chemistry Analysis...................................68
Illness History of Children................................75
Clinical Diagnosis Method (Clinical)..............78
Radiographic Method of Diagnosing Rickets
Rickets Diagnostic Method Using Clinical, Radiography
and 25-OH-D Results (Clin, X and D)...............87
Clinical Symptoms Plus Alkaline Phosphatase Method
Diagnosing Rickets Using Clinical, X-ray and Alkaline
Phosphatase (Clin, X and ALP).....................94
Predictive Power of Clinical Signs and Symptoms.....95
Rickets Rate Among North China Children............100
Clinical Method Using Only Significant Clinical Signs
(Sig. Clinical Signs).......................... 101
Seasonal Decline in Rates of Rickets......................102
Association of Risk Factors with Rickets..................103
Rickets and Health Outcomes........................107
Demographic Factors and Rickets.........................115
Age and Clinical Rickets..........................115
Gender and Rickets................................116
Absence of the Father.................................118
Dietary Factors and Rickets.............:................... 118
Vitamin D and Calcium.............................119
Age of Supplementing Solid Foods................ 122
Breast-feeding and Rickets........................124
Cultural/Behavioral Characteristics of Children and Rickets.127
Additional Descriptive Statistics.......................133
Instruments and Scales............................135
5. RESULTS II: QUALITATIVE DESCRIPTIONS OF RICKETS AND
Chinese Culture and Health..............................141
Maternal Health....................................... 143
Childs Health........................................ 145
A Healthy Child...................................145
Zuoyuezi: It is Part of Being a Woman.................147
Bugan! (Dont you dare!): Mothers Psychology.....150
Diet During Zuoyuezi..............................152
Taking the Child Outside...........................153
Other Zuoyuezi Customs.............................155
Attitude Toward Health Workers...........................157
Childhood Illness...................................... 158
Cause of Illness...................................158
Summary of Zuoyuezi and Shanghuo.........................159
Diet and Nutrition.......................................160
Xiaohua huhao (Poor Digestion).....................160
Environment and Weather..................................161
The Human in the Universe..........................162
Changing Culture and Customs.............................163
The Meaning of Rickets Among North China Mothers.........164
Causes of Rickets..................................164
Treatment for Rickets..............................166
Po-po Paternal Mother-in-law...........................167
Daughter- and Mother-in-law Relationship...........168
Desires for Children in 20 Years.........................169
6. DISCUSSION................................................. 172
Rickets Diagnosis and Prevalence.........................172
Interpreting Clinical Signs and Symptoms of Rickets
Cultural Beliefs and Perceptions of Clinical Signs and
The Process of Developing Rickets...................176
Clinical Diagnosis of Rickets.......................177
Suggestions for China Childs Health Workers........178
Different Results in X-ray Reading.........................179
Rickets Risk Factors.......................................181
Gender Effect on Rickets...................................186
Gender Differences in Rickets and Biochemical
Vitamin D-deficiency and Illness...........................188
Rickets in China and Associated Factors....................191
Healing from Spring to Fall................................195
Nutritional Deficiencies............................. 196
The Case for Calcium-insufficiency Rickets Globally........196
Comparing Rickets in China and Nigeria..............199
Summary of Calcium-insufficiency Rickets............200
Cause of Rickets...........................................201
Biochemical Assessment of the Relationship Between Rickets,
Nutrition and Disease......................................202
Vitamin D Deficiency................................203
Cultural Factors and Rickets...............................209
Rickets as Understood from the Perspective of Social
Methods to Reduce Rates of Rickets..........221
Proposed Intervention to Prevent Rickets....225
Components of Rickets Prevention Program....226
Lessons Learned................................. 228
Health Research in a Developing Country.....228
Developing Global Research..................230
Methods and Data Limitations......................232
Unique Contributions of this Study................233
A. GLOSSARY OF CHINESE TERMS AND COMMON
B. CHINAS RURAL HEALTH CARE SYSTEM....................237
C. RESEARCH IN THE MIDST OF THE SARS EPIDEMIC..........241
D. SUBJECT CONSENT FORMS...............................255
E. MOTHERS SURVEY FORM SPRING.......................262
F. CHILDS EXAM FORM SPRING..........................271
G. MOTHERS SURVEY FORM FALL........................ 274
H. CHILDS EXAM FORM FALL............................279
3.1 Map of China........................................................40
3.2 Map of Yuci City................................................... 40
4.1 Percentage reporting eating these foodstuffs often in the recent month
4.2 Zinc levels in spring.............................................. 70
4.3 Zinc levels in fall............................................. 70
4.4 Lead levels (pg/dl) among children in the spring................ 115
4.5 Mean calcium intake by age in spring...............................121
1.1 The concepts of the Social Cognitive Theory.........................12
2.1. Values for the variables used in calculating HeaLYs due to rickets in
Pakistan and China.................................................27
2.2 Biochemical indexes of girls aged 12-14 y in the Beijing area.......28
2.3 Serum levels of normal and rachitic infants in Shanxi Province (age 0-3)
2.4 Illness histories in households with and without rickets in Bangladesh....29
3.1 Relative nutritional content of foods commonly eaten by Chinese children
3.2 Score given for month of birth......................................50
3.3 Score given for time taken outdoors............................... 50
4.1 Descriptive statistics of child subj ects.......................... 65
4.2 Subject participation by village....................................67
4.3 Summary of spring and fall blood levels for children surveyed.......68
4.4 Percentage of children supplemented with vitamin D or calcium in the
recent month (spring data)..........................................71
4.5 Summary of 25-OH-D levels of children surveyed in spring and fall...71
4.6 Comparison of means of lab test results and in spring and fall......72
4.7 Gender differences in serum 25-OH-D and serum calcium...............73
4.8 Health perception and dietary differences between boy and girl children 74
4.9 Comparing illness rates among children from spring to fall..........76
4.10 Paired sample means and t-test results of growth parameters from spring to
4.11 Standards for clinical signs and symptoms of rickets................79
4.12 Comparing presentation of rickets signs and symptoms in spring and
4.13 Comparing clinical signs and symptoms of rickets with x-rays and alkaline
4.14 Radiographic rickets scoring results for Chinese and Western
4.15 Comparing clinical diagnosis versus radiographic diagnosis of rickets in the
4.16 Comparing clinical diagnosis versus radiographic diagnosis (Chinese reader)
of rickets......................................................... 86
4.17 Rickets rates using clinical, X-ray and 25-OH-D.......................88
4.18 Alkaline phosphatase levels of children with and without rickets
(Clin, X and D)................................................... 89
4.19 Multinomial logistic regression analysis of clinical signs and symptoms. 90
4.20 Diagnosing rickets by combining clinical diagnosis and alkaline
4.21 Comparing rickets in spring and fall using clinical and ALP results as
4.22 Mean serum 25-OH-D for each Clin, ALP outcome (spring)................93
4.23 Rickets (spring) rates by Clin, X and ALP...........................95
4.24 Multinomial regression analysis of rickets (Clin, X and ALP) and
clinical signs and symptoms of rickets............................. 96
4.25 Summary critique of clinical criteria for diagnosing rickets..........98
4.26 Rates of rickets using different diagnostic methods..................101
4.27 Change in rickets rates from spring to fall..........................102
4.28 Rickets rates (clinical) in spring and fall by age cohort............103
4.29 Logistic regression analysis of rickets and contributing risk
4.30 Logistic regression using univariate analysis to determine risk factors
contributing to rickets.............................................106
4.31 Relationship between rickets (Clin, X and D as binary) and other illnesses
4.32. Illness rates among children with and without rickets................109
4.33. Pneumonia history of children with 25-OH-D <12 ng/ml and
>30 ng/ml......................................................... 110
4.34 Bronchitis history of children with 25-OH-D <12 ng/ml and >30 ng/ml 110
4.35 Partial correlations between 25-OH-D levels and illness..............110
4.36 Blood chemistry and rickets...s..i...................................Ill
4.37 Serum calcium levels for Clin, X and D diagnosis....................112
4.3 8 Zinc levels of children with and without rickets.....................113
4.39 Hgb levels of children with and without rickets......................113
4.40 Partieil correlations between rickets by several criteria and lab results... 114
4.41 Rickets rates among boys and girls.......................:..........116
4.42 Rickets and gender using four diagnostic options (spring)............117
4.43 Rickets and gender using four diagnostic options (fall)..............118
4.44 Intake frequency of calcium-containing sources (spring)..............120
4.45 Relationship between calcium intake and serum 25-OH-D................121
4.46 Vitamin D supplementation pattern....................................122
4.47 Relationship between age of weaning and rickets......................123
4.48 Relationship between rickets and breast-feeding......................125
4.49 Dietary intake during spring and fall................................126
4.50 Zuoyuezi customs.....................................................129
4.51 Mean age of child when first taken outdoors according to month
4.52 Blood levels on three mothers various time periods post-partum.......130
4.53 Method of creating zuoyuezi intensity score..........................130
4.54 Spring data z-score results.........................................134
4.55 Mean dietary intake scores in spring and fall.......................134
4.56 Family possessions and the common signs of prosperity...............136
4.57 Economic status of Yuci respondents.................................137
4.58 Reported main family income.........................................138
6.1 Variables hypothesized to correlate positively with rickets.........182
6.2 Variables hypothesized to have no relation with rickets.............184
6.3 Hypotheses related to the phenomenon of rickets and the health of Chinese
6.4 Health problems associated with vitamin D insufficiency.............204
6.5 Health problems associated with zinc insufficiency..................207
6.6 Health problems associated with iron insufficiency..................208
6.7 Rickets described by the Social Cognitive Theory....................212
INTRODUCTION AND BACKGROUND
Purpose of the Study
Even though rickets is a preventable problem, it is reported that 26.7% of
infants and 15.9% of children under age five in China (Ge et al. 1999) suffer from
nutrition-related rickets (1990-97 survey results). This is a decline from 40.7% in
China from 1983 (Ma et al. 2002). This makes the prevalence of rickets in China
one of the highest in the world.
The biology of rickets is well understood (Greer et al. 1989; Crofton et al.
1995; Kruse 1995; Bishop 2000). Vitamin D- deficiency is recognized as the
primary determinant of rickets (Ho et al. 1985; Xu et al. 2000; Specker et al. 1992).
However, it is not known how ecological, cultural, and dietary variables interact to
contribute to the occurrence of rickets in Chinese infants at a population level. In
order to extend clinical successes in treating rickets into the population, it is
necessary to understand these variables. This study combines survey methodology
with qualitative research methods as a way of both determining and explaining
behaviors that are contributing to rickets, and will extend the research through a
particular focus on Chinas unique customs such as cloistering of the mother-infant
Research in tropical regions such as Nigeria where sun exposure is plentiful
has shown that calcium insufficiency alone can cause rickets (Thacher et al. 1997).
The relative role of calcium insufficiency as a contributor to rickets has not been
studied in the China context, and the topic is still a matter of intense investigation
around the world. This study will address the relative roles of calcium and vitamin
D-deficiency to rickets in north China, and demonstrate the broader consequences of
prevalent vitamin D-deficiency there.
There has been a lack of consistency globally in how rickets is diagnosed.
Consequently, it has been difficult to compare rates of rickets from one country to
another and over time in one country. This study will use methods of diagnosing
rickets in China that are based on international standards, thus allowing for
comparison of rates of rickets on a global scale. An economically feasible and
accurate method of diagnosing rickets in China will be introduced. This will allow
the epidemiology of rickets in north China to be determined, its public health
consequences understood, and its treatment facilitated.
Rickets is a disease that results from many factors, both biochemical and
environmental. However, much of the research on rickets has focused mainly on
physiologic and biochemical processes within the individual. Social ecological
models of rickets which merge the underlying biochemical processes with social and
cultural factors are needed. This is the goal of the present research.
In order to accomplish this goal, collaborative research between the
University of Colorado-Denver department of Health and Behavioral Science,
Shanxi Medical University, Shanxi Women and Childrens Hospital and the
American NGO Evergreen was carried out in Shanxi Province, China from April
2003 to March 2004.
Rationale for this Research
Main Research Questions
Question 1: What is the prevalence of rickets among children in north China?
Question 2: What are the determinants of rickets in Chinese children?
Question 3: What health outcomes are associated with vitamin D-deficiency and
Question 4: How do mothers and Chinese health workers in rural Shanxi
Province, China perceive rickets as a health problem?
Rationale for these Research Questions
Question 1. What is the prevalence of rickets among children in north China?
This study is conducted in Shanxi Province which has been recognized as one
of the two provinces in China with the highest rates of rickets (Kantha 1990). The
specific research location is Yuci Prefecture. Liu Qiaoe, current director who has
worked at the Yuci Maternal and Child Health Center for over twenty years, told the
author that in 1981 70% of children in Yuci had rickets. By 1987 this rate had
declined to 30%. In the spring of 2003 she reported that the rate of rickets in Yuci
was 20%, but this number was reported anecdotally, and had not been determined
Such high rates of rickets have created a high level of awareness and concern
in the Maternal and Child Health system in Shanxi Province, including the
establishment of Rickets Clinics in all childrens hospitals and health centers.
Furthermore, mothers of children suspected of having rickets often undergo
expensive laboratory testing, and those found to have rickets are sold an array of
calcium and other vitamin supplements. It is crucial that the true prevalence of
active rickets in Shanxi Province be determined, and a reliable method of diagnosing
rickets be agreed upon.
Question 2. What are the determinants of rickets in Chinese children?
It is hypothesized that rickets will correlate positively with the following
behavioral and dietary variables. The rationale for these hypotheses follows.
Late introduction of solid foods (after age 5 months)
Late weaning (after 12 months of age)
Failure to drink milk after weaning
Presence of paternal mother-in-law ipo-po) in the home
Good adherence to zuoyuezi (sitting out the month)
Exclusive breastfeeding among infants who do not receive vitamin D
supplements (Specker et al. 1992) (Greer et al. 1989) is a risk factor for rickets
(Bachrach et al. 1979; Elzouki et al. 1989; Pettifor et al. 1997; Kreiter et al. 2000).
Breast-feeding has been aggressively promoted in China in the recent two decades,
especially after beginning the baby friendly hospital initiative in 1991 (World
Health Organization 1989). Traditional practices in the rural areas result in infants
being breast-fed until nearly two years of age. This long duration of breast-feeding
results in relatively late introduction of solid foods or milk supplements that contain
vitamin D (around 8 months of age). Infants should be breast-fed until at least age
12 months (Tsang et al. 1997), but solid foods and milk supplements containing
vitamin D should be given starting at five months of age. By age nine months,
infants are dependent on supplemental foods (beyond breast milk) for over 90% of
their zinc and iron, 62% of their calcium and 35% of their protein. Timely and
nutritious supplemental foods are crucial for infants during their first year of life.
The rate of rickets was shown in a hospital-based study to be higher among
breast or mixed-fed infants versus artificially fed infants in China (Chen 1994).
However, as Prentice has reported, breast-feeding beyond 12 months of age, and
even up to two years, results in greater gains in length among north China children
(Prentice 1996). This study investigates the advantages and disadvantages of breast-
feeding in relation to growth and rickets, as well as the role of timing of introduction
of solid foods and milk drinking habits after weaning. In previous research delayed
weaning and delayed provision of solid foods were found to correlate with rickets
(Strand et al. 2003).
In patrilocal cultures (where after marriage women live in their husbands
extended family household), such as China, a woman is expected to be subservient to
the authority of her husbands mother (called po-po in Chinese) (Morsy 1978;
Kinoshita 1988; Gittlesohn 1991; Wolf 1985). During her years of child rearing, the
mother will have limited control and autonomy and po-po may engage in practices
which are not beneficial to her infant (Young 2000). If the mother-in-law determines
the method of raising and feeding the child, providing supplements to and educating
mothers about infant and child nutrition may only lead to an effect a generation later.
So cultural variables may affect the childs nutritional situation and must be
considered in any study of nutrition. This is the rationale for this study design
merging epidemiological and social science approaches. The role these customs play
in the development of rickets also needs to be understood in order to be able to
successfully prevent rickets in this context.
The Chinese custom of zuoyuezi (sitting out the month post-partum) (Cheung
1997) is strictly adhered to in north China. According to this custom, mother and
infant must spend the first month (sometimes the first 100 days) indoors on their bed.
Lack of sun exposure on their skin interferes with vitamin D synthesis for the infant
and nursing mother. Season of birth and northern location (~38 north) also
influences the amount of sun available for photoconversion of vitamin D metabolites.
Namgung et al. (1998) have shown that infants bom in the winter have serum 25-
OH-D levels three times lower than infants bom in the summer, if they are exposed
to the sun. Infants require 24 minutes per day of face exposure to the sun to maintain
adequate serum 25-OHD levels (11 ng/ml) (Ho et al. 1985). Traditional customs
such as zuoyuezi put infants in China at risk for developing rickets.
By clinical criteria, boys have higher rates of rickets in China than girls
(Strand et al. 1999). Other research has also shown a trend toward more rickets
among boys than girls. Chen (1994) reports a higher prevalence of rickets during the
first 18 months of life among boys (6.8%) than girls (4.6%) in Shanghai. Reporting
on cases of rickets in a North Carolina Medical Center from 1990-99, Kreiter et al
(2000) found 57% of the cases to be males. Elzouki et al. (1989) found the ratio of
males to females among Libyan children with rickets to be 10/6. In a longitudinal
study of vitamin D deficiency in Finland (n=l 0,366), Hypponen et al. (2001) found
the suspected prevalence of rickets to be 2.4% for boys (n=126) and 1.8% for girls
It is hypothesized that rickets will not be associated with:
Maternal knowledge of how to prevent rickets
Socio-economic status (SES) has been shown to be a powerful indicator of
poor health outcomes in varied contexts (Adler et al. 1994; Durkin et al. 1994).
Chen (1994) reports that poverty is a predictor of rickets in China. However, it is
hypothesized that people can use simple, local resources and customs to accomplish
prevention of rickets. This argument is based on the principles of positive deviance
research (Zeitlin 1991; Guldan et al. 1993). Positive deviance suggests that in every
context people are able to successfiilly meet their health needs, even in contexts of
poverty. While some poor people may be suffering from poor health, others will not,
thus obviating any economically explained health gradient. This is not to deny the
relationship between poverty and poor health, which has been demonstrated in many
contexts. The hypothesis is based on experience in China, where limited variability
in economic status among the people of a given community makes it difficult to
detect the role of economics in rural health.
The value of increased education cannot be questioned, especially the value
of increasing health information available to people. However, it is difficult to find a
direct relationship between knowledge about health and behavior change (National
Dairy Council 1985; Shepherd et al. 1985; Glanz et al. 1997). Previous research by
the author and others has found no correlation between maternal knowledge of
rickets and successful prevention (Strand et al. 2003), but other Chinese researchers
have found such a relationship. The relationship between childhood rickets and the
mothers knowledge about rickets and her overall educational level merits further
investigation, which this study has done.
Question 3. What health outcomes are associated with vitamin D-deficiencv
Rickets will be primarily due to vitamin D-deficiency, but also related to
Calcium deficiency will be high (>50%)
Vitamin D-deficiency will be high (>50%) and will correlate with rickets,
pneumonia, and diarrhea
Zinc-deficiency will be associated with pneumonia, diarrhea and stunting
Ht-for-Age z-scores will be lower among infants with rickets than among
Wt-for-Age z-scores will be normal for infants with rickets
Among 1248 girls age 12-14 in Beijing, 45.2% were found to be vitamin D-
deficient in winter, and 6.7% in summer (Du et al. 2001) (serum 25-OH-D level
<12.5 nmol/L). It appears most of the cases of rickets in China is due to vitamin D-
deficiency. The main cause is lack of sun exposure and lack of vitamin D through
oral sources. However, Thacher et al. demonstrated that nutritional rickets in
Nigerian children was not due to vitamin D deficiency, but was due to calcium
deficiency (Thacher et al. 2000), pointing out the need to consider the role of
calcium-deficiency in rickets. This is likely due to Nigerias proximity to the
equator. Even with high levels of melanin in their skin, normal sun exposure
supplies adequate vitamin D synthesis. Varying ecological contexts may explain the
different etiologies of rickets in Nigerian and Chinese populations. A nationwide
survey in China found that Chinese children only obtain 42.3% of RDA for calcium
(Wang 1997) and milk-drinking is not common. The north China diet is high in
cereals that bind calcium and make it unavailable (Sly et al. 1984). Vitamin D
deficiency as the main cause of rickets in Chinese children is the most plausible
explanation, but the role of calcium deficiency merits exploration as well.
Lead is a doubly charged cation like calcium, and is able to compete with
calcium in calcification. This results in poorly calcified bones and could lead to
rickets. Blood lead levels are tested in this study to test this possibility.
Infants with rickets and/or vitamin D-deficiency suffer in a myriad of ways.
In addition to the skeletal abnormalities arid compromised strength and mobility,
rickets is also associated with co-morbidities in infants such as pneumonia (Bamess
1987; Muhe et al. 1997), stunting, and gastrointestinal infections, all known killers of
infants and children. Vitamin D-deficiency in infancy is also suspected to lead to
adult morbidities such as Type I diabetes (Hypponen et al. 2001), hypertension
(Krause et al. 1998), and breast, ovarian, prostate and colorectal cancers (Vieth
1999:849). Patients with tuberculosis were found to have a deficiency of vitamin D
receptors (Wilkinson et al. 2000), so there is even a relationship between vitamin D-
deficiency and tuberculosis. The overall contribution of vitamin D-deficiency to
morbidity and mortality may greatly the levels that are commonly assumed.
Zinc insufficiency is associated with pneumonia (Brooks et al. 2004),
diarrhea, and stunting (Umeta et al. 2000; An et al. 2001). Therefore it is possible
that zinc insufficiency may confound the relationship between vitamin D-deficiency
and rickets. More likely, low zinc intake contributes to the context of overall
compromised nutrition in which rickets is occurring (Chen et al. 1992). The role of
zinc in the immune system implicates low zinc intake as a possible contributor to
many childhood illnesses (Shankar et al. 1998).
Stunting is one of the consequences of rickets in infants. During the period
of economic reforms (1980s), there was continuous overall improvement in the
growth of children, but disparities between rural and urban children were substantial
(Shen et al. 1996). During the 1980s, this disparity in growth between rural and
urban children was reduced. But by 1990, this improvement changed directions and
urban development began to outpace rural development again. Shen et al. (1996)
report 38% of rural children aged two to five had moderate stunting and 15% had
severe stunting, as compared with 10 percent and three percent among urban children.
It is likely that stunting among rural children is related to the prevalence of rickets.
Question 4. How do mothers and Chinese health workers in rural Shanxi
Province. China nerceive rickets as a health problem?
It is hypothesized that:
Mothers will consider rickets unimportant as a health problem
Local practices increasing rickets (zuoyuezi, extended breast-feeding) will be
shown to have an adaptive advantage even though some features may be
It is hypothesized that Chinese mothers are aware of the condition of rickets
but it is not their first concern when they think of their childs health. Rickets is a
chronic disease and its seriousness usually lessens as the child grows. Therefore
mothers may see rickets as somewhat normal, and of less concern to them than acute
illnesses that threaten the childs well-being such as diarrhea and respiratory
There are also consequences of chronic vitamin D-deficiency for women
themselves. Ninety percent of adult bone mass is established by age 19, so that
vitamin D-deficiency in childhood can result in compromised bone health, possibly
leading to higher rates of osteoporosis later in life (Vitamin D 1998), and worsening
of osteoarthritis (McAlindon et al. 1996), especially among women. Beijing
residents over age 60 show a prevalence of osteoporosis of 60-70% for women and
25-35% for men (Ge et al. 1999). Osteoporosis and osteoarthritis that arise in mid-
to late-life may not be associated by Chinese mothers with former vitamin D-
deficiency. Rather, according to Traditional Chinese Medicine concepts, these
problems are blamed on having had their joints exposed to cold wind or cold water
when they were pregnant or nursing (many years before). All these factors make it
challenging for mothers to value vitamin D-deficiency and/or rickets as health
Conceptual Framework of this Study
Social and Cultural Epidemiology
The study reported here is an epidemiologic study of rickets in north China
that has been broadened in order to explore the cultural determinants of this
condition and other common diseases of north China such as respiratory infection
and diarrhea. This is based on the belief that disease states are intrinsically tied not
only to pathogens and nutritional states but to cultural factors that increase exposures
and disease incidence.
This perspective makes the study unique in the context of biomedical
research on rickets. In a PUBMED search hundreds of references to rickets were
found, but most of the studies were clinical or biomedical in nature. One area of
mutual concern in these articles is the recent increase in rates of rickets among dark-
skinned, breast-fed infants in the U.S. (Ho et al. 1985; Kreiter et al. 2000; Welch, et
al. 2000). The general consensus is that vitamin D supplementation is needed for
these vulnerable populations but little attention is paid to the role cultural factors
play in this increase. Recent research has also identified the role of calcium
insufficiency without vitamin D deficiency in the development of rickets (Thacher et
al. 2000). Although this research investigated the role of Muslim religious beliefs in
the development of rickets, and assumed that mothers were constrained by food
restrictions, the role of these cultural practices were assumed and not measured,
failing to take advantage of theories derived from the social sciences to understand
This heavy emphasis on a primarily biomedical approach to rickets research
affects clinical care. At the 2000 Annual Meeting of the Dietetics Association Dr.
Irene Sills, a pediatrician from Albany Medical Center, New York, reported that
some patients with rickets referred to her had been told by orthopedic surgeons that it
was necessary to surgically fracture the childs legs and reset them to cure their
rickets. These physicians overlooked nutritional and environmental factors which
had caused these children to develop rickets, and that these same factors would also
preclude the healing of fractured legs, to say nothing of the risk and discomfort of
the unnecessary procedures. There may be cases when surgery is necessary to
correct the skeletal problems of rickets, but these are rare (Fischer 2004).
Early in the latter half of the 20th century public health research focused
increasingly on individual factors and risks to identify causes for health problems in
society. Health and disease in populations was often explained primarily in terms of
the characteristics of individuals. This reductionist approach has been heavily
criticized (Mechanic 1995), and in recent years it has become recognized that
ecological and contextual factors contribute to individual health as much as
individual factors. Failure to appreciate these contextual factors may impair
effective health promotion (Berkman 1995; Stokols 1996; Diez-Roux 1998; Ehiri et
al. 1999). This is particularly so in the global health context (Scrimshaw 2001).
Following is a review of the theoretical perspectives drawn from the social and
behavioral sciences that have provided a conceptual framework for the design and
interpretation of this study.
Social Cognitive Theory
Social Cognitive Theory (SCT) explains human behavior as a dynamic and
reciprocal interaction among the individual, his/her environment and his/her health
behavior (Baranowski et al. 1997). The model brings together the personal factors,
environmental factors and personal behaviors that work together to determine
peoples perspectives on health and their health-seeking behavior. Social Cognitive
Theory involves many concepts as described below (Table 1.1).
able 1.1. The Concepts of the Social Cognitive Theory.
Concept Definition Implications
Environment Factors physically external to the person Provide opportunities and social support
Situation Persons perception of the environment Correct misperceptions and promote healthful norms
Behavioral capability Must know what the behavior is and skills to perform it Promote mastery learning through skills training
Outcome Expectations Beliefs about likely results of action Model positive outcomes of healthful behavior
Outcome Expectancies The values that the person places on a given outcome, incentives Present outcomes of change that have functional meaning
Selfcontrol Personal regulation of goal-directed behavior or performance. Depends on self-efficacy. Provide opportunities for self- monitoring, goal setting, problem solving, and self-reward
Observational learning (modeling) Beliefs based on observing others like self and/or visible physical results Include credible role models of the targeted behavior
Reinforcements (rewards and punishments) Responses to a persons behavior that increase or decrease the chances of recurrence Promote self-initiated rewards and incentives
Self-efficacy Confidence in ability to take action and persist in action Approach behavioral change in small steps to ensure success; seek specificity about the change sought
Emotional coping responses Strategies or tactics that are used by a person to deal with emotional stimuli Provide training in problem solving and stress management
Reciprocal determinism Behavior changes result from interaction between the characteristics of the person, their behavior and the environment in which the behavior is performed; change is bidirectional Consider multiple avenues to behavioral change including environmental, skill and personal change
Social Cognitive Theory takes into consideration the fact that people exist in a
dynamic milieu of relationships, learning, observing, and changing expectations. In
this regard it serves as an appropriate theoretical perspective for analyzing health
behaviors in the collectivist Chinese society. Rickets is more than just an illness that
can be easily diagnosed in the clinic and treated with a medication. It is a disease
that occurs over a long period of time and which results in part from behaviors that
are culturally conditioned.
Specific dimensions of the Social Cognitive Theory are not being tested in
this study, but they have informed the formulation of the research questions and the
survey items. Discussion of each of the concepts of the Social Cognitive Theory
with regard to the results of this study will be offered in the Discussion chapter of
Cultural Explanatory Model of Illness
The Cultural Explanatory Model of Illness, formulated by Kleinman and
Kleinman (Skelton et ai. 1991; Kleinman et al. 1997), is based on the question of
how individuals explain and make sense of health and disease issues. In the global
context, this model is helpful in recognizing that a persons culture has its own set of
interests, emotions and biases (Fadiman 1997) which influence the health
perspectives of people in that culture.
Chinese people draw on a long and rich tradition of indigenous medical
knowledge and practice when they attempt to treat and prevent illness. Indigenous
concepts of health and illness overlap and are integrated with nearly all aspects of
social life. Kinship solidarity is highly valued in China (King 1991). This involves
filial piety, male authority, respect for elders, and the role of the paternal mother-in-
law in the home. This value influences how rickets as an illness is perceived and
controlled in China, especially in how it may limit the autonomy of the mother in her
child rearing practices. As a new member of the family, the new mother is expected
to fit in, and to help build family solidarity. This is done by complying with
extended family expectations, even with regard to child rearing. A second health
value is the abstract concept of equilibrium. Equilibrium can also be described as
rhythm, order or balance (You 1994). There is the rhythm of the four seasons, and
the rhythm of the human body within the rhythmic environment. Taking infants
outside may upset their cosmic rhythm and put them out of balance, thus making
them vulnerable to illness.
It is crucial for the mother to regain her rhythm after the trauma of delivering,
so there are many customs meant to accomplish this, such as the balancing of hot and
cold foods and prescriptions of what may be eaten post-partum. Zuoyuezi, the
custom of post-partum cloistering, is itself a balance-seeking practice. The concept
of balance among Chinese people is extended to generally mean a sense of order,
good health, and freedom from conflict (You 1994).
One of the three explanations for illness in the Traditional Chinese Medicine
(TCM) system is the intrusion of environmental factors into the body which upset the
bodys processes or rhythms (e.g., ideas about wind-fire, cold-hot, damp-arid). The
health of infants may be compromised when their rhythm is upset by cold wind on
their bodies. While taking infants outside may secure the advantage of sun-catalyzed
vitamin D synthesis, it may also upset the rhythm of these infants bodily processes.
The application of the cultured explanatory models approach is particularly
useful in determining how Chinese people view the relationship between the disease
of rickets and nutrition as well as between rickets and behavioral practices such as
swaddling infants and keeping them indoors. For example, local understanding of
rickets is related to the fact that in Chinese thought the bone represents fortitude, the
quality of ones character and courage (Tung 1994). So Chinese people may be
more concerned about the way rickets or osteoporosis reflects negatively on the
persons character than how it affects their ability to labor or exert themselves
Typically, personal cultural models are elicited by interviews. This study did
not attempt to determine individual mothers explanatory models for rickets. Rather,
the effort was to uncover the models employed by different groups of participants
mothers, mothers-in-law, and health care providers in a focus group setting. This
strategy permitted me to understand the particular models of rickets held by these
different groups and thus appreciate the degree to which these models may impair, or
enhance, subsequent efforts at health promotion.
Health and illness patterns in populations cannot be explained by biomedical
descriptions of disease. Cultural and social variables influence human behavior,
which relate to health outcomes. The tools and theories of the social sciences are
able to enhance research in ways that provide context and explanatory power to
epidemiologic observations (Dunn 1986; Inhom 1995). This section has introduced
the Social Cognitive Theory and Cultural Explanatory Models of Illness as two
social perspectives from which to analyze rickets among north China children.
These two perspectives informed the design of this research.
Arrangement of the Dissertation
This introduction is followed by a chapter on the theoretical and empirical
background, where previous research in the area of rickets and vitamin D deficiency
is introduced, as well as the physiology of rickets. Because of the impact of
Traditional Chinese Medicine on the health behaviors of Chinese people, and their
relative unfamiliarity to Westerners, it is introduced at length in this chapter as well.
The materials and methods chapter provides details about how this study was
conducted, including a defense of the validity of the methods. The results are
divided into two chapters, quantitative and qualitative. The qualitative results
chapter is rich in contextual and cultural information pertinent to rickets, and relies
extensively on the section on Traditional Chinese Medicine for explanation. The
discussion chapter summarizes the results of this study and integrates these results
with those of other research on rickets. A proposed intervention for controlling
rickets in China is also introduced. The appendices include survey forms used in
data collection as well as important contextual information that is not related directly
to the research questions, but may be of relevance to readers who have a broad
interest in China, including an introduction to the Rural China Health Care System
and an introduction to the 2003 China SARS epidemic and how it impacted this
THEORETICAL AND EMPIRICAL BACKGROUND
The Biology of Rickets
Basic Introduction to Rickets
Bone mineralization requires adequate circulating levels of vitamin D
metabolites to absorb calcium from the small intestine and to assist in the process of
bone metabolism (Fomon 1993). Under ideal conditions the body absorbs 40-50%
of the calcium consumed, but without adequate vitamin D, only 10-15% is absorbed
(Tufts University 1998). Vitamin D may be obtained from food sources such as fish
liver oils and eggs. However, it is difficult to obtain adequate vitamin D from
common dietary sources.
Vitamin D is also manufactured in the skin through the action of sunlight
(Webb et al. 1989; Vitamin D 1998). Since only about 10% of the vitamin D in the
body is derived from dietary sources (Hartman 2000) exposure to the sun is crucial.
Time spent outdoors is directly related to serum 25-OH-D levels (Kim et al. 2000).
Typically,the sun exposure of a person in a bathing suit of 1 minimal erythema dose
(which causes a slight pinkness to the skin) is the same as ingesting 20,000 IU of
vitamin D (Holick 2003). Therefore, exposure of hands, face, and arms or arms and
legs to 25% of a minimal erythema dose (about 5-15 minutes between 11 AM and 2
PM in Boston) will provide an adequate amount of vitamin D (Holick 2003).
Season also affects vitamin D levels. Infants bom in the winter have serum
25-OH-D levels three times lower than infants bom in the summer, if they are
exposed to the sun (Namgung et al. 1998). Failure to be exposed to the sun during
the first months of life when the skeleton is experiencing rapid growth causes infants
who are not given vitamin D supplements to develop rickets (Elzouki et al. 1989).
The physical symptoms of rickets include bowed legs or knock-knees, pectus
carinatum (chicken breast), pearling (bead-like enlargements on the costocondral
juncture of the ribs), inward bending of the interior ribs, frontal bossing, and/or
delayed eruption of the teeth. Deformities in the legs delay walking for the infant,
and affect the childs ability to walk and run. Deformities in the ribs decrease
respiratory capacity and contribute to increased respiratory infections. In severe
cases, pelvic deformities in girls can lead to increased risk of obstructed labor later in
life (Frisancho 1996).
Vitamins. Minerals and Bone Metabolism
Vitamin D and Calcium Metabolism and Storage. 7-dehydrocholesterol
(provitamin D) is irradiated in the skin to form cholecalciferol (vitamin D3). In the
event of prolonged exposure to the sun, vitamin D toxicity is prevented by the UV-
induced isomerization of cholicalciferol to biologically inert products such as
lumisterol and tachysterol. Cholecalciferol (vitamin D) is carried from the skin to
the bloodstream by vitamin D-binding protein. Vitamin D-binding protein is not
responsible for transporting vitamin D into cells; rather, 25-OH-D and l,25(OH)2D
enter cells in their free unbound forms (Holick 1995). Vitamin D is primarily
sequestered in the liver.
Calcium in the body is stored primarily in bones, but also in most body
tissues. Homeostasis is maintained primarily by the action of the parathyroid gland,
which secretes parathyroid hormone (PTH) in response to low serum calcium levels.
PTH stimulates hydroxylation of cholecalciferol (vitamin D) in the liver producing
25-OH-D. In the kidneys, hydroxylase performs a second hydroxylation of 25-OH-
D to produce the active form of vitamin D, l,25(OH)2D, as well as a host of other
vitamin D metabolites. 1,25(OH)2D then increases absorption of calcium in the
intestines. Calcium absorption in the gastrointestinal tract occurs primarily by a
saturable, active, transcellular process, and is vitamin D dependent (Tsang et al.
1997). But as calcium intake increases, paracellular passive diffusion of calcium
also occurs. At the same time, PTH stimulates calcium resorption from the bones.
PTH also acts to increase phosphorous excretion and decrease calcium excretion
from the kidneys. All these processes increase serum calcium levels, thus protecting
the heart and other muscles from the consequences of hypocalcemia. l,25(OH)2D
inhibits the synthesis of PTH directly through an interaction with the preparathyroid
hormone gene or secondarily, through an increase in circulating ionized calcium
(Tsang et al. 1997).
l,25(OH)2D is the active vitamin D metabolite, and 25-OH-D is biologically
inert before hydroxylation. But serum concentrations of 25-OH-D are stable and
serve as the best measure of vitamin D status. The normal level for serum 25-OH-D
used in this study is 12 ng/ml. There are experts calling for the recognition of 20
ng/ml as the normal level for serum 25-OH-D (Holick 2003), but this has not been
recognized as of yet.
Calcium deprivation promotes mild hyperparathyroidism which stimulates
the production of l,25(OH)2D, resulting in vitamin D deficiency. Low calcium
intake increases hepatic catabolism of 25-OH-D; therefore serum 25-OH-D
decreases due to increased serum PTH. This may explain the prevalence of rickets
found in sunny countries such as Nigeria where the diet is low in calcium, and high
in high-phytate foods which bind calcium and decrease its bioavailability.
Chinese researchers in Hong Kong have shown that children on a chronically
calcium-deficient diet are able to increase their calcium absorption compared to
children on a normal calcium diet (Lee et al. 1995). They found true fractional
calcium absorption (TFCA) to be higher among the group on a low calcium diet
(64.3 SD 10.7%) than among the study group on a normal calcium diet (55.6 SD
12.7%; p=0.015). Their results suggest that children with adequate vitamin D status
can adapt to a change in calcium intake by adjusting the efficiency of TFCA. Nordin
concurs, arguing that calcium requirements are highly variable and depend on many
factors (Nordin 2000).
While it is true that as calcium intake increases the total amount of absorbed
calcium also increases, the overall percent absorbed decreases. So in the Lee study,
while the low-calcium intake group did have higher true fractional calcium
absorption, the absolute amount of calcium absorbed in both groups was the same
(roughly 250 mg). This 250 mg still fell far short of the RDA for calcium of 800
mg/d, suggesting that the increased absorption in the low-calcium group had little
effect in improving calcium status.
There is new concern for the role of calcium-deficiency in the development
of rickets, and this study seeks to determine the role it plays in the rickets observed
in Chinese children.
Bone tissue has several types of cells. Osteoblasts function to synthesize the
extracellular matrix of bone, deposit this matrix in the appropriate amount and place,
and initiate its mineralization. Once mineralized in the bone, these cells are known
as osteocytes. Osteoclasts are the cells that function to resorb bone (Krane et al.
In concert with PTH, l,25(OH)2D mobilizes calcium stores from bone by
inducing the dissolution of bone mineral and matrix. l,25(OH)2D induces stem cells
to differentiate into osteoclasts which resorb bone.
Parathyroid Hormone. Parathyroid glands have a vitamin D-dependent
calcium-binding protein and possess nuclear receptors for l,25(OH)2D (Holick 1995).
This suggests that the parathyroid gland responds to both serum calcium levels and
l,25(OH>2D levels. In response to low serum calcium levels, the parathyroid gland
secretes parathyroid hormone, which stimulates hydroxylation of vitamin D to
l,25(OH)2D. The l,25(OH)2D then increases serum calcium levels through several
mechanisms. PTH is also excreted in response to phosphorous levels and stress
Chapuy et al. found serum PTH held a stable plateau concentration at 36
pg/ml as long as serum 25-OH-D values were higher than 78 nmol/1 (31ng/ml), but
increased when the serum 25-OH-D value fell below this threshold. When the 25-
OH-D concentration became equal to or lower than 11.3 nmol/1 (4.6 ng/ml), the PTH
values reached the upper limit of normal values found in vitamin D replete subjects.
Research in Shenyang, north China found an inverse association between 25-
OH-D and PIH concentration among elderly men and women, as well as among
male adolescents (Yan et al. 2000). However this association has not been found in
all studies. Vitamin D-deficient individuals may not necessarily exhibit elevated
PTH. Little correlation between 25-OH-D levels and PTH suggest that calcium
alone is mostly responsible for activating the parathyroid gland. As long as calcium
serum levels are normal, the parathyroid gland will not respond to vitamin D levels.
PTH receptors have been identified on osteoblast-like cells, but not on
osteoclasts. It may be that PTH is exerting its resorptive effect on bone indirectly by
stimulating one or more activities of osteoblasts, which in turn facilitates osteoclastic
bone resorption (Amaud 1991).
Mature osteoblasts possess vitamin D-receptors and respond to l,25(OH)2D
by increasing the expression of alkaline phosphatase (Holick 1995). Because
alkaline phosphatase, present in osteoblasts and other cells, can catalyze the
hydrolysis of inorganic pyrophosphate at neutral pH, this enzyme could play a role in
the regulation of mineralization by controlling the concentrations of pyrophosphate.
This is not totally clear. What is clear is the consistent observation of elevated
alkaline phosphatase when bone mineralization is occurring or trying to occur. ALP
levels are used as an important indicator of active rickets in this study.
The Physiology of Rickets
Vitamin D and calcium are absolutely necessary for strong and healthy
growing bones (Bishop 2000). When vitamin D is insufficient, it results in
nutritional rickets in growing individuals, and osteomalacia in nongrowing
individuals. Rickets appears to develop in three stages. First, inadequate vitamin D
intake and/or decreased exposure to sunlight results in low serum 25-hydroxyvitamin
D levels. Consequently intestinal absorption of calcium and phosphorous declines.
Second, in response to low levels of circulating calcium, parathyroid hormone is
released stimulating vitamin D synthesis in the kidneys. This stimulates bone
alkaline activity, which is associated with attempts at bone growth (Crofton et al.
1995); unfortunately; the result is the resorption of calcium from the bones as an
attempt to normalize calcium levels in the blood (Kruse 1995). Third, despite
increased release of parathyroid hormone, adequate quantities of vitamin D cannot
be produced in the kidneys, so that intestinal absorption of calcium and phosphorous
drops even further.
Vitamin D-deficiency rickets (VDR) occurs as a result of low vitamin D intake
in association with low exposure to sunlight. Low serum 25-OH-D levels, secondary
hyperparathyroidism (stimulating osteoblastic activity), and high bone turnover are
the hallmarks of rickets in children (Kruse 1995). Elevated bone alkaline
phosphatase (ALP) activity, which is known to correlate with growth velocity
(Crofton et al. 1995), has traditionally been recognized as one of the key biochemical
characterizations of rickets (Kruse 1995).
In rickets, the main problem is insufficient mineralization of newly forming
matrix (cartilage and bone). The epiphyseal growth plate increases in thickness, is
cupped, and reveals a haziness at the diaphyseal border. This is due to decreased
calcification of the hypertrophic zone and inadequate mineralization of the primary
spongiosa. The cartilage at the epiphyseal growth plate thickens because the cellular
sequence matures too slowly and the cartilage cells are disorganized (Goldring et al.
1995). Vitamin D deficiency slows chondrocyte maturation, which causes the
growth plate to expand greatly in height (Hunt et al. 1994).
Rickets appears to develop in three stages. First, inadequate vitamin D intake
and/or decreased exposure to sunlight results in low serum 25-hydroxyvitamin D
levels. Consequently intestinal absorption of calcium and phosphorous declines.
Second, in response to low levels of circulating calcium, parathyroid hormone is
released stimulating vitamin D synthesis in the kidneys. This stimulates bone
alkaline activity, which is associated with attempts at bone growth (Crofton et al.
1995); unfortunately; the result is the resorption of calcium from the bones as an
attempt to normalize calcium levels in the blood (Kruse 1995). (Ho et al. 1985;
Specker et al. 1992).
Although Greer and Marshall have shown that white breast-fed infants in the
United States do not need vitamin D supplementation (Greer et al. 1989), it is
generally agreed that vitamin D supplementation before six months of age for breast-
fed infants is beneficial to assure normal bom growth (Bishop 2000). A 2003 report
from the American Academy of Pediatrics (AAP) entitled, "Prevention of Rickets
and Vitamin D Deficiency: New Guidelines for Vitamin D Intake", recommends that
all healthy infants beginning during the first two months of life should receive
vitamin D supplementation to prevent rickets and vitamin D deficiency. The
recommendation includes breastfed infants. According to the report, "It is
recommended that all infants, including those who are exclusively breastfed, have a
minimum intake of 200 International Units (IU) of vitamin D per day beginning
during the first 2 months of life. In addition, it is recommended that an intake of 200
IU of vitamin D per day be continued throughout childhood and adolescence,
because adequate sunlight exposure is not easily determined for a given individual."
A copy of the report, first published in the April 2003 issue of Pediatrics, can be
found at http://www.aappolicy.org.
Rickets in infants is dependent on the mothers prenatal vitamin D status.
Infants bom of vitamin D-deficient mothers will be bom with correspondingly low
levels of circulating vitamin D levels in their blood (Namgung et al. 1998). 25-OH-
D levels in an infant exclusively breast-fed with no supplements can be depleted to
deficient levels within 8 weeks after birth (Tsang et al. 1997).
The breast milk of vitamin D-deficient mothers is also low in vitamin D;
though human breast milk is a poor source of vitamin D (4-60IU/L) under any
conditions. Without adequate sunshine exposure or vitamin D supplements for the
infant, even nutritious breast milk is unable to provide the infant with sufficient
A high fiber or vegetarian diet impedes calcium absorption and is implicated
in the development of rickets (Fomon 1993). Lignin, an important component of
wheat fiber, combines with bile acids and increases bile acid excretion. If vitamin D
attaches to the fiber-bile acid complex, it too might be unavailable. Chinese people
have a high fiber diet. This may partially explain their chronic calcium deficiency
(Chang et al. 1994) and high incidence of rickets. Poor absorption of calcium can
also lead to rickets. Phytate, a component of chlorophyll, is ubiquitous in vegetable
foods and whole grains (Zubay 1988). Infants fed vegetarian diets consume high
levels of phytate, resulting in decreased absorption of calcium.
Although vitamin D-deficiency is associated with rickets in the young and
osteoporosis in the elderly, it is actually prevalent among all age groups in America.
Thomas et al (Thomas et al. 1998) found 57% of a group of290 patients to be
vitamin D-deficient. Adequate sun exposure and vitamin D supplementation is
necessary at all stages of life, particularly for women.
Vitamin D-deficiency. Vitamin D-deficiency is generally recognized as a
serum 25-OH-D less than 30 nmol/L (12ng/ml) (Scanlon 2001). In northeast China,
13-48% of adults were found to have 25-OH-D levels below 25 nmol/L in the spring
season (Yan et al. 2000).
There has been a group calling for more attention being paid to the high
prevalence of subclinical vitamin D-deficiency. Vieth makes the claim that for
unclothed early humans, circulating 25-OH-D levels were much higher than is
regarded as normal now. They suggest increasing vitamin D intake to 800 IU/d
In related research, there is a theory that states that vitamin D deficiency in
utero, or in early infant development, programs the metabolic and endocrine systems
in ways that influence adult life. By programming is meant changes in structure and
function caused by environmental stimuli during critical periods of development.
With regard to bone development, there are two groups of researchers contributing to
Cooper et al (1997) have argued that growth during prenatal and early
postnatal life may be a determinant of peak adult bone mass. Studying elderly
residents in the UK, they found a statistically significant association between weight
at 1 year and bone mineral content (but not BMD) in their 70s. They conclude that
skeletal growth may be programmed during intrauterine or early postnatal life. Of
interest to the China context is the possibility that rickets in infancy may predispose
people to osteoporosis in adulthood.
An Australian research group led by McGrath build their case for vitamin
D-deficiency imprinting on the higher rates of multiple sclerosis, hypertension, and
other illnesses among northern dwellers, attributing these conditions to vitamin D-
deficiency in early development. McGrath has found vitamin D-deficient rats to
develop neurological symptoms resembling that of patients with schizophrenia
(McGrath 2001a; McGrath 2001b).
The mechanism by which this programming might occur will need to be
presented before it will be broadly accepted by the research community and
influence pediatric clinical practice.
Epidemiology of Rickets
Rickets Increasing in the USA
In 1925 Eliot (Eliot 1925) demonstrated cod liver oil (vitamin D
supplementation) and/or sun exposure was able to prevent rickets. Within a few
years, most milk processors began supplementing their milk with vitamin D. A
gradual decline began and by the 1960s, rickets was rare and regarded as a medical
curiosity (Harrison 1966). Today, public health officials and food producers have
successfully eliminated vitamin D-deficiency rickets as a health problem for most
Americans. Yet there are still groups of people in the U.S. at high risk for
In a study in the United States, 43 patients with rickets were reviewed
retrospectively (1986-2002). These patients were found to be primarily children
with darkly pigmented skin who were exclusively breastfed (DeLucia et al. 2003).
This alone is cause for concern, for the Black population in the U.S. is large, and to
their credit, increasing numbers are breast-feeding their children. These children
with rickets were also found to be weaned to low-dairy diets (86%), and without
vitamin D supplementation (15%). Only 22% had serum 25-OH-D levels below 15
ng/ml, and one of them improved with calcium alone, suggesting that calcium
insufficiency played an important role in their development of rickets. As will be
described later, vitamin D and calcium insufficiency both play a role in the
development of rickets under different age, environmental and dietary conditions.
Reports of an increasing number of cases of rickets in the US (Shah et al.
2000; Bachrach et al. 1979; Kreiter et al. 2000) have not been accompanied by
published prevalence rates, probably because the cases are two few to be statistically
meaningful. It seems the new cases being discovered are unique and are not
evidence of an epidemic. However, this increased attention to rickets has helped
focus attention on a previously neglected population at-risk for rickets, and has
raised the issue of vitamin D-deficiency.
Rickets in China
Vitamin D deficiency rickets has been reported to be Chinas number two
most common nutritional disease of children (Xu et al. 2000). The most
conservative prevalence rate of rickets among children under age five in China is
reported as 15.9%, with rates among infants of 26.7% (Ge et al. 1999). This is a
decline from 40.7% in China in 1983 (Ma et al. 2002).
The discipline of public health has struggled for years to know how to best
measure the impact of various diseases on the community. Child (ages 0-5)
mortality rate and life expectancy have been commonly used as measures of the
health of a population.
One of the tools that has been devised to measure the health of the public is
the DALY (Disability-Adjusted Life Years). The DALY is used by the World Bank
in assessing global disease burden (Murray et al. 1990; The World Bank 1993;
Gwatkin et al. 1999; Hyder et al. 2000). The DALY score which is generated
determines the relative importance (burden) of different diseases in a population by
factoring in their frequency (incidence or prevalence) and severity (the mortality and
the extent and length of serious morbidity). One DALY represents one lost year of
Nationwide Burden of Disease research in China has not addressed the
problem of rickets, so it is impossible to know the burden of disease due to rickets
relative to other conditions. The World Bank reported that for China in 1990,
musculoskeletal conditions were responsible for 1,330,000 DALYs lost. What
portion of these DALYs lost were attributed to rickets or osteoporosis is unknown as
neither condition was included in the analysis, even though both are known to be
highly prevalent in China, especially in the north (Ge et al. 1999).
In the absence of published data on burden of disease due to rickets in China,
a relative burden of disease due to rickets was calculated relying on other published
research. Hyder and Morrow did a large study of burden of disease study in Pakistan
(Hyder et al. 2000), which included rickets. In Pakistan rickets was responsible for
0.92 HeaLYs lost per 1000 people (the HeaLY is very similar to the DALY). For
comparison, the leading diseases in Pakistan were diarrhea (38.95 HeaLys), lower
respiratory tract infection in children (36.76), tuberculosis (35.74), birth diseases
(33.13), and injuries (32.94). Although much less of a burden than these common
diseases, rickets in Pakistan ranked higher than such highly prioritized health
problems as bronchitis, hepatitis, malaria and vitamin A deficiency. Rickets is thus
responsible for a considerable burden of disease in Pakistan.
Using Hyder and Morrows variables for calculating HeaLYs lost due to
rickets and an under-five prevalence of rickets for China of 15.9% (Ge et al. 1999)
(which was estimated to be age 1-5 incidence of 7.9 per 1000), resulted in rickets in
China being responsible for 2 HeaLYs lost per 1000 people (Table 2.1). China
numbers to compare this against do not exist, but it can be said that at these levels
rickets in China can be estimated to be twice as burdensome as rickets in Pakistan.
Table 2.1. Values for the variables used in calculating HeaLYs due to rickets in
Pakistan and China.
Pakistan (numbers from Hyder etal. (2000) Table 1) China estimates (created by author)
Incidence (per 1000 per year) 0.79 7.9
Case fatality ratio 0 0
Age at disease onset 2 1
Age at fatality 83
Expectation of life at onset 81.84 82
Extent of disability (l=death) 0.25 0.1
Duration of disability (yrs) 5 5
HeaLY 0.92 2
In November 2000 Cornell University and the Islamic Development Rank
hosted an international symposium entitled Improving Health and Economic
Development: Approaches to Preventing Diet-Related Rickets, in part to redress the
problem of global neglect of rickets as a public health problem. The published
proceedings highlight the importance of rickets in countries such as Bangladesh,
Nigeria, Mongolia, and Kuwait. China was mentioned as a country with high rates
of rickets, but no official data on China was reported.
Dietary Deficiencies Among Children in Rural China
Although Chinas economic development in the last 15 years has been
record-setting, this development has not been uniform throughout the country. Rural
areas have lagged behind. The result is that even though the problem of caloric
malnutrition has been largely resolved in rural China, rural children still suffer from
a host of nutritional deficiencies. A 1992 survey in China found that 87% of rural
children in China under age 6 consume less than 50% of the RDA for calcium.
Of continual concern in China is the problem of vitamin D-deficiency.
Vitamin D deficiency in winter is highly prevalent, even among 12-14 year old girls
(Table 2.2), whom one would expect to be outside enough to satisfy their vitamin D
needs through solar means.
Table 2.2. Biochemical indexes of girls aged 12-14 y in Beijing. (Du et al. 2001)
25-OH-D (nmol/L) Winter 12.8 6.7 (64) 13.0+9.6(108)
Summer 23.88.7 30.2 11.9
Calcium (mmol/L) Winter 2.31+0.07(60) 2.26+ 0.12 (96)
Summer 2.3 9 0.08 2.34+ 0.09
Parentheses indicate the number of paired data.
P<0.05 showing a significant main effect of season and area.
Deficiency defined as 25-OH-D <12.5 nmol/L, plasma Ca <2.25 mmol/L
Children with rickets in Shanxi Province have been shown to exhibit
significantly lower 25-OH-D levels than controls (Table 2.3). Serum calcium levels
among these children are normal and not different between rachitic children and
controls, while alkaline phosphatase levels are significantly elevated among rachitic
Table 2.3. Serum levels of normal and rachitic infants in Shanxi Province (age 0-3).
___________________________________(Zhang et al. 1999).________________________
25-OH-D (nmol/L) Ca (mmol/L) ALP (IU/L)
Control (n=30) 49.88+13.55 . 2.280.13 149.67 31.33
Rickets (n=82) 36.23+20.23* 2.21 0.41 214.35 115.92*
Deficiency defined as 25-OH-D <12.5 nmol/L, plasma Ca <2.25 mmol/L, Zn <70 pg/dL Elevated
alkaline phosphatase defined as >115 IU/L.
Studies of illnesses related to rickets in China are scant, but work in
Bangladesh has found a strong relationship between rickets, pneumonia and diarrhea
Table 2.4. Illness histories in households with and without rickets in Bangladesh.
_________Two week recall. Derived from Table 7 (Hassan et al. 2000),__
Disease Rickets households Non-Rickets households
URI 32.3% (n=145) 44.0% (278)
Pneumonia (ARI) 15.6% (70) 2.2% (14)
Diarrhea 12.5% (56) 6.0% (38)
Anemia 6.7% (30) 8.7% (55)
Asthma 5.6% (25) 4.3% (27)
Social and Cultural Factors in Rickets Epidemiology
Diet and nutrition are very much culturally determined. As a foreign
researcher it is important to be aware of cultural influences on diet. Because of the
strong influence Western medicine and Western research has on multilateral
organizations such as UNICEF, it is doubly important to be attentive to cultural
issues influencing diet and nutrition in the developing world, and to allow the social
sciences to inform research. This section will introduce some issues pertinent to this
study as viewed from the social sciences.
From an evolutionary perspective, the mother must balance protein and
energy in her breast milk so that her sacrifice is in ideal balance with the infants
need (Fomon 1993). Consequently, in conditions of poor nutrition, it is possible that
the mother may breast-feed at the expense of her own health, or she curtail breast-
feeding early and deprive the child of critical nutrition1. The women in this study are
not living at such a margin of poverty that nutritional competition between mother
and child should be an issue, but it may have been historically.
Cultural variations in duration and style of breast-feeding are great (Quandt
1986). In some cultures, women can maintain breast-feeding a long time even after
beginning supplemental foods, and even when working (Winikoff et al. 1989).
Duration of breast-feeding and implications for rickets will be examined in this
The Chinese family is patrilocal, where the girl is married off to the mans
family, and her identity as a member of her birth family is lessened, and in some
aspects, such as inheritance, is entirely lost. Patrilocality reinforces traditional
customs/practices and limits maternal autonomy (Wolf 1972; Judd 1989).
Chinas historical perspective on the role of women was based on the Three
Obediences and the Four Virtues (sancong side) of a proper woman. This is the
moral foundation for the patrilocal system in China. The Three Obediences are:
1. Obedience to father before marriage
2. Obedience to husband after marriage
3. Obedience to son after husbands death
The Four Virtues are morality, proper speech, modest manner and diligent work.
A second important cultural standard is that guiding interpersonal
relationships. It is called The Three Cardinal Guides and the Five-Constant
Virtues (sangang wuchang). The Three Cardinal Virtues are:
1 Historically, women in China could earn money nursing babies for women who didnt want the
inconvenience of breast-feeding or had work or other duties that kept them from breast-feeding.
These women were called naima (milk mothers). It is now very rare in China.
1. Ruler guides his subjects
2. Father guides his son
3. Husband guides his wife
The Five Virtues are benevolence, righteousness, propriety, wisdom and
fidelity (ren, yi, li, zhi, xiri).
These traditional values highlight the historical role of women in Chinese
culture. The woman was required to be obedient to the men in her life, transitioning
from her own father, then to her husband, and finally to her son. So she has little
decision-making autonomy in her life.
Communism dealt a significant blow to the public display of traditional
customs such as ancestor veneration and Confucianism, but it did not destroy them.
In this way communism did more to elevate the role of women in society than
democracy and Western influence has done in similar cultural contexts such as
Taiwan and Korea. With the relaxing of political repression, these values are more
public again as people search for a moral or cultural order to guide them. Women in
China today live in relative freedom, but with the strong undercurrent of the
traditional values highlighted above. 71% of rural mothers receive no prenatal care
and 78% of infants are bom in the home (Schwarz et al. 2001). Hospital deliveries
are occuring at higher rates in rural China now, but it is still an issue of importance.
Decision-making and Authority in the Home
Maternal decision-making is not an individualistic choice designed to
maximize self-interest. Rather it is a dynamic activity strongly influenced by power
relationships in the family and society (Morsy 1978). So it is difficult to study the
nutritional habits of children (it is not only the mother making the decision), and
even harder to influence the nutritional behavior of the mother, because she is being
influenced by a number of others in the household. Dietary decisions are intimately
connected to problems that have to do with other larger and more public arenas of
social life (Appadurai 1991). Therefore, individual decision-making cannot be
independently analyzed, and information about proper nutrition alone cannot directly
and independently determine maternal feeding behaviors for her children, because
such decisions are influenced by so many other social factors.
Food allocation practices tend to prioritize adult men over adult women in the
family. Daughters-in-law are often at the bottom of the feeding priority list, and are
often expected to do most of the heavy domestic work (Gittlesohn 1991). This is
true of women in rural China; who also contribute up to 80% of agricultural labor
Relatives, friends and neighbors also have a significant impact on infant
feeding practices, and their advice and encouragement contributes to a successful
lactation experience (Bryant 1982). Unsupportive husbands are especially able to
discourage breast-feeding; however, in China po-po (the paternal mother-in-law)
often acts in the place of her son (the .husband), and she has the most important
influence in the mothers life. The roles of husband and po-po in the Chinese home
are in transition, and how their changing role influences child-rearing practices and
health outcomes needs to be understood.
Chinese society is communal in nature so that individual expression is looked
down on and conformity to the groups expectations is encouraged. There is even
the popular phrase in China, The bird who sticks his head out first is the one to get
it shot off. (Qiang da ehutou niai.) How this social pressure influences child-
rearing will be analyzed in some detail.
Traditional Health Beliefs and Practices
Health concepts that relate to the development of rickets in the child, such
post-partum cloistering, are based in traditional Chinese health beliefs. In China,
Traditional Chinese Medicine is the discipline that comprises the scholarly,
professional basis of these health beliefs (much like Western biomedicine undergirds
lay practices in the U.S.). In order to understand the rationale for beliefs and
practices that contribute to rickets in children, it is necessary to understand TCM.
History of TCM
Traditional Chinese Medicine is as old as Chinese history. Shen Nong
(-1900 BCE) was famous for tasting and trying the effects of hundreds of herbs in
order to safely administer them to patients. He made a record of 365 medicinal
materials of botanical, animal and mineral origin. From 1100-1600 BCE the
writings of Huangdi were compiled to create the famous Huanedi Neiiine (Internal
Medicine of HuangdiV This book described the internal organs and their functions,
and especially their relationship with the environment. Acupuncture was also
described in this book. TCM as a medical profession was formalized during the
Yuan Dynasty (1279-1368 CE), primarily for those who failed to establish a career
in the civil service, the first choice.
Traditional Chinese Medicine (TCM) has profoundly affected the health
behavior and attitudes of Chinese people. Since 1949 the Chinese government has
steadfastly supported and subsidized TCM (Rosenthal 1990). TCM is broadly
practiced, comprising forty percent of the health care delivered in China (Hesketh et
al. 1997). Since the 1970s, TCM has been increasingly welcomed in the West
(Scheid 2002), with therapies such as acupuncture having been validated by many
criteria as an acceptable therapeutic intervention (NIH Consensus Development
Panel on Acupuncture 1998).
Source of Illness in TCM
In TCM concepts the source of illness is not pathogens, but is a weak body
unable to handle environmental threats (qinfan) such as wind, cold, heatity, humid
and dryness. Of all the environmental threats wind is considered the worst This is
one reason cloistering is practiced by new mothers, to avoid contact with the wind.
Illness is seen as a sign of some lack or weakness, not primarily a result of bacterial
or viral attack. This is described by the poem:
If you have good qi in you,
Illness (external forces) cant enter your body.
If bad qi has already entered your body,
. . It means you were already ill.
TCM theory seldom considers an illness manifestation to be the direct result
of a single detectable cause. For example, gastric ulcers to the Western physician
may be diagnosed as an H. pylori infection in the stomach and treated with an
antibiotic. But the Chinese physician would be reluctant to attribute it to one cause.
As Kaptchuk has written, .. .search for the causes may be a hopeless pursuit because
most disease states are the indirect outcome of a constellation of circumstances
rather than the direct result of single determinant factors (Kaptchuk 1983:263).
As part of the Chinese belief in the body as a microcosm of the universe,
illness is also given a providential explanation; namely, that illness is a disruption of
the physical order of the body caused by a disruption of the moral order of the
universe. Therefore a breach in human relationships can manifest itself in physical
illness (Harrison 2004). Family harmony is considered necessary for good health.
Therapy focuses not on treating illnesses, but on strengthening the affected
body system through the use of nutrition and herbal preparations. Treatment does
not focus on using synthesized drugs to destroy or weaken pathogens. TCM theory
holds that supplementing with foods is superior to supplementing with medicines
(yaobu bum shibu).
Caring for the ill is a way in which family roles are acted out through the
expression of affection. Thus people respond to illness in ways that reinforce the
order of the family and household (Harrison 2004). The post-partum mother is
considered to be ill, and the ways family members go about caring for her are an
expression of their shared roles. Complying with her mother-in-laws expectations is
a necessary process of fitting into the family. It would be difficult to oppose the
expectations she has of what is the proper way to recover from delivery.
The Concept of the Human in the Universe
In Chinese philosophy, the human person comes from the universe and is in
unity with it (Wan 1998:34-39). There are many Chinese phrases that illustrate this
belief, such as Confucius Tian ren heyi (Heaven and the person are one). The
human person is a part of the greater cosmos like a given organ in the body is a part
of the entire body. As a person moves from one environment to another he or she
needs time to adjust to it. If a persons body is strong, they can adapt quickly and
easily, if it is weak, for example .they have just delivered a baby, they are vulnerable
to attach from environmental forces such as wind and cold. This is also why taking
the child outdoors to sun bathe is considered by Chinese parents to be risky.
Rickets and Traditional Health Beliefs
Ancient History of Rickets in China
Rickets has been a part of Chinese life from antiquity (Meng 2002). It was
said of Confucius that as a child he had a large and a depressed fontanelle (soft spot
on his head). As an adult he had a hunched back and a blocky-shaped head. Famous
historian Sima Qian (writing in 105 BCE) wrote of Emperor Qin Shihuang (214 BCE)
that he had a pronounced chicken breast [pectus carinatum).
The word goulou (rickets) can be found in Menezi Dashens which comes
from the Spring-Autumn period (770-476 BCE). Goulou referred to the stooped
shoulders and bent back of the aged, as well as bowed legs and chicken breast. From
history Chinese associated osteoporosis with rickets, and even considered rickets to
be a form of premature aging. In Confucius The Book of Poems goulou is
associated with widened wrists that look like the claws of the chixiao bird. At that
time if a young man had this condition, it was a blessing because he was exempted
from conscription to war. Rickets has been a problem in China for a long time. This
is different from the history of rickets reported in England, where it was the advent
of tall buildings blocking the sun and hazy, polluted air blocking out the UVB rays,
that lead to an increase in rickets. In China, rickets is a function of widespread
nutritional insufficiency and habits that keep mothers and infants out of the sun.
The book Childrens Pharmacology Secrets (JGaoer yaozheng zhijue) by
Song Dynasty doctor Qian Yi describes rickets as five delays: delayed walking,
feet wobbly when walking, delayed eruption of teeth, teeth not strong, hair growing
slowly and hair not dark black. Ming Dynasty doctor Lu Baisi in his book Hundred
Questions for Children talks about the five weaknesses of rickets: soft head, soft
neck, weak hands, weak feet, and weak muscles. From then on rickets was called the
five delays and the five weaknesses.
The current signs identified with rickets have been observed in China for
thousands of years. Very early, the symptoms of chicken breast, hunched back,
deformed joints, bow legs, X-shaped legs and widened wrists were widely
recognized as abnormal, and sent Chinese medicine doctors to search for clues.
Analysis of Rickets in Traditional Chinese Medicine
TCM teaches that rickets results from four problems with the child, his
internal organs are fragile, his basic health inherited from his parents is poor, his
early nutrition is insufficient, and his spleen and kidneys are lacking (Yu 2002).
TCM teaches that the kidneys are connected to and sustain the bones (shen zhu gu).
Kidneys store what is called jing the energy or essence. If a child is slow in
physical development, has late or incomplete fontanelle closure, or has poor general
skeletal development, it is attributed to kidney jing being inadequate (shenjing buzu)
(Kaptchuk 1983:65). The prescribed treatment to enrich the kidneys is the herbal
rehmannia 6. Longmu zhuangu is another medicine used to treat rickets. TCM
i . . ' . .
practitioners contend that these medicines can cure rickets in the absence of vitamin
Several concepts pertinent to rickets in China have been introduced in this
section. People are vulnerable to environmental forces, like wind, which can
contribute to illness. When a person is weak, such as a new mother post-partum, she
is especially susceptible to illness, so she should stay indoors. Family roles are
played out in the process of recovery, so the mother-in-laws role in caring for the
new mother is an expression of her authority and status as family matriarch.
Illnesses, like rickets, result from a cluster of causes and require a similarly broad
The biology of rickets is well understood, forming a sound foundation on
which to conduct this research. Two areas of particular interest currently are the
relative role of calcium insufficiency as a contributor to rickets, and the broader
implications of prevalent vitamin D deficiency. Research in tropical regions where
sun exposure is plentiful, such as Bangladesh and Nigeria, has shown that calcium
insufficiency alone can cause rickets. This study will explain the relative roles of
calcium and vitamin D deficiency to rickets. A second important contribution of this
study is the consequences for these infants and children of spending many
consecutive months in the context of severe vitamin D-deficiency and then the
results of adequate summer sun exposure.
There has been a lack of consistency globally in how rickets is diagnosed.
Consequently, it has been difficult to compare rates of rickets from one country to
the next and over time in one country. This study will describe the basic
epidemiology of rickets, comparing the results of different methods of diagnosis,
calculating a true prevalence for active rickets, and determining the public health
consequences of rickets.
Nutrition and nutrition-related illnesses are bound up with cultural values and
customs. In order to reduce the rates of rickets in China it is necessary to understand
the correlates contributing to or found in association with rickets. Some of these
correlates are culturally-bound behaviors that are not easily understood by the
outside researcher. This study will combine survey methodology with qualitative
research methods as a way of both determining and explaining behaviors that are
contributing to rickets, and to extend the research by particular focus on Chinas
unique customs such as zuoyuezi and shanghuo. Dietary behaviors contribute to
rickets as well and will be duly assessed.
MATERIALS AND METHODS
Characteristics of Subjects Involved
Research Location: Yuci. Shanxi Province. China
Jinzhong Prefecture, Yuci City (popn 500,000) lies in the eastern part of
Shanxi Province, China (see Figure 3.1 and 3.2 for maps of the research area). Yuci
City was chosen as the study site because it is rural, but has transportation access to
make research possible. Furthermore, Yuci has not been previously exposed to
foreign health workers.
Yuci, Shanxi Province is a 1500 year old city. Historically it was the center
of commerce and banking for north China, but in the past 100 years it lost ground in
this regard. Heavy industry and agriculture are the pillars of the Yuci economy
The villages in the Yuci area have many paved roads, although they are in
great disrepair. Compared to many villages in north China, transportation in Yuci is
considered good, and the farmers are more successful. Details on the population will
be presented in the results section of this dissertation.
The Yuci Maternal and Child Health station has been recognized by Shanxi
Province health officials as being a model. In our work we have found this to be true.
In 1982 the Yuci Maternal and Child Health station did a large rickets survey in Yuci
discovering a rickets prevalence of around 70%. These results contributed to the
national rickets study in 1983 that revealed rates of rickets for all of China of 40.7%
(Ma et al. 2002). They have been working to combat rickets since then. (For more
information on the structure of Chinas rural health care system refer to Appendix B.)
Figure 3.1. Map of China.
Subjects were mothers and their infants between 12-24 months of age, known
to be the age of highest risk for rickets. Power analysis determined that a sample
size of 300 was sufficient to yield statistically significant results. A multi-stage
cluster sampling (non-probability) method was used. Based on availability of
electricity and a space large enough in which to conduct research, we selected three
townships, one in east Yuci (Changning Township), one in west Yuci (Zhangqing
Township) and one in south Yuci (Chenkan Township) (see Figure 3.2). These three
townships also represented a range from mountainous and remote to more local and
wealthy. Villages (the cluster) were then selected randomly and all qualifying
women and their infants within each village were invited to participate.
In total, 14 villages were sampled. Excluded from the study were mothers
lacking the cognitive ability to participate (1), immigrants to Yuci (0), infants bom
before 37 weeks gestation (2), infants away from the village for the majority of their
life (0), and infants who were significantly disabled, mentally or physically, in
fashions unrelated to rickets (1). Because village doctors were needed to coordinate
the study, villages without a designated village doctor were excluded from
Transportation in this region of China is very challenging. It was difficult to
pack up, transport and set up the X-ray machine and blood collection equipment (the
Township Health Center Clinics did not have lab or X-ray equipment available).
These issues forced us to exclude some remote villages which had few eligible
children and lacked space or electricity to do our work.
The original study plan was to enroll 300 subjects. Because of the epidemic
emergence of sudden acute respiratory syndrome, or SARS, in this region, we were
forced to stop the research at 250 subjects (see Appendix C for a report on doing this
research in the midst of the SARS epidemic). This compromised our statistical
power. However, I felt it was necessary to sympathize with the fears of my local
colleagues, and not force them to continue collecting data, even though no health
officials had as of yet demanded that we stop (but they would have in a couple days).
A smaller number of subjects may have compromised our ability to detect non-
random variability in some of our variables of interest, for example, the small
number of cases of pneumonia.
In order to understand the cultural context in which these mothers live, and
their health perceptions, six focus group discussions were convened. Study subjects
were mothers of infants age 12-24 months (3 focus groups), grandmothers of young
children (age range 45-60) (1 focus group), Traditional Chinese Medicine (TCM)
doctors (1 group), and MCH workers in Yuci (1 group). With the assistance of the
Yuci City Maternal and Child Health Hospital, villages from different regions in the
Yuci area were selected for the research. The villages selected had not participated
in the spring round of quantitative data collection. There were no major known
differences between these villages and the villages who participated in our
quantitative survey. They were located distantly from one another, and represented
both mountainous and plains regions. The selection of distant villages was also to
minimize the potential for contamination among participants between each focus
group, and to minimize bias that might come from an unexpected visit to one of
these areas by an MCH official.
The village doctor from each village chose 5-6 eligible mothers or
grandmothers from the village and invited them to participate in a focus group. At
least one mother of an infant diagnosed with rickets was included in each focus
The Yuci Public Health Bureau representative identified TCM doctors, and
Yuci City Maternal and Child Health Hospital workers identified MCH workers
whom they considered to be respected and skillful in their areas, and invited them to
participate in two additional focus groups.
Mothers who participated in the spring round of quantitative data collection
were given a small toy truck or car for their child in the spring, and a small English-
Chinese picture book in the fall session. Focus group participants were given a gift
of 15 yuan value ($2) for their participation.
Process of Collaboration
On March 25,2003 the workers from Shanxi Medical University, Shanxi
Women and Childrens Hospital, Yuci Women and Childrens Hospital and
Evergreen began meeting to finalize the experimental design and methodology. This
involved several sessions and discussions. The research team was made up of the
following institutions and individuals:
Shanxi University: Li Sihan, Ni Shuhua, Gao Zhuqi
Shanxi Women and Childrens Hospital: Zhang Peiying, Xi Weiping, Yang Jianping,
Yuci Women and Childrens Hospital: Liu Qiaoe, Jin Meimei, Wei Yuqin, Zhang
Haizhen, Li Caihong, Zhao Jinping, Jia Yihong, Wang Aiai
Chenkan Township Health Clinic: Zhao Runshou
Evergreen: Judy Perry, Guo Jinzhi, Guo Jinbiao, Mark Strand, and drivers Zhou
Bing and Sun Zhizhong
One of the first concerns in implementing the study design was that Chinese
mothers would not be interested in allowing their children to have blood drawn, and
would thus decline participation in the study. This concern also arose during the
process of gaining IRB approval from the Shanxi Province Public Health Bureau
Department of Science and Research. Here the concern was with proper handling of
the blood by the investigators. It was necessary to give strong proof that all blood
would be handled by Chinese workers and analyzed in a legitimate Chinese
laboratory, with no risk of foreigners extracting DNA or exporting blood products
out of country. The second major concern of the IRB was that if too large a
compensation was given to mothers for their childrens participation it would be
coercive. These issues were worked out to everyones satisfaction (consent forms in
Chinese and English are provided in Appendix D). To everyones surprise, the
mothers were so concerned for their childrens health that not only was the
participation rate high, there were even mothers who did not qualify for the study
who still tried to participate.
All workers were given a set of supplies including a clipboard, notebook, pen,
drinking cup and snacks as a way to show appreciation for their involvement. Yuci
MCH Center medical workers were paid the locally agreed upon fee of 40 yuan per
day ($5 USD), and the provincial level workers were paid a per diem of 50 yuan a
day ($6.10 USD) for their involvement in the study. This money was divided so that
half was paid to the worker her or himself, and half was paid to the department
which they vacated while participating in the research.
The survey had been translated in advance, and then it was pilot-tested with
several mothers in the Yuci MCH Hospital to make sure that the questions were clear
and understandable locally. See Appendices E-H for English and Chinese copies of
the survey instruments used.
Permission to survey a given village was secured from the director of the
Township Health Center which provided services to the villages of interest (a
township is the administrative unit just above that of the village). We gave him a
copy of the Research Protocol and the IRB permission form. His permission allowed
us to then work directly with the Township Health Center Maternal and Child Health
worker to contact the women.
A week in advance of surveying the mothers and examining their children,
the Township Health Center Maternal and Child Health worker delivered to the
eligible women in selected villages a brochure entitled For the Health of your
Child, which contained a description of the study and an invitation to participate.
The township MCH worker was also given a copy of the informed consent form,
which she was instructed to orally introduce to eligible mothers, as well as informing
them when we would be in their community.
The research team included 2 pediatricians who were available to provide
consultation for children presenting with health problems. Fortunately, we never had
any complicated cases, and we were aware of no negative results from the blood
drawing, X-rays or physical examinations. Anxious mothers were mostly concerned
about problems such as spots on their childs finger nails, skin rashes (often from not
bathing), light colored hair and such.
As noted above, this research project involved the collaboration of
researchers from the University of Colorado at Denver, Shanxi Medical University
(SMU), Shanxi Province Women and Childrens Hospital, Yuci City Women and
Childrens Hospital and the American NGO Shanxi Evergreen Service, with
participation of researchers from the Shanxi Province Public Health Bureau (PHB).
Professor Li Sihan was responsible for blood work and laboratory analyses at the
Shanxi Medical University. He also has collaborators in medical laboratories in
Beijing who assisted in setting up the procedures used and provided test standards.
Dr. Zhang Peiying arranged technicians and pediatricians from Shanxi Childrens
Hospital to perform physical examination on infants. Judy Perry, RN, MPH, of
Evergreen was responsible for quality control in the physical exam team. Director
Tian Xiangyi of the Shanxi Province Public Health Bureau managed all political and
legal matters. Dr. Jin Meimei, pediatrician and vice-director of the Yuci Women and
Childrens Hospital escorted researchers to village clinics and arranged liaison work
with the village doctors and the families involved. Guan Xiaoli, department head of
radiology at Shanxi Childrens Hospital, was responsible for radiography, which was
actually done by Dr. Zhao Runshou of Chenkan Township Health Center.
Responsibilities of Scientific Personnel Involved in the Research:
Mark Strand Co-Primary investigator
Train in survey method and focus group method
Li Sihan Co-Primary investigator
Liaison with government and health officials
Oversee blood drawing and blood analysis
Final authority on unexpected research roadblocks
Zhang Peiying Liaison with Maternal and Child Health system
Train staff in rickets and other clinical diagnoses
Jin Meimei Arrange village visits, site preparation
Coordinate the role of village doctors
Experimental Design and Methodology
The research design involved prospective identification of infants with and
without rickets, and the retrospective evaluation of caretaker knowledge and
behaviors that relate to the childs illness or health. It is a cross-sectional design
with a longitudinal component. Although rickets can result from vitamin D-receptor
deficiency and hypophosphatemia, diagnoses of such rare conditions was beyond
the scope of this study. In this study, all cases of rickets were considered to be
nutritional rickets (vitamin D- and/or calcium-deficiency).
While the results of population-based studies may be less precise than case-
control studies, they are less affected by selection bias. Consequently, it is assumed
that the results of this study can be safely generalized to other rural areas in north
Data analyses were carried out using SPSS (version 11.0). Outcomes and
risk factors were compared using a variety of statistical methods, such as t-tests for
continuous variable, paired t-test for the changes of measures from spring to fall, chi-
square tests for comparing proportions. Mantel-Haenzel test was used to examine
the change of rickets rates from spring to fall. Univariate logistic regressions were
fitted to examine the association between rickets and risk factors. For the same
purpose and using both spring and fall data, a nonlinear mixed model (Generalized
Estimating Equations) was fitted using the status of rickets as dependent variable.
Results were presented as odds ratios with their p-values. Several correlation
coefficients (Phi, Pearson and rho) were used and p-values presented.
Power analysis done prior to data collection indicated 300 subjects would be
necessary to detect statistical significance. Due to the SARS epidemic, only 250
subjects were included in the study, compromising the power of the final data.
Additionally, because this is an observational study, the priority is on determining
risk factors potentially related to the dependent variable. While all statistical tests
were performed using an alpha of 0.05, in most cases p<0.10 is used as the criterion
of statistical significance.
Food Intake Score. In order to calculate relative amounts of calcium, vitamin
D, zinc, iron and protein that the children were consuming, a food frequency chart
was used in the survey administered to the mothers. The chart elicited consumption
of foods known to be the most common foods of children locally. Mothers were
asked to indicate the frequency with which her child ate a particular food item (not at
all, occasionally or often).
The intake scores for each of calcium, vitamin D, zinc, iron and protein were
calculated in the following ways. Consulting the following food chart (Table 3.1),
each food item was given a score for the relative amount of calcium, vitamin D, zinc,
iron and protein it contained. This number was then multiplied by the frequency
with which that food was eaten (Not at all=l, Occasionally=2, Often=3).
Table 3.1. Relative content of foods eaten by Chinese children.
Relative content of the following
Food item Calcium (mg) Vitamin D (IU/1) Zinc (mg) Iron (mg) Protein (gm) Caloric value (cal)
Milk ('A pint) 288 -100 (supplemen ted),-6 (unsupp.) 4 0.2 8.5 166
Soy products ('A cup beans) Made with CaS04? 88 (4oz dofu=12 0-392 mg) 1 4 (dofu?) 10 94
Green leafy vegetables (lettuce) Low -2 High High
Cucumbers (Huang gua) A medium 5 1.4 0.2 0.4 6
Green beans ('A cup) 65 2.4 1.1 2.4 35
Fruits (in any form) ('A cup applesauce) 5 0.2 0.5 0.3 50
Potatoes, sweet potatoes (1 med) 11 2.3 0.7 2 83
Cereals (cereal formula) (% cup) 10 4 1.1 (must be fortified) 3.5 102
Com products (2 in square of wo- wo tout) 29 6 0.7 3.2 93
Millet soup (xi fan) (A cup millet) 1.9 10.9% of grain is protein
Table 3.1 (Cont.). Relative content of foods eaten by Chinese children.
Relative content of the following
Food item Calcium (mg) Vitamin D (IU/1) Zinc (mg) Iron (mg) Protein (gm) Caloric value (cal)
Rice porridge (Mi tang) (% cup rice) 7 4.3 02 2.1 103
Mantou (1 steamed bun) 35 0.3 4.7 150
Crackers (1 saltine) 1 6.6 0.1 0.3 14
Noodles (dou mian, long xu mian) (% cup) 10? 6 1.2? 5.5 162
Eggs (yolk and white) (1 boiled) 26 24.6 (0.6 Mg) 14 1.3 6.1 77
Meat products (ground up) (1 med pork chop) 8 23 2.1 16.1 233
Liver (beef) (100 gm) 4 8 48 2.9 8.8 86
Vitamin D supplements (vitamin pills containing vitamin D, fish liver oil, etc.)
Brown sugar water (hong tang shui) 11 0.4 0 52
Sun Score. The Sun Score is made up of two variables, month of birth
(reflects amount of UVB available to the child three months after birth) and the
number of days after birth the child was taken outside. These two sub-scores added
together create the childs total Sun score, ranging from 0.5 to 7 (Tables 3.2 and
The score given for month of birth was created by consulting the information
published by Webb et al. (1989) on UVB available in northern climates for each
month of the year.
Table 3.2. Score given for month of birth.
Month of birth Month that child was 3 months old (likely taken outdoors) Month of Birth Score
1 January 4 2
2 February 5 3
3 March 6 4
4April 7 4
5-May 8 4
6-June 9 3
7-July 10 2
8 August 11 1
9 September 12 0
10-October 1 0
11 November 2 1
12 December 3 2
Table 3.3. Score given for time taken outdoors.
Number of days after birth taken outside Time outdoors Seme
All forms were written in English and Chinese. The Chinese version was
pre-tested through three modified versions. Three female workers from the Yuci
Maternal and Child Health Hospital were trained to conduct interviews. They
conducted most of the interviews with subjects in the local language, Mandarin
Chinese. Mark Strand conducted 27 interviews (10.8% of total), also in Mandarin.
The interviewer was blinded to the condition of the subjects child, however, the
interviewer could often see the child being held by grandmother or sometimes by the
mother herself. Interviews were conducted in the local clinic in a place separate
from the blood drawing and physical exam. The interview was the first event in the
process, so mother and child were full of anticipation and had not yet been subjected
to physical exam or blood drawing. Dietary information was also obtained at this
Physical Examination and Biochemical Measures
The second stage in the process was physical examination in the Township
Health Center. Trained local clinicians carried out physical examination and
determined the severity of rickets based on a list of clinical signs and symptoms.
Three examiners worked together on any given day. It was not necessary to test
inter-rater reliability on these examiners as diagnostic decisions were rendered by
The scale for diagnosing rickets was provided by Dr. Xi Weiping of the
Shanxi Province Maternal and Childrens Hospital, which came from a China
Ministry of Public Health document on rickets prevention (China Ministry of Health
A radiology team took x-rays of each childs wrists.
Infant illness history was determined by maternal interview. There were
three questions used to assess the childs illness history.
1. Has your infant experienced any illnesses in the last week? What was it?
2. Has your child experienced any illnesses in the last month? What was it?
3. In the recent six months, has your child experienced any of the following
illnesses? Listed were diarrhea, recurring respiratory infection, dysentery,
pneumonia, upper respiratory tract infection and anemia.
The first two questions were open-ended, allowing the mother to report what
she remembered. The third question was meant as a prompt to the mother to help her
consider in a more focused way whether her child had experienced any of the
illnesses listed. Definitions used for illnesses followed the WHO IMCI (Integrated
Management of Childhood Illness) definitions (World Health Organization 2001) as
Diarrhea = Diarrhea of more than 3 loose watery stools a day lasting for a
few days or more (not routine soft stools, but out of the ordinary diarrhea).
Recurring respiratory infection = More than three times in recent six months
Dysentery = Diarrhea by above criteria and containing blood
Pneumonia = Diagnosed by a doctor. (Defined as Respiratory infection with
1. Fast breathing, 2. Lower chest wall indrawing, or 3. Stridor.)
Upper respiratory tract infection = Respiratory infection with fast breating
Anemia = Has your child been tested for anemia? Were you told by a doctor
that he/she had anemia? Deficiency defined as blood Hemoglobin <11.5 g/dl.
Stunting = Height for age z-score below -2SD
Underweight = Weight for age z-score below -2SD
Wasting = Weight for height z-score below -2SD
To avoid misclassification bias, it was crucial to accurately diagnose rickets,
the main dependent variable in this study. Because there is disagreement in what to
use as a gold standard for measuring rickets in China and the U.S. this proved more
difficult than expected. In the U.S. positive signs of rickets on X-ray is considered
the gold standard. In China, x-rays and alkaline phosphatase testing are used to
diagnose rickets, but economic considerations as well as habit have preserved the use
of physical signs and symptoms as the most widely embraced standard. Give these
differences, the following diagnostic procedures were adopted.
1. All infants were assessed for rickets by clinical exam according to the criteria
shown in Appendix F. Results were dichotomous, positive or negative
determination of rickets. A positive clinical diagnosis of rickets required the
presence of 3 or more primary clinical signs, or 2 or more primary clinical
signs plus 2 or more secondary clinical signs. Clinical symptoms were not
included in the clinical diagnosis of rickets because they proved to be non-
specific for rickets. This diagnostic procedure followed the time-tested
Chinese standards, but was made stricter because of such a high prevalence
of positive cases.
2. Radiography of both wrists was taken for each child. Dr. Zhao Runshou
collected and processed all x-rays. Each child was assessed for radiographic
rickets using the scale provided in Thacher et al. (2000). This resulted in a
score from 0-8 (4 points possible for each wrist). A score of one or more was
considered a positive indication of rickets. These x-rays were read Dr. Guan
Xiaoli, chief of radiology at Shanxi Childrens Hospital.
3. Serum 25-OH-D levels were assayed on 180 children (there was inadequate
funding to test all children). Using 25-OH-D <12 ng/ml as the cut-off for
deficiency, children were categorized as either deficient or normal on this
4. Blood alkaline phosphatase levels were assayed for each child. Children
with alkaline phosphatase level >250IU/L were considered to have a
positive diagnosis of rickets by that criteria.
A final diagnosis of rickets was determined by making a composite of the
results of each of the above diagnostic measures, depending on which method of
diagnosis was used. This will be introduced in the results section of this dissertation
when the results for each diagnostic method are presented.
Body weight was measured using a beam-balance platform scale with a
multipurpose bed-chair mounted on the platform. The scale was manufactured in
Wuxi, China at the Wuxi Scale Factory (Ertong Chen (Child scale), model number
RGT-50, maximum weight 50 kg). Chinese parents are reluctant to allow their
infants to be fully unclothed, so infants wore one layer of standard clothing. The
weight of a representative set of clothes was subtracted from the measured weight.
Recumbent length was measured using an aluminum measuring board
manufactured in China by Nantong Instruments Factory, Nantong, China (Yingyouer
shengao zuogao ji (Child and infant length measuring board), Sangong pai brand,
model YSL-1, serial number 3037). The measuring range was 50-1050 mm, with a
variability of 1 mm.
Weight-for-age, and height-for-age z-scores were calculated using ANTIIRO
(version 1.02,29 June 1999, CDC/WHO), which is based on United States NCHS
standards. It was concluded that the broad use of these standards in anthropometries
made comparison with other published reports more feasible.
Collection of serum. Two ml of blood (non-fasting) was collected by
venipuncture on the inside of the elbow by manual extraction using a 5 mL plastic
disposable injection syringe. Before collection 1% nitric acid, ionized water, iodine
was used to disinfect the area from which the blood was to be drawn. The blood
sample was then manually extracted into glass test tubes or Eppendorf tubes for
analyses. All plastics, glassware and other material coming into contact with the
specimen were free of any contamination. 10pL of blood was used for the
measurement of hemoglobin, which was added in a colorimetric tube with 2.5ml
hemoglobin assaying test solution, well shaken up for measuring. Measuring
alkaline phosphatase required 30pL of whole blood. It was placed in the sampling
well of the alkaline phosphatase reactive strip for measurement, after 2 drops of
dilution was added first. Assaying blood lead required 20pL whole blood, which
was added in 20pL nitric acid solution, well shaken up for measurement.
Serum isolation: The whole blood was centrifuged 15 minutes at a speed of
3000r/min to separate out the serum. The serum was stored in a -20C refrigerator.
All equipment used in collection such as Eppendorf tubes and micropippette
were soaked in 10% nitric acid for 24 hours, then washed clean by distilled water,
ionized water, and dried to use. In pipetting, a new tip was used each time.
All blood work was done in the laboratory of the Department of Nutrition
Sciences at the Shanxi University Medical University.
Standards of abnormal for these tests were defined as:
. 25(OH)D <12 ng/ml
Calcium <90 pg/ml (2.2nmol/L)
Zinc <0.60 pg/mL
Alkaline phosphatase >250IU/L
Phosphorous <38 mg/L
Lead >10 pg/dl
Serum 25-OH-D Assay. The 25-Hydroxyvitamin D I Radioimmunoassay
Kit is manufactured by DiaSorinlnc., 1951 Northwestern Ave, Stillwater, MN
The DiaSorin 25-OH-D assay consists of a two-step procedure. The first
procedure involves a rapid extraction of 25-OH-D and other hydroxylated
metabolites from serum or plasma with acetonitrile. Following extraction, the treated
sample is then assayed using an equilibrium RIA procedure. The RIA method is
based on an antibody with specificity to 25-OH-D. The sample, antibody and tracer
are incubated for 90 minutes at 20-25C. Phase separation is accomplished after a 20
minute incubation at 20-25C with a second antibody precipitating complex. A
NSB/Addition buffer is added after this incubation prior to centrifugation to aid in
reducing non-specific binding.
Collected blood samples (see above) were allowed to clot at room
temperature (15-25 C). Samples were centrifuged for 15 minutes using
approximately 760 X g* to obtain hemolysis free sera. To maintain integrity of the
sample, no additives or preservatives were added.
Serum samples were stored at -20 C in glass vials well sealed to prevent
desiccation of the sample. Spring blood samples were stored for three weeks before
analysis. Fall blood samples were stored for two weeks before analysis. DiaSorin
claims an internal study has shown samples to be stable for up to 9 weeks when
stored at-20 C.
Extraction and Assay:
1. Set up labeled 12 X 75mm disposable glass tubes for each standard,
control and patient sample.
2. Add 500pL of acetonitrile to each tube.
3. Place pipette tip containing 50uL of standard, control or patient sample
below the surface of the acetonitrile and slowly add into the acetonitrile.
4. Vortex for 10 seconds.
5. Centrifuge using 1200 X g* for 10 minutes at 20-25 C.
6. Pipette duplicate 25 pL aliquots from the supernatant onto separate
appropriately labeled 12 X 75mm tubes taking care not to disturb the pellet.
7. Assay supernatants according to the assay procedure.
1. Allow all reagents and samples to equilibrate to room temperature. Do
not allow reagents to reach temperatures above 25 C.
2. Set up labeled 12 X 75mm disposable glass tubes in duplicate.
3. Add reagents as follow:
a. . Total count tubes
50|xL of 125I 25-OH-D
1.0ml of NSB/Addition buffer
b. Nonspecific binding tubes (NSB)
25 pL of 0 standard (extracted)
50pL of 125125-OH-D
1 .OmL of NSB/Addition buffer
c. Standards, controls, and unknown samples
25pL of standard, control, or unknown sample (extracted)
50pL of 125125-OH-D
l.OmL of 25-OH-D antiserum
4. Vortex gently without foaming and incubate for 90 (+/- 10) minutes at
5. Add 500pL of DAG precipitating complex (DAG precipitating complex
should be mixed thoroughly before and during use) to all tubes except the
total count tubes.
6. Mix tubes well and incubate for 20-25 minutes at 20-25 C.
7. Add 500 uL of NSB/Addition buffer to all tubes except the total count
tubes. Vortex gently to mix tubes well. Use caution when performing this
step to avoid splashing due to high liquid volume in tube.
8. Centrifuge all tubes for 20 minutes at 20-25 C 1800Xg*, except the
9. Decant the supernatants, except the total count tubes, using a foam rack
tube holder or equivalent by inverting the rack into an appropriate waste
container. Place the inverted rack into absorbent paper for 2-3 minutes. Blot
the tubes gently to ensure all liquid is removed.
10. In a gamma scintillation counter, count each tube for a minimum of 1
minute. Each tube should was counted for a sufficient time to achieve
Hemoglobin Assay. Add 20pL whole blood to a colorimetric tube pre-
loaded with 5ml hemoglobin test solution. Shake it up. Use 721 spectrophotometer
to assay the color (at less than 540 wavelength).
Measurement of Hemoglobin: Drab kins method. The test solution and the
standard are both in compliance with the Hematology Institute of the Chinese
Academy of Medical Science.
Serum Calcium and Zinc Assay. Dilute 100 pL serum 15 times with 1%
nitric acid. Use atomic absorption spectrophotometer (Atomic Absorption
Spectrophotometer, Japan DaoJin AA670) to measure the concentration of calcium
and lead by flame width and turbidity.
The wavelength for calcium is 422.69 ran, and for zinc 213.9nm. Flame is
air -acetylene instruments slot. Volume of flow of air and acetylene and light
electric current is adjusted satisfactorily according to the instruction book of the
To make the zinc standard curve, separately take 0.1, 0.2, 0.4, 0.8,1.6 ml of
1 pg/ml standard application solution. 1% nitric acid solution is added to each to
make 10ml resulting in 0.1, 0.2, 0.4, 0.8,1.6 pg/ml.
To make the calcium standard curve, separately take 0.1, 0.2,0.4,0.8 ml of
100 pg/ml standard application solution. 1% nitric acid solution is added to each to
make 10ml resulting in 0.1, 0.2, 0.4,0.8 pg/ml.
The calcium concentration of the quality control blood serumIs 78110
pg/ml, the actual measure is 102.25 pg/ml.
Serum Zinc. Use the flame atomic absorption spectrophotometric method.
Blood serum controls are provided by the China Academy Science of Medicine
Institute of Environment and Health. The zinc concentration of the quality control
blood serum is 0.61-0.8 lpg/ml, the actual measure is 0.75pg/ml. In statistical
analyses, serum zinc values greater than 1.1 pgm/ml were considered to result from
contamination and were discarded.
Lead Assay. To assay blood lead, use the Graphite Oven Atomic Absorption
Spectrometry. Lead blood controls are supplied by the China Academy Science of
Medicine Institute of Environment and Health
Add 20pL whole blood to a corked tube pre-loaded with 160pL ionic water
and mix. Then add in 20pLl% nitric acid solution to dissolve. Use atomic
absorption spectrophotometer, in the method of graphite oven, to measure the
concentration of lead. At the same time do a standard series and quality control
Test condition: wavelength 217nm.
To make the standard lead curve separately take 10,20,40, 60, 80pL of the
0.2pg/ml standard application solution. Into each of these add 1% nitric acid
solution up to 200jjL, resulting in 10,20,40, 60, 80pg/ml.
Bone Alkaline Phosphatase. The Bone Alkaline Phosphatase Test Kit (IZS-
ISO AP TEST PROCEDURE) is manufactured by the Zhongsheng Beikong Bio-
technology and Science Company, Inc., No 6, Zhongguancun Road, Beijing 100083,
The Bone Alkaline Phosphatase test kit used is a partial quantitative assay in
that it uses a colorimetric process requiring subjective determination by the naked
1. Add 2 drops of washing solution into the well of the plasma separator,
wait for 3 min to wet plasma separator membrane.
2. Collect 30pL of peripheral blood using a sampling capillary (up to the
blue line on capillary) then add it into the well of the plasma separator.
3. Add 2 drops of washing solution to flush plasma down to the reactive
4. Tale away the plasma separator from the test cassette and a reactive
membrane is exposed on the reactive strip.
5. Add 2 drops of washing solution to the reactive membrane
6. Move forward the reactive strip and let the reactive membrane locate at
the right center of the test window.
7. Add a drop of color developing agent on the reactive membrane and keep
for 9 min in a 37 C incubater.
8. Move back the reactive strip to original position and add a drop of stop
solution on the reactive membrane.
9. Compare the color of the reactive dot with the color of the standard
palette within 20 min.
Qualitative Data Collection
Conditions perceived as pathological in the West, but very prevalent in
developing situations, are often seen as part of the normal trajectory of human
development. This may be the case with rickets in north China. Therefore, a focus
groups methodology was used to determine whether rickets is perceived locally as a
health problem and to help describe the context in which infants are being raised.
Focus group facilitators who speak the local dialect were hired and trained to
moderate these focus groups, and to record data from the sessions (Basch 1987).
The facilitators were Mark Strand, Guo Jinzhi, Jia Yihong and Zhao Jinping.
Prior to conducting the focus groups, I wrote my anticipated answers to each
of the focus group questions as a way of bracketing my opinions. This was to allow
for a comparison of my personal opinions and the actual opinions of those who
participated in the focus groups. These bracketed opinions were also used in data
Each focus group session involved a group of 5-6 selected individuals being
guided in a discussion by trained facilitators. Four 90-minute focus groups were
held among mothers (three groups of young current mothers and one group of older
mothers who are now grandmothers), and two 90-minute focus groups were held
among health workers (one group of MCH workers and one group of traditional
Chinese medicine doctors). Open-ended questions were used to guide the focus
group discussion (questions used are in the Appendix). These focus group sessions
were audio taped and later transcribed verbatim in Mandarin Chinese. A notebook
was used to record decisions, modifications and other miscellaneous observations to
provide a clear accounting of the methodology used.
Focus Group Focus Group Location Focus Group Participants
Young mothers 1 Dongzhao Township, Xiyao Village. Mothers home. 6 women
Young mothers 2 Zhuangzi Township, Zhuangzi Village 6 women
Young mothers 3 Huangcai Township, Huangcai Village. Held in township clinic. 6 women
Grandmothers Dongyang Township, Dongyang Village 5 women between ages 48-55
Township MCH Workers Evergreen Office in Yuci 5 women
Yuci TCM Doctors Evergreen Office in Yuci 6 men and 1 woman (TCM experience 10, 30, 20,12, 34, 18 years)
The focus group transcripts were analyzed independently by four members of
the research team (one foreigner and three Chinese researchers). The objective of
this analysis was to determine common themes (called codes) that tend to be shared
by the focus group participants. All four researchers then met for a day of discussing
the results of their independent analysis. Based on this analysis, a unified set of
codes was created along with important quotes from the transcripts. The results of
the initial coding, choices of codes and emerging themes was checked by mothers
and health workers to confirm that what was observed in the transcripts reflects the
opinions of the focus group participants. Microsoft Word was used to code this data.
This research was done under the authority of the Shanxi Province Public
Health Bureau as a collaboration between Shanxi Medical University, Shanxi
Province Women and Childrens Hospital, the University of Colorado at Denver, the
Yuci Women and Childrens Hospital and Shanxi Evergreen Service. It was carried
out through the Yuci Women and Childrens Hospital and the Yuci City Public
Health Bureau. We relied on village doctors to contact the subjects and secure then-
consent to participate in the study.
All subjects were given a full oral explanation of the study. Verbal
informed consent was obtained from all subjects according to procedures approved
by the Institutional Review Boards of the University of Colorado-Denver (Human
Subjects Protocol number 875) and Shanxi Province Public Health Bureau ((Shanxi
Province Public Health Bureau Office Document No. 22,2002). Consent forms are
included in Appendix D. Children are a vulnerable population and special care was
taken to minimize risks to them. Interviews took place in the township clinic, and
included the principal or a co-investigator, or a research assistant. Consent to
participate in all the aspects of the study was seemed on the day of the study. The
mother consented on behalf of her child. Focus group participants were informed in
detail at the time the focus group was convened, and all agreed verbally to participate.
Consent forms and all data were stored under the lock and key in the Evergreen
offices, Yuci City.
Needed medical care was provided to any subjects participating in the study.
First, a qualified pediatrician was available to provide immediate attention to any
infants found to be ill. Second, so that study subjects did not suspect that we had
come to find illnesses or sell medicines, we did not treat them, charge them for tests,
or sell them medications at that time; rather, we made clear recommendations to
patients and/or parents as to what they should do for their child. Third, after all test
results were collated (about a month in the spring, due to S ARS delay), a pamphlet
was provided to each mother with a report of their childs health condition and lab
results, plus basic information on rickets prevention and/or treatment.
RESULTS I: QUANTITATIVE
The main research questions of this study are:
Question 1: What is the prevalence of rickets among children in north China?
Question 2: What are the determinants of rickets in Chinese children?
Question 3: What health outcomes are associated with vitamin D-deficiency and
Question 4: How do mothers and Chinese health workers in rural Shanxi
Province, China perceive rickets as a health problem?
The first three questions will be addressed in this chapter, and the fourth will
be primarily explained in the following chapter presenting qualitative results. First
the study population will be described, followed by descriptive results regarding the
blood chemistry and illness outcomes of these children. This information is
necessary for understanding the overall health of these children, which will then aid
in understanding the problem of rickets introduced afterwards.
Children and Mothers
As shown in Table 4.1, the children in this study ranged in age from 12 to 24
months. The gender distribution showed a higher percentage of male participants,
but this imbalance is a reflection of Chinas current gender disparity. The
implementation of the Family Planning Policy in 1979 limits one child to each
family. The gender disparities can be explained in part by the strong preference for
boys, and the slight excess of male infants in most populations. In China,
currently 117 boys are bom for every 100 girls, precisely the ratio found in this study.
Table 4.1. Descrii ptive statistics of child subjects.
Number of subjects 250 176
Mean age (months) 18.2 3.7 23.3 3.8
Age range (months) 12-24 17-29
Male 52.4% 53.4%
Female 47.6% 46.6%
The average age of mothers in the survey was 28.26 years 4.17 in the
spring. These mothers were older than may have been expected in rural China where
marriage and child-bearing are believed to begin at age 20. Partial explanation is
that for nearly 30% of them, it was their second or third child. But even so, it shows
that Chinas family planning promotion of later marriages and later child-bearing has
had a significant effect in delaying child-bearing.
Of these 250 participants 3 were the adoptive mother of the child examined, 2
were fathers speaking on behalf of their wives, and 2 were paternal grandmothers
speaking on behalf of the childs mother. Because the study was stopped early due
to SARS, limiting the final total number of cases, and after discarding 12 subjects
because they did not meet the study criteria, it was decided to keep these 7 in the
study because their answers did not differ significantly from the mothers who
answered for themselves.
69.5% of mothers surveyed have completed only junior high school (9 years
of education) and 10.9% having only completed primary school. 2.8% are illiterate
and 16.8% have completed senior high school or college. 43.6% of these mothers
know at least one way to prevent rickets in their child.
Participation Rate. The participation rate in this study was 80.5% in the
spring round of data collection and 57.3% in the fall (Table 4.2). The return rate in
the second round of data collection was 71.26%. Village collaborators reported that
some mothers felt that they had already learned what their childs health situation
was from the first round of data collection and didnt considered it worth their while
for their children to undergo the same tests.
Table 4.2. Subject participation by village.
Spring Fall Return rate
Tow n- ship Nam e Vill- age Nam e Villag e Total Popul- ation #of eligibl e child- ren # who partici paled Partici -pation Rate # who partici pated # who did not return Partici- pation rate
Xiuwen Nan yao 1443 22 22 100.00 % 17 5 77.27% 77.27%
Xiu wen 1403 20 14 70.00 % 11 3 50.00% 78.57%
Chen kan 1234 22 18 81.82 % 14 4 63.63% 77.78%
Chu cun 1403 20 18 90.00 % 12 6 60.00% 66.67%
Guo cun 2875 42 21 50.00 % 16 5 38.10% 76.19%
1 Zhangqing Zhan g qing 3621 57 46 80.70 % 28 18 49.12% 60.87%
Yong kang 3475 47 42 89.36 % 30 12 63.83% 71.43%
Beih u qiao 1167 12 10 83.33 % 5 5 41.67% 50.00%
1 Changning Dong chan g ning 2284 18 18 100.00 % 11 7 61.11% 61.11%
Xi chan g ning 1994 12 10 83.33 % 7 3 58.33% 70.00%
Bage da 843 8 7 87.50 % 7 0 87.50% 100.00 %
Xian gli 689 8 6 75.00 % 6 0 75.00% 100.00 %
Ludo wan 528 10 8 80.00 % 6 2 60.00% 75.00%
Nanh e liu 1096 9 7 77.78 % 6 1 66.67% 85.71%
Total 14 24055 307 247 80.46 % 176 71 57.33% 71.26%
Urba n 3 0 3
Total surveyed 250 176 74
Blood Chemistry Analysis
As a measure of health and nutritional status, blood tests were done on
participating children in both the spring and fall rounds of data collection. These
results are tested against rickets and other health outcomes in sections to follow.
Standards for normal and percentage of children with blood test results out of normal
are reported in Table 4.3.
Table 4.3 Summary of spring and fall blood chemistry.
Test norm Spring Percent abnormal Fall Percent abnormal Comments
Hgb <11.5 gm/dl 14% 27.8% Less meat and milk consumed in the summer Lots of activity need more. Growing fast
Zinc <0.60 gg/ml 14.8% 39.1%
Calcium <90 gg/ml 1.8% 14.4%
ALP > 250IU/L* 10.9% 10.5%
Lead>10gg/dl 63.5% 54.9% Many children are at risk for neurological damage due to lead poisoning
25-OH-D <12 ng/ml 65.3% 2.8% This degree of vitamin D deficiency is very large
* The Chinese BALP Test Kit used suggested the fo lowing diagnostic criteria: <200 IU/L Normal;
201-249 Subclinical rickets; >250 Clinical rickets.
Hemoglobin. Iron and Zinc. Vegetables and fresh fruit are rarely available in
rural Shanxi in the winter, but they are abundant in the summer. It would thus seem
that the summer diet would be more balanced and more nutritious than the winter
diet. It was surprising to find that blood levels of hemoglobin, zinc and calcium had
declined from spring to fall (see Table 4.2). The two microelements assumed to be
provided in more sufficient amounts with increase in age hemoglobin and zinc -
both declined significantly. The percentage of children with blood Hgb levels below
11.5 gm/dl nearly doubled, increasing from 14% to 27.8%. This represented a mean
Hgb level decline during the 5 month interval from 12.5 to 12 gm/dl (t=4.07, dfl46,
p=.000), with only 4.1% of children below 10 gm/dl. This compared with 1.8%
below 10 gm/dl in the spring time. So the worsening of Hgb status, possibly related
to decline in iron intake, is consistently observed.
Both Hgb, dependent on iron, and zinc levels are related to meat intake in the
diet. Local health workers reported that children eat very little meat in the summer
because it is too hot after being cooked so it is uncomfortable to eat in the summer,
and difficult to store (Figure 4.1)1. Finally, hot days increase thirst but decrease the
appetite. Mothers report that in the summer children consume large amounts of
sweetened drinks which suppress appetites at meal times. The proportion of children
who drink sweetened drinks often or occasionally increased from 47.3% in spring
to 83.4% in fall.
Those reporting drinking milk were not the same ones who reported drinking
sweetened drinks, so it is unlikely that sweetened drinks1 2 were mistakenly reported
as milk. There was no relationship between those reporting drinking milk,
sweetened drinks and soy milk.
Figure 4.1. Percentage reporting eating these foodstuffs often in the recent month.
1 Only 35% of families report owning a refrigerator, many of which are not actually being used
because people arent used to using them, and they consume a lot of electricity which people are
loathe to waste.
2 These sweetened drinks, such as WaHaHa, report to have miniscule amounts of milk in them, and
call themselves milk, so some people assume they contain substantial amounts of milk, which they
These changes in diet may account for the declines in zinc and iron levels.
In terms of protein intake, the decline in egg, soy and meat consumption was
compensated for by green beans.
Zinc Levels of Surveyed Children Spring Zinc Levels of Surveyed Children Fall
.14 30 .45 .61 77 SB 1.08 .14 .28 .43 .58 .72 .67 1.02 1.16
22 .38 .53 .69 .85 1.00 1.18 21 .36 JSO .85 .80 .84 1.08
Serum zmc (rrtcrogrrfdl) Serum airtc (rricrogm/ml)
Figure 4.2. Zinc levels in spring. Figure 4.3. Zinc levels in fall.
Zinc insufficiency in populations tends to result in more frequent and severe
infections, including diarrhea and pneumonia. Mean zinc level declined from .77
.19 ng/ml to .66 .21 gg/ml from spring to fall, and this decline was statistically
significant (paired samples t=.3.3, p=.001, n=87). Also, the number of children
whose serum zinc levels fell below .60 gg/ml increased from 14.8 to 39.1%. from
spring to fall. The small decline in intake of zinc-containing foods coupled with a
significant decline in serum zinc levels shows a trend toward worsened zinc status
from spring to fall (Figures 4.2 and 4.3).
Vitamin D. It has long been held that rickets is primarily a result of vitamin
D deficiency. 65.3% of surveyed children were vitamin D-deficient (25-OH-D less
than 12 ng/ml) in the spring cycle of data collection. In this northern climate, with
extended time indoors in the winter, and only 4.2% of children given vitamin D
supplements often in the recent month (see Table 4.4), vitamin D-deficiency is
inevitable (Table 4.5). After a summer of sun exposure the percentage vitamin D-
deficient was reduced to 2.9%.
Table 4.4. Percentage of children supplemented with vitamin D or calcium in the
________________________recent month (spring data).____
(n the last month did your child eat supplements?
Vitamin D Supps (n=237) Calcium Supps Percent (n=245)
No 90.3 59.2
Occasionally 5.5 20.8
Often 4.2 20.0
Total 100.0 100.0
Table 4.5. Summary of 25-OH-D levels of children surveyed in spring and fall.
25-OH-D level (ng/ml) Number (n=179) % Number (n=172) %
<5 60 33.5% 0 0
5-11.9 57 31.8% 5 2.9%
12-19.9 34 19.0% 9 5.2%
21-39.9 21 11.7% 97 56.4%
>40 7 3.9% 61 35.4%
Total 179 172
Calcium. Mean calcium levels declined from 152.330.8 to 117.527.9
pg/ml (t=9.0, p=.000, 105 df) from spring to fall. This represents an increase in the
number of children whose serum calcium was insufficient (defined as serum
calcium <90 pg/ml) from 1.8% to 14.4%. This decline in serum calcium levels
occurred during the same time that clinical rickets rate declined from 41.6 to 17.0%.
Although serum calcium is not a diagnostic test for rickets, this observation should
be investigated in more detail, as is done below.
All the blood tests showed a statistically significant change from spring to
fall (Table 4.6).
Table 4.6. Comparison of means of lab test results and in spring and fall.
Test Abnormal standard Spring Fall Paired samples t-test
n Mean SD n Mean SD
Blood Hgb <11.5 gm/dl 228 12.5 1.2 170 12.0 1.2 4.07, p=.000
Serum Zinc <0.60 pg/ml 87 .77 .19 87 .66 21 3.328, p=.001
Serum Calcium <90 pg/ml 165 152.4 29.8 171 116.0 28.3 9.007, p=.000
ALP >250IU/L 230 224.4 18.7 174 219.7 17.9 1.89, p=.060
Lead >10pg/dl 217 15.7 11.0 175 12.4 12.5 6.469, p=.000
Serum 25-OH-D <12 ng/ml 177 13.8 29.7 172 51.1 57.4 -5.703, p=.000
Serum calcium declined significantly from spring to fall, but the mean
calcium levels were still well above the normal range of 90 pg/ml. However, the
percentage of children whose calcium level fell below the normal range of 90 pg/ml
increased from 1.8 to 14.4%. It must be asked whether this decline in serum
calcium is causing health problems. For example, one common result of calcium
insufficiency is night terrors (spasms) while sleeping. In the spring, 14% of mothers
reported their children to experience night terrors. Levels of serum calcium were
found to be negatively correlated with having night terrors (partial correlation
controlling for gender and age =-.178, p=.012), suggesting that the decline in serum
calcium levels was leading to calcium insufficiency and possibly to negative health
outcomes. Correlation between having night terrors in spring and fall is .362
(p=.000), so the same children are continuing to experience this problem. The
possibility of calcium insufficiency among these children exists.
Gender and Blood Chemistry. There is a positive relationship between being
male and blood 25-OH-D levels (rho=.13, p=049) and between males and blood
calcium levels (rho=.132, p=.045). Male children have higher blood levels of both
25-OH-D and calcium. Independent samples t-test shows a difference in the mean
levels of these two blood tests by gender, although the statistical significance is not
strong (see Table 4.7).
Table 4.7. Gender differences in serum 25-OH-D (ng/ml) and serum calcium
Childs gender N Mean Std. Deviation T-test
Serum 25- OH-D3 (ng/ml) Female 84 10.21 12.91 -1.76 (p=.082)
Male 93 16.96 34.49
Serum calcium (microgm/ ml) Female 81 148.51 28.43 -1.64 (p=. 103)
Male 84 156.08 30.74
Equal variances not assumed.
This leads one to ask whether there are dietary or behavioral differences
between boy and girl children. There is some difference between gender and the
perceptions mothers have of themselves and of their children, and of gender and
meat consumption (see Table 4.8). Only 12.7% of mothers of daughters considered
themselves to be very healthy, but 20.2% of mothers of sons considered
themselves to be very healthy (Z=1.72, p=.085). Mothers of sons are either truly
healthier, or they are just feeling better about themselves.
When asked if they considered their child to be very healthy, only 8.5% of
mothers of sons agreed, but 14.3% of mothers of daughters agreed. Independent
samples z-test confirms the significance of these differences (z=-1.79, p=074).
Mothers of sons are less convinced of their sons good health than mothers of
daughters, even though the data will show the only difference in health outcomes
between boys and girls is prevalence of bronchitis (males 22.1%, females 12.8%,
Spearman rho=.12, p=.056) and prevalence of rickets as measured by clinical
Table 4.8. Health perception and dietary differences between boy and girl children
Childs gender N Mean Std. Deviation Independet samples Z- test
In the last month did /our child eat meat (no, occasionally, or often)? 0 Female 119 1.93 .634
1 Male 130 2.07 .625
What do you perceive your child's health to be? 0 Female 119 1.92 .443
1 Male 130 2.02 .431
What do you perceive your own (mother) health to be? 0 Female 118 1.92 .405
1 Male 129 1.82 .441
Of further interest is the apparent concern of po-po for grandsons over
granddaughters. When asked how significant their po-pos role is in child-rearing,
29% of mothers of sons said great influence, while only 18.3% of mothers of
daughters said great influence. This relationship was shown to have a Spearmans
correlation of -.15 (p=.048). Mothers and po-po are more concerned about
grandsons than granddaughters.
It is possible that having a son elevates ones status in the village. When
asked, If you use a new method of child-rearing, what is the attitude of the
villagers? 36.6% of mothers of sons replied that they would be supported, while
only 23.3% of mothers of daughters said they would be supported (rho=-.132,
p=.08). While this relationship is not strong, it supports the trend toward gender
bias in favor of sons. In rural China, the birth of a son is heralded with much
fanfare, while the birth of a daughter is hardly even announced, being considered a
If asked whether their child had eaten meat often in the last month, 23.1%
of mothers who had a son said yes, but only 16.8% of mothers of daughters (z=-1.71
(p=.089). There were no differences found in milk drinking patterns among boys
Surveyed girls were found to be weaned from the breast about a month
earlier than boys (15.4 versus 16.4 months), however this difference was not
statistically significant (t=-l.l 1, p=.268).
Boy children were found to have higher blood 25-OH-D and calcium levels
than girls. Boy children eat more meat than girls, but no differences are found on
other dietary variables. Mothers of sons consider their own health to be better than
mothers of daughters, while mothers of sons consider their sons health to be poorer
than mothers of daughters. Po-pos influence in child-rearing is greater with a
grandson than a granddaughter. Mothers of sons feel themselves to have more
status in the village as measured by their ability to freely use new child-rearing
methods without being criticized. Clearly there is a parenting psychology biased
toward boy children that may be influencing boys higher serum levels of vitamin D
and calcium. Even though this study has not been able to determine the causal
factor, it is likely working through preferential treatment toward boy children.
Illness History of Children
Illness history in this study was dependent on maternal recall. Table 4.9
below shows that rates of pneumonia, bronchitis, URI and rickets declined a
significant amount from spring to fall. The rate of diarrhea declined an insignificant
amount and the rate of anemia (reported) increased an insignificant amount.
dren from spring to fall.
Illness Percent with this Illness p-value (chi- square) n Comments
Diarrhea 25.6% 20.2% .141 175 Expect more diarrhea in summer
Pneumonia 9.2% 0% .000 173 Attributable to 25-OH-D increase?
Bronchitis 17.7% 6.1% .000 173
URI 14.1% 7.4% .032 174
Anemia 1.6% 3% .301 174 Hgb levels also declined
Wt-Age z score <-2 SD 3.3% 4.3% .422 169
Ht-Age z score <-2 SD 5.3% 6.4% But mean z- score increased
Rickets (clinical) 41.6% 17% .000 175
Discussion with Dr. Eric Simoes, who does research on rickets and illness in
the developing world (Muhe et al. 1997), suggested that in the developing world it is
reasonable to expect an attack rate for diarrhea in children of 1 episode/child/year,
and an attack rate for pneumonia in children of 0.5 episodes/child/year (Simoes
2002). Spring illness results show an attack rate for diarrhea of .51/child/year, and
for pneumonia of .18/child/year. These results are somewhat artificially low
because it does not account for multiple cases of an illness in the same child.
Mothers were simply asked if their child had a given else at all in the recent six
months. However, the illness burden in these children is less than was anticipated
using crude prediction estimates.
Height for age z-scores showed an increase from spring to fall, so the
childrens height is getting closer to the NCHS standard. This may reflect the
remarkable healing of rickets during that same time period. The percent of
malnourished children (weight-for-age z-score below 2 standard deviations below
the mean on NCHS standards) increased from spring to fall.
Table 4.10. Paired sample means and t-test results of growth parameters from spring
Mean (SD) Mean (SD) t-test p-value n
Ht-age z-score -.41 (1.19) -.28 (1.07) -2.58 .011 169
Wt-age z-score -.08 (1.12) -.25 (1.08) 3.98 .000 169
In summary, through the summer these childrens weight-for-age decreased,
height-for-age increased and they have lower blood levels for most of the micro and
macronutrients tested in this study. One can assume that during the summer, these
growing children experienced more overall activity than in the previous winter.
This increase in physical activity coupled with growth demands, increase childrens
nutritional needs. Based on dietary and blood micronutrient measures, it appears
that this need is not being met. The slowing in weight gain may be a consequence
of failing to meet the nutritional demands of children during the summer season.
However, as of the fall, these children experienced less illness and much of
the rickets they had was healed, so something beneficial was occurring as well. The
two most obvious changes are development and increased sun exposure.
Rickets is a multifactorial disease, so that the method of diagnosing rickets
varies in different contexts and different countries. Furthermore, some of the most
well validated methods of diagnosing rickets, such as analysis of the wrist x-ray, are
not appropriate methods to use in contexts of poverty or for screening purposes. It
is costly to the family, and introduces unnecessary radiation exposure to the vast
majority of those children screened and found to be negative. Analysis of the value
of the clinical signs and symptoms which are used by health workers to diagnose
rickets in China is therefore extremely important.
Chinas method of diagnosing rickets based on clinical observation has been
used for a long time. It has guided Chinas rickets researchers and pediatricians in
the 30-year long struggle to control rickets. However, it has been increasingly
recognized that clinical signs and symptoms alone are not sufficient evidence to
diagnose rickets (Thacher et al. 2002). Because clinical signs and symptoms may
represent healed rickets, the results of other illnesses, or represent a normal range of
different somatotypes, they cannot provide a definitive diagnosis of active disease.
In the following sections several methods of diagnosing rickets will be
introduced, along with the rickets rates found by each method. The purpose is to
calculate an accurate prevalence of active rickets in these children, to determine
which signs and symptoms should be confidently used in clinical diagnosis of
rickets in China and to suggest the most reliable way to diagnose rickets in the
Chinse context. These methods may also be useful to application in diverse
developing country contexts.
The methods of diagnosis along with the terms used to refer to them in this
dissertation are as follows:
Clinical Diagnosis Clinical
X-ray Results only Radiography
Clinical, X-ray and 25-OH-D Results Clin, X and D
Clinical and Alkaline Phosphatase Results Clin, ALP
Clinical, X-ray and Alkaline Phosphatase Results Clin, X and ALP
Clinical Method Using Only the 5 Significant Clinical Signs Sig.
Clinical Diagnosis Method (Clinical!
Since 1984 China has been using a clinical standard for diagnosing rickets
based on a list of signs and symptoms (Table 4.11). A symptom is what the
patient/parent reports as bothersome to them. In the table below, this includes the
first three conditions night terrors3, restlessness and hyperhidrosis. A sign is what
the clinician sees that is abnormal. This includes all the remaining conditions
presented in Table 4.11.
3 The word night terrors, which sounds strange to English readers, comes from the Chinese phrase for
this symptom, ye jing. It really means restlessness or spasms while sleeping.
Table1 .11. Standards for clinical signs and symptoms of rickets.
Condition Classification Points Standard for a positive symptom
Night terrors Symptom 1 + Restless sleeping, Sequent spasms of the limbs while sleeping
Restlessness Symptom 1 +Easily irritated, doesnt want to stay still
Hyperhidrosis Symptom 1 + Excessive sweating unrelated to ambient temperature, especially when nursing or crying
Leg pain on walking Primary sign 1 +Child refuses to walk, cries when forced to walk, or exhibits pain when walking
Occipital rubbing Secondary sign 1 + Hair on the back of the head is rubbed off
Frontal bossing, Primary sign 1 + Squarish shaped head, X- or saddle-shaped lines visible at the suture lines
Delayed closing of anterior fontanelle Secondary sign 1 +Anterior fontanelle closes after 18 months of age
Delayed teething Secondary sign 1 +First tooth erupts after ten months of age
1 + Can feel the breast protuding
Pectus carinatum (chicken breast) Primary sign 2 -H- Can see the breast protruding
Primary sign 1 +- Sternum is inverted slightly (exclude congenital cases)
Funnel chest 2 ++ Sternum is inverted severely (exclude congenital cases)
1 -i- Bead-like enlargements palpable on the costochondral cartilages of the ribs
Pearling Primary sign 2 -H- Can see bead-like enlargements on the costochondral cartilages on the ribs
Harrisons groove Primary sign 1 + Outward flaring of costal margins, along with inward bending of the ribs
1 + <3 cm distance between the knees while standing
0 (Bow) legs Primary sign 2 ++ >3-6 cm distance between the knees while standing
X-legs (Knock knees) Primary sign 1 + <3 cm distance between the ankle bones while standing
2 H- >3 cm between the ankle bones while standing
Wrist enlargement Primary sign 1 +Wrists excessively wide, in China called bracelet
+ Moderate, ++ Severe
The Chinese clinical standard defines rickets as the presence of 2 primary
signs plus one symptom, or 2 primary signs plus 2 secondary signs. The variables
used in this method of diagnosis, and their frequencies in spring and fall are
presented in Table 4.12 below. Hyperhidrosis (excessive sweating) is the only
symptom or sign that increased significantly from spring to fall. Most signs and
symptoms decreased in frequency, suggesting a decline in the rate of rickets from
spring to fall.
Table 4.12. Presentation of rickets signs and symptoms in spring and fall.
Symptoms or Signs Percentage Presenting this Symptom (%)
Night terrors 14 14
Restlessness 25 23
Hyperhidrosis 11 25
Leg pain 2 4
Occipital rubbing (zhentu) 25 12
Frontal bossing (fanglu) 45 28
Delayed closure of anterior fontanelle 18 12
Delayed teething 12. 1.5
Pectus carinatum 17 16
Funnel-shaped chest 0.4 0.5
Pearling 74 67
Harrisons groove 72 46
O-shaped legs 5 1
X-shaped legs 5 1
Wrist widening 5 1
Because of the difficulty inherent in interpreting the clinical symptoms that
the mothers report, and because of the non-specificity of these symptoms (night
terrors, restlessness, and sweating), the three self-report symptoms were excluded
from the clinical diagnostic criteria used below. Additionally, because of the high
frequency that many of clinical signs were reported, a decision was made to narrow
the clinical diagnostic criteria. The resulting method was significantly more
stringent, requiring 3 primary signs or 2 primary signs plus 2 Secondary signs for a
positive diagnosis of rickets. Chinese research colleagues, all long-time rickets
researchers, agreed to these changes for the purpose of better discrimination in the
context of the research. Applying these more stringent criteria, the prevalence of
rickets was 41.6% in the spring (n=250) and 17% in the fall (n=188).
It could be argued that rickets in China begins developing prenatally,
worsens during the period of post-partum cloistering, or zuoyuezi, worsens or
improves depending on whether it is winter or summer, and then finally improves
significantly after the age of 18 months when the children are outside more and
eating a more varied diet. Thus, it must be asked whether the children in this study
had rickets before age 12 months, and have been gradually healing and will continue
to heal as they age. Controlling for season, the only signs and symptoms of rickets
that declined with aging were occipital rubbing (spending less time in bed as they
grow older) and frontal bossing. These data suggest that the healing of rickets in
these subjects is not entirely a function of aging that began before the age of one.
Paired sample correlation of clinical rickets diagnosis in the spring and fall
is .234 (p=.001). This observation merits two explanations. Some childrens rickets
persisted from spring to fall while many children showed signs of healing. This
result also demonstrates consistency in diagnosis by the examiners from spring to
fall so that children not healed were consistently found at both time points to have
rickets. The Mantel-Haenszel chi-square statistic is 8.356 (p=.004), showing that
the observed decline in rates of clinical rickets was significant.
Validation of Clinical Criteria for Rickets Diagnosis. The next step in
evaluating diagnostic methods was to test which of the clinical criteria for
diagnosing rickets was found to be associated with other more objective methods,
such as radiographic analysis and measures of alkaline phosphatase.
Simple correlation analysis comparing clinical signs and symptoms and x-
ray results show that only the sign of 0-shaped legs correlated positively with
radiography (.132, p=.056) as read by the Western radiologist. None of the clinical