Sex differences in lipids and body shape, but not diabetes, increase at puberty. Hong Kong Chinese are mainly first or second generation migrants from China, who have shared an economically developed environment for years, but grew up in very different environments in Hong Kong or contemporaneously undeveloped Guangdong, China. We assessed if environment during growth had sex-specific associations with lipids and body shape, but not diabetes.
Methodology and Principal Findings
We used multivariable regression in a population-based cross-sectional study, undertaken from 1994 to 1996, of 2537 Hong Kong Chinese residents aged 25 to 74 years with clinical measurements of ischaemic heart disease (IHD) risk, including HDL-cholesterol, ApoB, diabetes and obesity. Waist-hip ratio was higher (mean difference 0.01, 95% CI 0.001 to 0.02) in men, who had grown up in an economically developed rather than undeveloped environment, as was apolipoprotein B (0.05 g/L, 95% CI 0.001 to 0.10), adjusted for age, socio-economic status and lifestyle. In contrast, the same comparison was associated in women with lower waist-hip ratio (−0.01, 95% CI −0.001 to −0.02) and higher HDL-cholesterol (0.05 mmol/L, 95% CI 0.0004 to 0.10). The associations in men and women were significantly different (p-values<0.001). There were no such differences for diabetes.
Citation: Schooling CM, Lam TH, Thomas GN, Cowling BJ, Heys M, Janus ED, et al. (2007) Growth Environment and Sex Differences in Lipids, Body Shape and Diabetes Risk. PLoS ONE 2(10): e1070. https://doi.org/10.1371/journal.pone.0001070
Academic Editor: J. Jaime Miranda, London School of Hygiene and Tropical Medicine, Peru
Received: June 11, 2007; Accepted: October 5, 2007; Published: October 24, 2007
Copyright: © 2007 Schooling et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The study was supported by the Hong Kong Health Services Research Committee (HSRC# 411026), the University of Hong Kong Committee on Research and Conference Grants, the Hong Kong Research Grants Council (grant 407/94m) and the Hong Kong Society for the Aged. The funding source had no involvement in the design, collection, analysis, and interpretation of data; in the writing of the report; or in the decision to submit the paper for publication.
Competing interests: The authors have declared that no competing interests exist.
Secular trends in ischemic heart disease encompass several features without clear explanation. Firstly, there is a widening of sex difference in premature ischemic heart disease (IHD) with economic development , , which cannot be fully explained by oestrogen in women  or smoking in men . Secondly, the social patterning of ischemic heart disease in men in western populations reversed with economic development over the 20th century, from association with higher to lower social position , although this may be the case for women . Mounting evidence from locations outside long-term industrialised countries makes it increasingly clear that the prevalence and social patterning of IHD or its risk factors is not universal but is epidemiologic stage specific –. Finally, in countries with a historically more rapid and recent trajectory of economic development from pre-industrial levels, such as South Korea , there is evidence of social inequalities in IHD risk in women but not men . In addition, in long-term industrialised countries with a longer history of economic development there is sometimes a more marked pattern of social inequalities in IHD in women than men –, so that the social patterning of IHD risk particularly in men appears to be related to the history and trajectory of socio-economic development.
A (set of ) factor(s) associated with economic development which specifically and universally increased the risk of IHD in men would explain these trends, because it would lead to greater male risk with economic development; it would occur first in the socially advantaged and it would obscure or reduce the generally protective effect of social advantage. In industrialized countries it is well known that a more atherogenic lipid profile and body shape emerges in boys at puberty whilst the reverse happens in girls, due to the action of sex-steroids –. These sex-specific pubertal changes have previously been suggested as contributing to the earlier development of IHD in men . In contrast, although diabetes is a risk factor for IHD, diabetes risk appears to be more strongly related to another hypothalamic-pituitary-endocrine axis, i.e. growth hormone  and so should not be affected by pubertal sex-steroids –. With economic development over the last century, there has been little increased rate of diabetes in men compared with women . Figure 1 shows these hypothetical causal pathways. However, it is not known whether these sex-specific pubertal changes are a universal environmentally independent feature of growth and development or an epidemiologically stage specific product of growth in an economically developed environment.
Men and women currently living in Hong Kong share an adult environment, but grew up in very different environments depending on whether they migrated from pre-industrialized Guangdong province or were born in Hong Kong, which was relatively more economically developed than Guangdong province throughout the 20th century.
Most Chinese people in Hong Kong are first or second generation migrants from southern China, mainly from the neighbouring province of Guangdong. The major waves of migration were in the late 1940s, early 1950s and the late 1970s . Migrants were mainly young people looking for work , with, until very recently, little migration back to China after 1949. At that time gross domestic product (GDP) per head in China was similar to pre-industrialized western Europe . Guangdong province was an economic backwater until the establishment of the Special Economic Zones near Hong Kong in 1978. In contrast industrialization and economic growth started in the first half of the 20th century in Hong Kong . By the 1930s living conditions in Hong Kong were better than in the neighboring provinces with workers sending remittances back to China . By 1952 per capita GDP was half that of western Europe and by 1995 had overtaken western Europe . We have previously used this natural experiment to examine whether childhood migration was related to self-reports of diabetes, hypertension, hypercholesterolemia or heart disease, obtained from telephone interviews . We previously found no difference in the age-adjusted risk of self-reported IHD (and/or angina) by sex or growth period spent in China or Hong Kong, although self-reported IHD was more common in women than men, suggesting the self-reports were incomplete, thus biasing any relation towards the null. In the present study, we used clinical measurements in a sub-sample to investigate whether the associations of some specific dimensions of IHD risk (body shape and lipid levels) but not others (diabetes) varied by sex according to growth environment location (Guangdong, Hong Kong or both) for Hong Kong residents.
Materials and Methods
The sampling frame and methods of this population-based cross-sectional study in Hong Kong have been described elsewhere in detail . In brief, a probability sample was used to obtain the WHO MONICA  recommendation of about 200 participants in each of five 10-year, adult, sex-specific age groups. Randomly generated telephone numbers were used to select households, and within households a Kish grid was used to randomly select respondents. Almost all Hong Kong households (99%) have a telephone . From 1994 to 1996, 7174 Chinese (25–74 years) took part in telephone interviews (a response rate of 78%). The study was supplemented by a relatively small number (556) age 55–74 specifically recruited because there were fewer people of this age in the population. Of the 7730 interviewed, 2900 (38%) came for physical examination. The attendees and non-attendees match the population and non-attendance bias should be small . Appendix S1 shows a comparison of the total sample and those who came for physical examination by socio-demographic characteristics and self-reported health status. For place of birth, exercise, smoking status, self-reported medically diagnosed diabetes or hypertension and general health, the Cohen effect sizes were negligible (<0.1). Effect sizes for job activity, a proxy for socio-economic status, (0.16) and education (0.23) were slightly larger but still acceptable. We excluded pregnant women and those with serious diseases such as cancer and those who were hospitalized. The detailed methods of measurement have been reported . The University of Hong Kong Ethics Committee approved the study and participants gave prior, written, informed consent.
Our primary exposure is growth environment obtained from place of birth, age and length of stay in Hong Kong. Growth environment was categorized as: 1) pre-industrial, i.e. born in Guangdong province and migrated to Hong Kong at age 20 years or older, 2) mixed, i.e. born in Guangdong and migrated to Hong Kong before age 20 years, and 3) more developed, i.e. born and raised in Hong Kong. We chose age 20 years as the cut-off to allow for the completion of possibly slower growth in Guangdong province and to maximise the limited number who migrated as adults given digit preference for migration at age 20 years. We specifically used a separate group for people who migrated before the age of 20, because they experienced potentially detrimental environmental change during growth , even though their average experience might be similar to either group. We excluded people born elsewhere to reduce confounding by other unmeasured differences.
Outcome specification of IHD risk
The main outcome measures were risk factor for IHD in which differences between the sexes become more pronounced at puberty due to the action of sex steroids. At puberty, HDL-cholesterol, body shape and possibly apolipoprotein B (ApoB) change towards a less cardio-protective pattern in males and a more cardio-protective pattern in females –, –. Body shape was proxied by waist-hip ratio. HDL-cholesterol, ApoB and waist-hip ratio are all risk factors for IHD, although not necessarily optimal predictors –.
A secondary outcome measure was a risk factor for IHD whose precursors also change at puberty but more likely as a function of growth hormone than sex-steroids, i.e. type 2 diabetes –. Diabetes is also a well established risk factor for IHD . Type 2 diabetes was classified using the World Health Organization definition of self-report of previously medically diagnosed diabetes or a fasting plasma glucose ≥7.0 mmol/L or a post-load reading of ≥11.1 mmol/L for those who completed a standard oral glucose tolerance test (consisting of 95% of those not on diabetic medications).
Chi squared tests were used to compare characteristics between groups. Locally weighted regression and scatter plot smoothing (LOWESS) was used to examine risk factors across the age-range. The association of each risk factor with growth environment adjusted for potential confounding factors was assessed in men and women separately using multiple regression. Mean differences and 95% confidence intervals (CIs) were calculated from linear regression for continuous outcomes. Odds ratios and 95% CIs were calculated from logistic regression for diabetes. Whether growth environment had a different association in men and women was assessed from the heterogeneity across strata and the significance of interaction terms. We tested for interactions by running models with and without the interaction term and examining the statistical significance of the likelihood ratio test of the difference between the two models on the relevant chi-squared distribution. A two-tailed p value of less than 5% was considered statistically significant.
Potential confounding factors considered were education, type of housing, use of alcohol, smoking and exercise categorized as in Table 1 and age. First we adjusted for age (model 1). Next, we added all other confounders (model 2). Finally, we also adjusted HDL-cholesterol, ApoB and diabetes for waist-hip ratio and body mass index (model 3). Each analysis used all relevant complete data (available for 97%).
Of the 2900 participants, 5 were excluded because they were pregnant, 1 was excluded because place of birth was unknown, 350 were excluded because of birth outside Guangdong or Hong Kong, and 7 migrants from Guangdong were excluded because their duration of residence in Hong Kong was unknown. Of the remaining 2537, 1586 were Hong Kong born and 951 were migrants from Guangdong (Table 2). Just over half the migrants (482) had grown up in Guangdong and mostly migrated to Hong Kong in their 20s (72%) and had lived in Hong Kong for many years; mean 28.7 years in men and 26.9 years in women. The other 469 had migrated from Guangdong to Hong Kong before the age of 20 years.
People born and raised in Hong Kong were more likely to have attained higher education than the other two groups (Table 1). In contrast men from Guangdong were more likely to smoke, take no exercise and report poorer self-rated health than the other groups and these associations remained after adjustment for age (data not shown). There was no use of lipid-lowering medication.
Differences by growth environment in waist-hip ratio and ApoB in men and women were present across the age-range, although less clearly so for HDL-cholesterol particularly in men (Figure 2). However there were no obvious differences for diabetes.
Waist-hip ratio and ApoB were higher in men who had grown up in Hong Kong than in men who had grown up Guangdong (Table 3), adjusted for age (model 1) or additionally adjusted for socio-economic status and lifestyle (model 2). In contrast, waist-hip ratio was lower and HDL-cholesterol was higher in women who had grown up in Hong Kong than in women who had grown up in Guangdong, adjusted for age (model 1) or additionally adjusted for socio-economic status and lifestyle (model 2). These different associations with waist-hip ratio, HDL-cholesterol and ApoB for men and women by growth environment were significant. Further adjustment of HDL-cholesterol and ApoB for waist-hip ratio and body mass index (model 3) attenuated the associations.
Diabetes risk was similar for men and women who grew up in Hong Kong or Guangdong (Table 4) in all three models, although men and women who had experienced environmental change during growth tended to have a higher risk of diabetes. There was no evidence that the association of diabetes risk with growth environment varied between men and women. This finding was not due to lack of power due to using a dichotomous outcome, as there was also no evidence of different associations with fasting plasma glucose or post-load plasma glucose by growth environment in men and women (interaction p-values 0.28 and 0.41 respectively in model 2).
Taking advantage of a unique natural experiment generated by migration, we found that different environments during growth were associated differently in men and women with the IHD risk factors affected by sex-steroids at puberty. Men, who had grown up in a relatively more economically developed environment (i.e. Hong Kong) had a more male shape and less favourable lipids than men who had grown up in a pre-industrial environment (economically undeveloped Guangdong). In contrast, women who had grown up in a relatively more economically developed environment had a more feminine shape and more favourable lipids than women who had grown up in a pre-industrial environment. These findings on lipids were attenuated slightly by adjustment for obesity, as wait hip ratio and body mass index suggesting there may be some underlying commonality. However, growth environment was not related to diabetes in men or women, although environmental change during growth to a more developed environment was associated with a non-significantly higher risk of diabetes.
To our knowledge, these findings are unique. There are few other ethnically homogeneous, developed populations, whose adult environments are similar, but whose growth environments were radically different. Nor, have male-female differences in IHD risk factors specifically been reported in such a setting, although there are some similar male-female differences in the long-standing Japanese migrant community in Brazil . However, our findings are consistent with the ecological observation that a female preponderance in obesity disappears with economic development . Moreover, our specific observation that growth environment can have different associations in men and women with HDL-cholesterol and waist circumference but not a pre-cursor of diabetes (higher Hba1c) has previously been demonstrated, but not explained, in the United Kingdom . On the other hand our findings are not consistent with IHD or its risk factors being associated with poor childhood conditions , . However, almost all such studies come from long-term industrialized countries where the reversal of the social patterning of ischemic heart disease in men has already taken place, and the gradation between levels of childhood conditions is not comparable with the difference between pre-industrial conditions in Guangdong and economically developed Hong Kong. There is one case-control study from Hong Kong where acute myocardial infarction had little association with education or childhood crowding, but was associated with a non-working mother, infrequent meat consumption as a child and inadequate food intake as a child . These inconsistent findings are difficult to interpret as most cases did not answer the questions on childhood diet . On the other hand, that higher levels of some risk factors were seen in the group who migrated from a pre-industrial environment to a relatively more developed environment during growth, might be consistent with accelerated or ‘catch-up’ growth increasing cardiovascular risk , . However there is little evidence that accelerated growth has different associations by sex, or is specific to lipids and body shape but not diabetes. Moreover, the biological mechanisms by which accelerated growth increases risk are not clearly understood.
Given the nature of our study there are many limitations and provisos. We took advantage of an uncontrolled natural experiment, so we are not showing precise point estimates of differences in lipids and body shape in men and women by growth environment. Economic differences between Guangdong and Hong Kong were possibly smallest during the early 20th century and greatest after 1945; nevertheless there was probably always a relative difference in favour of Hong Kong. Moreover, we found reasonably consistent differences across the age range (Figure 2). The participants may not have been completely representative; they may be better educated, however adjustment for education (model 2 compared with model 1) made little change to the estimates suggesting that education was not a major confounder. The long life expectancy and low rate of cardiovascular disease in Hong Kong make it unlikely that the findings result from premature adult deaths; after about 8 years of follow-up 29 of the 2900 participants had died from cardiovascular disease.
Total life exposure rather than exposures during growth may be the crucial factor; however we essentially compared men and women with similar total life exposures and found different associations by growth environment. Dietary data is not available for all these participants, so we cannot exclude the possibility that migrant men retained a diet specifically protective for obesity and lipids, but not diabetes, whilst the migrant women did not, although it is not easy to envisage such a diet. None of the participants reported using lipid lowering drugs, which makes treatment differences by sex and place of origin unlikely. In addition, the migrant men were more likely to smoke and to report poor self-rated health than Hong Kong born men. Moreover, the specificity of these findings to body shape and lipids but not diabetes makes it unlikely that they are the result of either different lifestyles by sex and place of origin, differential residual confounding in men and women or self-selection of healthy migrant men. Finally, only 38% of those originally surveyed attended the clinic for physical examination. Our findings would be biased if, in those from the original sample who did not attend for examination, there were systematically different associations of sex and growth environment to lipids, body shape or diabetes. Comparing the attenders with the non-attenders we found little evidence that the associations of sex and growth environment to socio-demographic characteristics or health varied with attendance status (Appendix S1).
Notwithstanding these provisos our study provides some aetiological evidence that increased risk of IHD in men compared to women may be related to growing up in a more economically developed environment. There is a plausible biological mechanism by which the dramatically improved living conditions associated with the epidemiological transition could have opposite effects on lipids and body shape in men and women, without causing sex differences in diabetes risk. Cholesterol from the diet or other sources is needed for the synthesis of sex-steroids. Animal experiments show that under-feeding reduces testosterone – and oestrogen levels  at puberty. Even in universally well-fed human populations, slight changes in diet affect sex-steroids in girls  although not in boys , where larger nutritional differences are needed . Sex steroids are responsible for the sexual dimorphism in body shape and lipids that arises at puberty , , . However, changes in pre-cursors of diabetes at puberty appear to be more strongly related to another hypothalamic-pituitary-endocrine axis, i.e. growth hormone –. Lipids and central obesity track across the life course from adolescence . We speculate that nutritionally driven increases in sex-steroids at puberty, manifested by earlier and more intense puberty seen in economically developed environments  but not in very poor environments , permanently increase sexual dimorphism in shape, HDL-cholesterol and ApoB with correspondingly increased IHD risk in men compared to women.
These findings may be consistent with a biologically plausible hypothesis, but they are only very preliminary, and require confirmation, because there are potential implications for the large proportion of the global population currently undergoing rapid economic development. Our hypothesis implies that transition to a high fat diet at or before puberty could generate a subsequent epidemic of heart disease in men and that growing boys should perhaps be a target for intervention. On the other hand in girls the potentially cardio-protective effects of the nutritional transition need to be weighed against potentially greater risk of other disease in later life.
Rather than being entirely an intrinsic sex-specific property, increased risk of IHD in men compared to women may be associated with growing up in an economically developed environment, thus providing a socio-biological explanation for the secular trends and sex differences in IHD seen with economic development. In considering the developmental origins of health, puberty, as well as fetal and early life, should be considered as a key developmental window, with potentially life-long effects on health. Understanding the long-term impacts of different patterns of pubertal development are of key relevance to the large proportion of the global population currently under-going epidemiological transition with rapidly changing patterns of pubertal development.
We would like to thank the members of The Hong Kong Cardiovascular Risk Factor Prevalence Study Steering Committee and the study staff.
Analyzed the data: CS. Wrote the paper: CS. Other: Contributed to the hypothesis, design and writing of the paper: GL. Critically reviewed the manuscript for intellectual content: BC GT TL GL MH EJ. Conceived and designed the original study: TL EJ. Advised on the hypothesis: MH. Advised on data analysis: BC. Conceived and designed the hypothesis tested here: CS.
- 1. Lawlor DA, Ebrahim S, Smith GD (2001) Sex matters: secular and geographical trends in sex differences in coronary heart disease mortality. BMJ 323: 541–545.
- 2. Nikiforov SV, Mamaev VB (1998) The development of sex differences in cardiovascular disease mortality: a historical perspective. Am J Public Health 88: 1348–1353.
- 3. Lawlor DA, Ebrahim S, Smith GD (2002) Role of endogenous oestrogen in aetiology of coronary heart disease: analysis of age related trends in coronary heart disease and breast cancer in England and Wales and Japan. BMJ 325: 311–312.
- 4. Gonzalez MA, Rodriguez AF, Calero JR (1998) Relationship between socioeconomic status and ischaemic heart disease in cohort and case-control studies: 1960-1993. Int J Epidemiol 27: 350–358.
- 5. Smith GD (2004) Life course approaches to socioeconomic differentials in health. In: Kuh D, Ben Shlomo Y, editors. A life course approach to chronic disease epidemiology. Oxford: OUP. pp. 77–115.
- 6. Ezeamama AE, Viali S, Tuitele J, McGarvey ST (2006) The influence of socioeconomic factors on cardiovascular disease risk factors in the context of economic development in the Samoan archipelago. Soc Sci Med 63: 2533–2545.
- 7. Kim MH, Kim MK, Choi BY, Shin YJ (2005) Educational disparities in the metabolic syndrome in a rapidly changing society–the case of South Korea. Int J Epidemiol 34: 1266–1273.
- 8. Marmot M (1992) Coronary heart disease: rise and fall of a modern epidemic. In: Marmot MG, Elliott P, editors. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. Oxford: OUP. pp. 3–19.
- 9. Wong SL, Donnan SP (1992) Influence of socioeconomic status on cardiovascular diseases in Hong Kong. J Epidemiol Community Health 46: 148–150.
- 10. Yu Z, Nissinen A, Vartiainen E, Hu G, Tian H, et al. (2002) Socio-economic status and serum lipids: a cross-sectional study in a Chinese urban population. J Clin Epidemiol 55: 143–149.
- 11. Maddison A (2003) Development Centre Studies: The World Economy: Historical Statistics. Paris: OECD.
- 12. Avendano M, Kunst AE, Huisman M, Lenthe FV, Bopp M, et al. (2006) Socioeconomic status and ischaemic heart disease mortality in 10 western European populations during the 1990s. Heart 92: 461–467.
- 13. Langenberg C, Kuh D, Wadsworth ME, Brunner E, Hardy R (2006) Social circumstances and education: life course origins of social inequalities in metabolic risk in a prospective national birth cohort. Am J Public Health 96: 2216–2221.
- 14. Manor O, Eisenbach Z, Friedlander Y, Kark JD (2004) Educational differentials in mortality from cardiovascular disease among men and women: the Israel Longitudinal Mortality Study. Ann Epidemiol 14: 453–460.
- 15. Jaross W, Baehrecke M, Trubsbach A, Seiler E (1981) Effects of sexual maturation on serum lipoproteins. Endokrinologie 78: 28–34.
- 16. Berenson GS, Srinivasan SR, Cresanta JL, Foster TA, Webber LS (1981) Dynamic changes of serum lipoproteins in children during adolescence and sexual maturation. Am J Epidemiol 113: 157–170.
- 17. Srinivasan SR, Sundaram GS, Williamson GD, Webber LS, Berenson GS (1985) Serum lipoproteins and endogenous sex hormones in early life: observations in children with different lipoprotein profiles. Metabolism 34: 861–867.
- 18. Kirkland RT, Keenan BS, Probstfield JL, Patsch W, Lin TL, et al. (1987) Decrease in plasma high-density lipoprotein cholesterol levels at puberty in boys with delayed adolescence. Correlation with plasma testosterone levels. JAMA 257: 502–507.
- 19. Roemmich JN, Rogol AD (1999) Hormonal changes during puberty and their relationship to fat distribution. Am J Human Biol 11: 209–224.
- 20. Stozicky F, Slaby P, Volenikova L (1991) Longitudinal study of serum cholesterol, apolipoproteins and sex hormones during puberty. Acta Paediatr Scand 80: 1139–1144.
- 21. Morrison JA, Barton BA, Biro FM, Sprecher DL (2003) Sex hormones and the changes in adolescent male lipids: longitudinal studies in a biracial cohort. J Pediatr 142: 637–642.
- 22. Arslanian S, Suprasongsin C (1997) Testosterone treatment in adolescents with delayed puberty: changes in body composition, protein, fat, and glucose metabolism. J Clin Endocrinol Metab 82: 3213–3220.
- 23. Wickman S, Saukkonen T, Dunkel L (2002) The role of sex steroids in the regulation of insulin sensitivity and serum lipid concentrations during male puberty: a prospective study with a P450-aromatase inhibitor. Eur J Endocrinol 146: 339–346.
- 24. Hero M, Ankarberg-Lindgren C, Taskinen MR, Dunkel L (2006) Blockade of oestrogen biosynthesis in peripubertal boys: effects on lipid metabolism, insulin sensitivity, and body composition. Eur J Endocrinol 155: 453–460.
- 25. Moran A, Jacobs DR Jr, Steinberger J, Cohen P, Hong CP, et al. (2002) Association between the insulin resistance of puberty and the insulin-like growth factor-I/growth hormone axis. J Clin Endocrinol Metab 87: 4817–4820.
- 26. Goran MI, Gower BA (2001) Longitudinal study on pubertal insulin resistance. Diabetes 50: 2444–2450.
- 27. Saad RJ, Keenan BS, Danadian K, Lewy VD, Arslanian SA (2001) Dihydrotestosterone treatment in adolescents with delayed puberty: does it explain insulin resistance of puberty? J Clin Endocrinol Metab 86: 4881–4886.
- 28. Gale EA, Gillespie KM (2001) Diabetes and gender. Diabetologia 44: 3–15.
- 29. Tsang S (2004) A Modern History of Hong Kong. Hong Kong: Hong Kong University Press.
- 30. Hong Kong Government (1955) Hong Kong Annual Report-1954. Hong Kong.
- 31. Dahlman CJ, Aubert J-E (2001) China and the knowledge economy. seizing the 21st century. Washington, D.C.: World Bank.
- 32. Ngo TW (2003) Industrial History and the Artiface of Laissez-faire Colonialism. In: Faure D, editor. Hong Kong: A Reader in Social History. Oxford: Oxford University Press. pp. 543–571.
- 33. Butters HR (1939) Report on Labour and Labour Conditions in Hong Kong. Hong Kong.
- 34. Schooling M, Leung GM, Janus ED, Ho SY, Hedley AJ, et al. (2004) Childhood migration and cardiovascular risk. Int J Epidemiol 33: 1219–1226.
- 35. Lam TH, Liu LJ, Janus ED, Bourke C, Hedley AJ (1999) The relationship between fibrinogen and other coronary heart disease risk factors in a Chinese population. Atherosclerosis 143: 405–413.
- 36. WHO MONICA Project (1988) Geographical variation in the major risk factors of coronary heart disease in men and women aged 35–64 years. WHO Statistics Q 41: 115–137.
- 37. Janghorbani M, Ho SY, Lam TH, Janus ED (2003) Prevalence and correlates of alcohol use: a population-based study in Hong Kong. Addiction 98: 215–224.
- 38. Lam TH, Liu LJ, Janus ED, Bourke C, Hedley AJ (1999) The relationship between fibrinogen and other coronary heart disease risk factors in a Chinese population. Atherosclerosis 143: 405–413.
- 39. Lam TH, Liu LJ, Janus ED, Bourke C, Hedley AJ (1999) The relationship between fibrinogen and other coronary heart disease risk factors in a Chinese population. Atherosclerosis 143: 405–413.
- 40. Gluckman PD, Hanson MA, Beedle AS (2007) Early life events and their consequences for later disease: a life history and evolutionary perspective. Am J Hum Biol 19: 1–19.
- 41. Jaross W, Baehrecke M, Trubsbach A, Seiler E (1981) Effects of sexual maturation on serum lipoproteins. Endokrinologie 78: 28–34.
- 42. Stozicky F, Slaby P, Volenikova L (1991) Longitudinal study of serum cholesterol, apolipoproteins and sex hormones during puberty. Acta Paediatr Scand 80: 1139–1144.
- 43. Morrison JA, Barton BA, Biro FM, Sprecher DL (2003) Sex hormones and the changes in adolescent male lipids: longitudinal studies in a biracial cohort. J Pediatr 142: 637–642.
- 44. Ballantyne C, Arroll B, Shepherd J (2005) Lipids and CVD management: towards a global consensus. Eur Heart J 26: 2224–2231.
- 45. Yusuf S, Hawken S, Ounpuu S, Bautista L, Franzosi MG, et al. (2005) Obesity and the risk of myocardial infarction in 27,000 participants from 52 countries: a case-control study. Lancet 366: 1640–1649.
- 46. Walldius G, Jungner I, Holme I, Aastveit AH, Kolar W, et al. (2001) High apolipoprotein B, low apolipoprotein A-I, and improvement in the prediction of fatal myocardial infarction (AMORIS study): a prospective study. Lancet 358: 2026–2033.
- 47. Pischon T, Girman CJ, Sacks FM, Rifai N, Stampfer MJ, et al. (2005) Non-high-density lipoprotein cholesterol and apolipoprotein B in the prediction of coronary heart disease in men. Circulation 112: 3375–3383.
- 48. Arslanian S, Suprasongsin C (1997) Testosterone treatment in adolescents with delayed puberty: changes in body composition, protein, fat, and glucose metabolism. J Clin Endocrinol Metab 82: 3213–3220.
- 49. Saad RJ, Keenan BS, Danadian K, Lewy VD, Arslanian SA (2001) Dihydrotestosterone treatment in adolescents with delayed puberty: does it explain insulin resistance of puberty? J Clin Endocrinol Metab 86: 4881–4886.
- 50. Wickman S, Saukkonen T, Dunkel L (2002) The role of sex steroids in the regulation of insulin sensitivity and serum lipid concentrations during male puberty: a prospective study with a P450-aromatase inhibitor. Eur J Endocrinol 146: 339–346.
- 51. Hero M, Ankarberg-Lindgren C, Taskinen MR, Dunkel L (2006) Blockade of oestrogen biosynthesis in peripubertal boys: effects on lipid metabolism, insulin sensitivity, and body composition. Eur J Endocrinol 155: 453–460.
- 52. Moran A, Jacobs DR Jr, Steinberger J, Cohen P, Hong CP, et al. (2002) Association between the insulin resistance of puberty and the insulin-like growth factor-I/growth hormone axis. J Clin Endocrinol Metab 87: 4817–4820.
- 53. Goran MI, Gower BA (2001) Longitudinal study on pubertal insulin resistance. Diabetes 50: 2444–2450.
- 54. Freire RD, Cardoso MA, Shinzato AR, Ferreira SR (2003) Nutritional status of Japanese-Brazilian subjects: comparison across gender and generation. Br J Nutr 89: 705–713.
- 55. Seidell JC (2005) Epidemiology of obesity. Semin Vasc Med 5: 3–14.
- 56. Galobardes B, Smith GD, Lynch JW (2006) Systematic review of the influence of childhood socioeconomic circumstances on risk for cardiovascular disease in adulthood. Ann Epidemiol 16: 91–104.
- 57. Pollitt RA, Rose KM, Kaufman JS (2005) Evaluating the evidence for models of life course socioeconomic factors and cardiovascular outcomes: a systematic review. BMC Public Health Jan 20;5: 7.
- 58. Donnan SP, Ho SC, Woo J, Wong SL, Woo KS, et al. (1994) Risk factors for acute myocardial infarction in a southern Chinese population. Ann Epidemiol 4: 46–58.
- 59. Donnan SP, Ho SC, Woo J, Wong SL, Woo KS, et al. (1994) Risk factors for acute myocardial infarction in a southern Chinese population. Ann Epidemiol 4: 46–58.
- 60. Gluckman PD, Hanson MA, Beedle AS (2007) Early life events and their consequences for later disease: a life history and evolutionary perspective. Am J Hum Biol 19: 1–19.
- 61. Singhal A, Lucas A (2004) Early origins of cardiovascular disease: is there a unifying hypothesis? Lancet 363: 1642–1645.
- 62. Adam CL, Findlay PA (1997) Effect of nutrition on testicular growth and plasma concentrations of gonadotrophins, testosterone and insulin-like growth factor I (IGF-I) in pubertal male Soay sheep. J Reprod Fertil 111: 121–125.
- 63. Nolan CJ, Neuendorff DA, Godfrey RW, Harms PG, Welsh TH Jr, et al. (1990) Influence of dietary energy intake on prepubertal development of Brahman bulls. J Anim Sci 68: 1087–1096.
- 64. Slob AK, Vreeburg JT, Van der Werff ten Bosch JJ (1979) Body growth, puberty and undernutrition in the male guinea-pig. Br J Nutr 41: 231–237.
- 65. Ronnekleiv OK, Ojeda SR, McCann SM (1978) Undernutrition, puberty and the development of estrogen positive feedback in the female rat. Biol Reprod 19: 414–424.
- 66. Dorgan JF, Hunsberger SA, McMahon RP, Kwiterovich PO Jr, Lauer RM, et al. (2003) Diet and sex hormones in girls: findings from a randomized controlled clinical trial. J Natl Cancer Inst 95: 132–141.
- 67. Dorgan JF, McMahon RP, Friedman LA, Van Horn L, Snetselaar LG, et al. (2006) Diet and sex hormones in boys: findings from the dietary intervention study in children. J Clin Endocrinol Metab 91: 3992–3996.
- 68. Campbell BC, Gillett-Netting R, Meloy M (2004) Timing of reproductive maturation in rural versus urban Tonga boys, Zambia. Ann Hum Biol 31: 213–227.
- 69. Arslanian S, Suprasongsin C (1997) Testosterone treatment in adolescents with delayed puberty: changes in body composition, protein, fat, and glucose metabolism. J Clin Endocrinol Metab 82: 3213–3220.
- 70. Arslanian S, Suprasongsin C (1997) Testosterone treatment in adolescents with delayed puberty: changes in body composition, protein, fat, and glucose metabolism. J Clin Endocrinol Metab 82: 3213–3220.
- 71. Saad RJ, Keenan BS, Danadian K, Lewy VD, Arslanian SA (2001) Dihydrotestosterone treatment in adolescents with delayed puberty: does it explain insulin resistance of puberty? J Clin Endocrinol Metab 86: 4881–4886.
- 72. Wickman S, Saukkonen T, Dunkel L (2002) The role of sex steroids in the regulation of insulin sensitivity and serum lipid concentrations during male puberty: a prospective study with a P450-aromatase inhibitor. Eur J Endocrinol 146: 339–346.
- 73. Hero M, Ankarberg-Lindgren C, Taskinen MR, Dunkel L (2006) Blockade of oestrogen biosynthesis in peripubertal boys: effects on lipid metabolism, insulin sensitivity, and body composition. Eur J Endocrinol 155: 453–460.
- 74. Moran A, Jacobs DR Jr, Steinberger J, Cohen P, Hong CP, et al. (2002) Association between the insulin resistance of puberty and the insulin-like growth factor-I/growth hormone axis. J Clin Endocrinol Metab 87: 4817–4820.
- 75. Goran MI, Gower BA (2001) Longitudinal study on pubertal insulin resistance. Diabetes 50: 2444–2450.
- 76. Ovesen L (2006) Adolescence: a critical period for long-term tracking of risk for coronary heart disease? Ann Nutr Metab 50: 317–324.
- 77. Hauspie RC, Vercauteren M, Susanne C (1997) Secular changes in growth and maturation: an update. Acta Paediatr Suppl 423: 20–27.