Birthweight, childhood growth and hypertension in adulthood

Mingfang Zhaoa, Xiao Ou Shua, Fan Jinb, Gong Yanga, Hong-Lan Lib, Da-Ke Liub, Wanqing Wena, Yu-Tang Gaob and Wei Zhenga

a Center for Health Services Research and Vanderbilt-Ingram Cancer Center, Vanderbilt University, Nashville, TN 37232–8300, USA.
b Department of Epidemiology, Shanghai Cancer Institute, Shanghai 200032, PR China.

Xiao Ou Shu, MD, PhD, Vanderbilt University Medical Center, Health Services Research, 6th floor, Room 6009, Medical Center East, Nashville, TN 37232–8300, USA. E-mail: xiao-ou.shu{at}mcmail.vanderbilt.edu

Abstract

Background Low birthweight (BW) and childhood growth have been hypothesized to be associated with an increased risk of hypertension in later life.

Methods We analysed data among 13 467 women with a recalled BW from the Shanghai Women’s Health Study. Cases included those with a self-reported hypertension with (‘confirmed cases’) or without (‘possible cases’) antihypertensive medication(s) use. Logistic regression was used to derive adjusted odds ratios (OR) and 95% CI.

Results Birthweight was inversely associated with the odds of early onset (at age 20–40 years) hypertension in a dose response manner (P for trend = 0.01). This association is stronger for ‘confirmed’ hypertension (only OR for ‘confirmed’ hypertension are referred to subsequently). Being heavier or taller than average at 15 years of age were both related to elevated odds of early onset hypertension. Women who had a low BW but were heavier than average at age 15 were more than four times (OR = 4.63, 95% CI: 2.40–8.94) more likely to have an early onset hypertension, and those who had a low BW and became taller at 15 years of age had an OR of 1.87 (95% CI: 1.05–3.31). A significant interaction between BW and weight at age 15 was observed (P = 0.04).

Conclusion Our study suggests that low BW, particularly if accompanied by accelerated childhood growth, may increase the risk of early onset hypertension in adulthood.

Keywords Birthweight, childhood growth, hypertension, China, developing country

Accepted 23 May 2002

Hypertension, one of the most important risk factors for cardiovascular diseases (CVD),1 is the second leading cause of death in China.2 In China, people with hypertension were five times more likely to develop stroke than those with normal blood pressure.3 In Western populations, 30% of total mortality attributable to CVD could be prevented if blood pressure could be reduced by 10 mmHg.4 Although extensively studied, the aetiology of hypertension cannot be fully explained by genetic factors and adulthood risk factors such as age, body mass index (BMI), physical activity, and cigarette smoking.3,5 It has been suggested that factors related to the intrauterine and postnatal environments may contribute to the development of hypertension.5 Intrauterine development may ‘programme’ fetal anatomy, physiology, and metabolism and thus may affect the risk of diseases including hypertension in later life.6–8 This hypothesis has been evaluated in animal models and low birthweight (BW) or intrauterine nutrition deficiency was found to be associated with high blood pressure and/or resulted in adult hypertension.9–11

Although the association between BW and blood pressure measurements in childhood or adulthood has been extensively evaluated,12–16 a limited number of epidemiological studies have examined the association of BW with hypertension in adulthood and the results are not entirely consistent.17–20 Analytical studies on childhood weight or height and adult blood pressure or hypertension are few.18,20 A recent study suggested that accelerated growth between ages 7 and 15 years increases the risk of hypertension in later life.20 The combined effect of BW and childhood growth on hypertension risk in adults is unclear. Using data from the baseline survey of the Shanghai Women’s Health Study (SWHS), we examined the effect of BW and childhood growth on the risk of adult onset hypertension.

Methods

Shanghai Women’s Health Study
The SWHS is a population-based prospective cohort study of all women who were aged 40–70 years and lived in the seven study communities in Shanghai, China, during March 1997 to May 2000. Of 81 271 women approached for the study, in-person interviews were completed with 75 322 women, a participation rate of 92.7%. After excluding women (n = 273) who were <40 or >70 years at the time of interview, 75 049 remaining women constituted the SWHS cohort. Among the non-participants (n = 5949), 2407 (3.0%) refused to participate, 2073 (2.6%) were absent during the recruitment period, and 1469 (1.8%) were excluded for miscellaneous reasons, including mental disorders and communication problems.

The baseline survey was accomplished through a self-administered questionnaire and an interviewer-administered questionnaire. The self-administered questionnaire, delivered in person, was used to obtain information related to demographic characteristics including BW, medical and menstrual histories, lifestyle factors, and lifetime occupational history. Birthweight was recalled in kilograms. Study participants were asked whether they had been diagnosed by a physician with hypertension and other selected chronic diseases, including diabetes, coronary heart disease, and stroke. For each disease reported, information on age at first diagnosis was also collected. A face-to-face home interview, scheduled at first home visit, was conducted to obtain further information on usual dietary habits, medication(s) use, reproductive history, physical activities, physical development and body size. During the in-person interview, the interviewer first checked with the study participant for completeness and consistency of the information provided in the self-administered questionnaire. For those women who had difficulties completing, or had not yet completed, the self-administered questionnaire, the questionnaire was administered by the interviewer. During the interview, information on using specific medication(s), including antihypertensive medication(s), was obtained. Ever use of specific medication was defined as using >=3 times per week for >2 months. Duration and frequency of medication(s) use were also inquired. Information on relative weight and height at age 15 and 20 years was collected by asking women to compare their weight and height to their peers. Women were also asked to recall their weight in kilograms and height in centimetres at age 20 years. All participants were measured for their current weight, circumferences of the waist and hip, and sitting and standing height, by trained interviewers according to a standard protocol. All anthropometric measurements were measured twice and a third one was administered if the difference between the first two measured readings exceeded the tolerance limits (1 cm for height and 1 kg for weight). The average of the two closest readings was used in current analyses.

Subjects included in the current analyses
Of 75 049 SWHS participants, 464 who were diagnosed with hypertension before age 20 years were excluded from current analyses, as it would be difficult to establish the temporal sequence of the effect of childhood growth on hypertension. Of the remainder, 13 467 provided information on BW and were included in our analyses. Prevalent cases (n = 2373) included all women with self-reported physician-diagnosed hypertension. Controls (n = 11 094) included those without a history of physician-diagnosed hypertension. Hypertension cases were further classified into two groups, ‘confirmed cases’ and ‘possible cases‘, according to the use of antihypertensive medication(s). ‘Confirmed cases’ included those cases who ever used antihypertensive medication(s) (n = 1433), one of the hypertension definitions previously used in epidemiological studies,17–20 and the ‘possible cases’ (n = 940) were those without regular antihypertensive medication(s) use. Age at first diagnosis of hypertension refers to the age when hypertension was first diagnosed. Early onset hypertension included those cases with a physician-diagnosed hypertension at ages 20–40 years; while late onset hypertension had onset at ages 41–70 years. Low BW was defined as weight <2500 g at birth. The perceived weight and height at age 15 was recorded as about average, a little heavier/ taller, a little thinner/shorter, heavier/taller and thinner/shorter than their peers. A little heavier/taller and heavier/taller were further categorized as above average; while a little thinner/ shorter and thinner/shorter were grouped as below average in weight and height at age 15 years. The BMI was calculated by dividing the subject’s weight in kilograms by the square of the subject’s height in metres (weight [kg]/height [m2]). Because Chinese women in general have a relatively low BMI and a low prevalence of obesity, all anthropomorphic measurements were categorized according to the quartile distributions of controls instead of using conventional cut-off points for obesity.

Statistical analyses
Unconditional logistic regression models were used to obtain maximum likelihood estimates of the adjusted odds ratios (OR) and their 95% CI.21 Multinomial logistic regression models were applied where the outcome variable (hypertension) has more than two categories (i.e. ‘confirmed cases’ and ‘possible cases‘). Tests for interactions utilized the likelihood ratio test by adding the interaction term(s) (of two or more variables of interest) to the model that already included the main effect variables and confounders. Tests for trend across levels of the exposure variables were performed by treating ordinal-score variables as continuous in logistic regression models. One-way ANOVA with Tukey multiple comparisons was used to test the differences in means among categories. Based on a prior hypothesis that the effect of early life exposures is more likely to manifest in young adulthood, we analysed data for all subjects combined and by age of hypertension onset. All P-values were two-sided. (Analyses were also carried out for all cases combined first and subsequently focussed on ‘confirmed cases‘.)

Results

Table 1Go presents the prevalence of hypertension by demographic factors and selected risk factors among all participants and separately among women who provided BW information. The age range of 75 049 women was 40–70 years (mean: 52.1 years). Among the whole cohort, 23.8% of women reported having been diagnosed with hypertension. The prevalence of hypertension increases steadily with age and BMI levels, and tended to be higher among women who were less educated. Among women with recalled BW, a similar pattern was observed, with the exception that a higher prevalence was observed among postmenopausal than premenopausal women. The average age for women with a recalled BW was 48.1 years, 4 years younger than that in the whole cohort. As current BMI may be correlated with weight and height at age 15 years and may be in the causal pathway between childhood growth and subsequent hypertension, we did not include current BMI for adjustment. All subsequent analyses were adjusted for age, age2, and education levels, and were performed only among those women with BW data.


View this table:
[in this window]
[in a new window]
 
Table 1 Associations of body mass index (BMI) and selected demographic factors with a history of hypertension in the Shanghai Women’s Health Study
 
The associations of BW and physical development in early life with a history of hypertension among those with BW data are presented in Table 2Go. Overall, BW was inversely related to the prevalence of hypertension in a dose-response manner (P = 0.07). Women with a history of hypertension were more likely to be heavier (OR: 1.26 and 1.43; 95% CI: 1.07–1.48 and 1.11–1.83) or thinner (OR = 1.21; 95% CI: 1.07–1.36) than those with average weight and were more likely to be taller (OR = 1.14; 95% CI: 0.99–1.30) than those with average height at age 15. The relative and estimated weight, height and BMI at age 20 were not strong predictors of hypertension among all women. Further analyses with cases stratified according to antihypertensive medication(s) use status showed that the associations of BW and weight and height at age 15 years were more pronounced for ‘confirmed cases’ (right two columns in Table 2Go), than for ‘possible cases’ (data not shown). A similar pattern was observed for weight, height, and BMI at age 20. Subsequent analyses were restricted to ‘confirmed cases’.


View this table:
[in this window]
[in a new window]
 
Table 2 Birthweight, weight and height at early life and the history of hypertension in adult life among the birthweight sub-cohort
 
Table 3Go presents the results of analyses stratified by age at first diagnosis of hypertension. The significant associations of BW and weight and height at age 15 with hypertension were confined to those with early onset. No significant associations of perceived weight and height and BMI at age 20 with the odds of early onset hypertension were observed. Higher BMI at 20 was, however, associated with increased risk of late onset hypertension.


View this table:
[in this window]
[in a new window]
 
Table 3 Birthweight, weight and height at early life and the history of hypertension in adult life stratified by the age of hypertension onset
 
The joint effects of BW and weight or height at age 15 are evaluated in relation to the prevalence of early onset hypertension (Table 4Go). Compared to women with average weight at age 15 and BW of 2500–3249 g, women who were above average weight at age 15 and had a low BW had more than fourfold elevated odds of hypertension onset at age 21–40 years. A statistically significant interaction between BW and weight at age 15 years was observed (PBW & weight = 0.04). Compared to women with average height at age 15 and BW of 2500–3249 g, women who were above average in height at age 15 and had a low BW had 87% higher risk of having hypertension diagnosed during ages 21–40. The interaction between BW and relative height at age 15 was not statistically significant (P = 0.42).


View this table:
[in this window]
[in a new window]
 
Table 4 Combined effects of birthweight and adolescent weight or height in relation to the prevalence of early onset hypertension
 
Table 5Go presents the joint effects of both weight and height at age 15 and BW in relation to the prevalence of early onset hypertension. The reference group included women with average weight and height at age 15 and BW of 2500–3249 g. Women who were above average weight and below average height at age 15 and had a low BW had the highest odds of having early onset hypertension (OR = 7.64, 95% CI: 2.85–20.44). Those who were above average for both weight and height at age 15 and had a low BW had more than fivefold increased odds of early onset hypertension (OR = 5.47; 95% CI: 1.74–17.26).


View this table:
[in this window]
[in a new window]
 
Table 5 Joint effects of birthweight and weight and height at age 15 years on the early onset hypertension
 
Discussion

In this population-based study, we found that BW was inversely associated with the prevalence of early onset hypertension in a dose-response manner. Women who were heavier or taller during childhood were at an elevated risk of developing early onset hypertension. The highest risk was found among women who were born with a low BW and were heavier but shorter than their peers during childhood. Our analyses found an interaction between BW and weight at age 15 years with respect to prevalence of hypertension.

An inverse association between BW and blood pressure has been consistently observed in many studies conducted in different countries, ethnic groups, age groups, and sexes.15,16 Of the 80 studies reviewed,16 only one study was conducted in China. In a Chinese study of 309 men and 318 women with hospital birth records, BW was found to be negatively related to systolic and diastolic blood pressure in adulthood,22 suggesting that higher BW may be associated with a lower risk of hypertension in Chinese adults. Epidemiological studies on the association between BW and hypertension prevalence or incidence are few.17–20 A negative association between BW and hypertension prevalence has been reported, with a dose-response relation, among men17,20 and women,20 but not among women in another study.18 In one study, the inverse association was only seen among older women,19 consistent with the conclusion of Law et al.’s review article, that the inverse associations between BW and blood pressure (or hypertension) became stronger with advanced age.14 In our study, we found a linear relationship between BW and the prevalence of early onset hypertension, but not late onset hypertension. Similar to our study, the definitions of hypertension among all of those studies17,18,20 except one19 were self-reported physician-diagnosed hypertension17,18 or ever use of antihypertensive medication.20 Only one study used ever use of antihypertensive medication or having elevated blood pressure measurements as the definition of the disease.19

Only one study has reported the association between growth in childhood and hypertension in adulthood.20 In a cohort study of 7086 subjects conducted in Finland, the association of BW, childhood growth from 7 to 15 years old and hypertension in adulthood was assessed in a subgroup of 1958 subjects.20 These researchers collected data from birth records and school health records, and identified the prevalent hypertension cases through personal identifier linked to the nationwide database of the Social Insurance Institution’s Register. In addition to the inverse association between BW and blood pressure, Eriksson et al. observed that weight and height at age 7–15 years were positively associated with blood pressure, and were independent predictors of hypertension in adulthood. Based on their observation, they concluded that hypertension developed through fetal growth retardation and ‘catch-up’ growth in early childhood. Our study suggested a similar pattern. Women who were born with a low BW, and became heavy or both heavy and tall at age 15, had a 5- to 7-fold increased odds of developing early onset hypertension.

In our study, we found that the inverse association between BW and the prevalence of hypertension was more evident among those with early onset hypertension but not late onset hypertension. The mean age at study enrolment is greater among the women with late onset hypertension (mean age 59 years) than those with early onset hypertension (mean age 54 years). Because low BW was also associated with increased mortality from CVD and/or other causes,23,24 it is possible that some women born with low BW may not survive long enough to enter our study. Therefore, the association between BW and the prevalence of hypertension, particularly among those with late onset hypertension, may be attenuated or even diminished due to the poor survival among those with a low BW.

It is biologically plausible that BW and childhood growth may jointly affect the risk of hypertension. In rabbits, a significantly reduced glomeruli number was observed in rabbits with a low BW.25 In Brenner and Chertow’s review article, they suggested that BW is an indicator of nephron number and that nephron number is inversely associated with blood pressure and positively associated with BW.26 Therefore, it is possible that accelerated growth during childhood imposes a greater demand on the reduced number of renal cells among women with a low BW, and thus induces progressive kidney damage and subsequently hypertension. It is also possible that the observed positive association is due to an unadjusted confounding effect from maternal or genetic factors that are related to both BW and/or childhood growth and risk of development of hypertension, or a chance finding given the multiple comparisons made and the small sample size in some of the analyses. Nevertheless, the biological plausibility of the hypothesis on BW, accelerated childhood growth and hypertension risk means that more studies on this topic are warranted.

The major strength of our study is the large sample size, population-based study design and the very high response rate. The major limitation of this study is that both outcome and exposures of interest are self-reported. Cases and controls of the study were defined based on self-reported history of hypertension at the baseline survey of the SWHS. Using medical chart information,27 physician’s claims28 or the criteria proposed by the Joint National Committee on Detection, Evaluation, and Treatment of High Blood Pressure (JNC III)29 as the gold standard, the validity of self-reported hypertension was reported to be low in telephone surveys,27 and fair to high in face-to-face home interview.28,29 Using a self-administered questionnaire, similar to ours, the validity of self-reported physician-diagnosed hypertension was found to be very high30 and fairly high even among those without antihypertensive medication(s) use.31 The home-completed questionnaire may allow the respondents more time for recall, to discuss it with other family members, and to consult their physicians for more accurate answers. Furthermore, the addition of ‘physician-diagnosed’ to question wording may have partially eliminated false positives, a phenomenon observed in Klungel et al.’s study.31 We compared the prevalence of hypertension in our study with that reported from the 1997 China Health and Nutrition Survey (CHNS), a study with hypertension defined by JNC III and conducted in cities adjacent to Shanghai among Chinese women with a similar age range of participants to SWHS.32 We found that the prevalence of hypertension in our study was similar to that in CHNS across BMI ranges (Figure 1Go), suggesting that self-reported hypertension history in our study is relatively accurate. However, misclassification, particularly misclassifying undiagnosed cases as controls, is likely. Nevertheless, neither study participants nor interviewers were aware of our study hypothesis. Such misclassification, therefore, is likely to bias the results towards null.



View larger version (15K):
[in this window]
[in a new window]
 
Figure 1 Comparison of prevalence of hypertension among Chinese women from the China Health and Nutrition Survey (CHNS) and the Shanghai Women’s Health Study

 
Information on BW was only available for about 18% of study participants, raising the question of potential selection bias. However, we did not find that women with recalled information on BW differ within the whole cohort of women on demographic and known hypertension risk factors. Digit preference of recalling BW to the nearest 50 or 100 g was evident, but it did not differ by hypertension status, indicating that this bias was likely to be non-differential. Validation studies have suggested a high correlation between self-reported BW data and those from hospital records.33,34 Recalled BW has been suggested to have a higher validity among younger women than older women.35 This may be one of the explanations for why a positive association of BW and hypertension was observed among young women, but not among older women, in our study. Relatively high accuracy of recalled adolescent weight (r = 0.84) and height (r = 0.92) with overestimates among lean girls and underestimates among obese girls, was observed in Must et al.’s study,36 indicating that recalled past weight and height is a valuable surrogate for the true measurements. Because misclassification error is generally less for categorized data than continuous data, applying the relative weight and height measurements in our study may have lessened the degree of misclassification. In our study, women were asked to recall their weight and height at age 20 years both in measured (kilogram and metre) and relative scales. One-way ANOVA with Tukey multiple comparison tests were conducted to test the differential categorization of relative terms with respect to the means of recalled anthropometric measurements (Table 6Go). Results of the analyses revealed expected gradients in height, weight and BMI by relative weight and height categories.


View this table:
[in this window]
[in a new window]
 
Table 6 Descriptive statistics for the recalled weight and height in digits by their relative terms at age 20 years in Shanghai Women’s Health Study
 
In summary, we found that growth in utero and childhood may influence the risk of development of hypertension in adulthood. Low BW and accelerated growth in childhood may work synergistically to increase the risk of early onset of hypertension. Our findings suggest that prevention of hypertension should be initiated as soon as life begins.


KEY MESSAGES

  • Birthweight was inversely associated with the prevalence of early onset hypertension among Chinese women in a dose-response manner.
  • Being above average in weight and/or height at age 15 years was related with an elevated presence of early onset hypertension among Chinese women.
  • Low birthweight and accelerated growth in childhood may work synergistically to increase the risk of early onset hypertension.

 

Acknowledgments

This work was supported by NCI grant R01CA70867. The authors want to thank Dr Wong-Ho Chow for the helpful comments.

References

1 Wright HT Jr, Hammonds VC. Hypertension: Epidemiology and contemporary management strategies. In: Wong ND, Black HR, Gardin JM (eds). Preventive Cardiology. Maidenhead: McGraw-Hill Companies, Inc., 2000, p. 133.

2 Chen XS, Ge KY. Nutrition transition in China: the growth of affluent diseases with the alleviation of undernutrition. Asia Pacific J Clin Nutr 1995;4:287–93.

3 Ueshima H, Zhang XH, Choudhury SR. Epidemiology of hypertension in China and Japan. J Hum Hypertens 2000;14:765–69.[CrossRef][ISI][Medline]

4 Rose G. Sick individuals and sick populations. Int J Epidemiol 1985; 14:32–38.[Abstract]

5 Barker DJP. Rise and fall of Western diseases. Nature 1989;338:371–72.[CrossRef][ISI][Medline]

6 Godfrey KM, Barker DJP. Fetal nutrition and adult disease. Am J Clin Nutr 2000;71(Suppl.):1344S–52S.[Abstract/Free Full Text]

7 Barker DJP. Maternal nutrition, fetal nutrition, and disease in later life. Nutrition 1997;13:807–13.[CrossRef][ISI][Medline]

8 Byrne CD, Phillips DI. Fetal origins of adult disease: epidemiology and mechanism. J Clin Pathol 2000;53:822–28.[Abstract/Free Full Text]

9 Persson E, Jansson T. Low birth weight is associated with elevated adult blood pressure in the chronically catheterized guinea-pig. Acta Physiol Scand 1992;145:195–96.[ISI][Medline]

10 Langley-Evans SC, Phillips GJ, Gardner DS, Jackson AA. Role of glucocorticoids in programming of maternal diet-induced hypertension in the rat. J Nutr Biochem 1996;7:173–78.[CrossRef][ISI]

11 Manning J, Vehaskari VM. Low birth weight-associated adult hypertension in the rat. Pediatr Nephrol 2001;16:417–22.[CrossRef][ISI][Medline]

12 Law CM, Shiell AW, Newsome CA et al. Fetal, infant, and childhood growth and adult blood pressure. Circulation 2002;105:1088–92.[Abstract/Free Full Text]

13 Miura K, Nakagawa H, Tabata M, Morikawa Y, Nishijo M, Kagamimori S. Birth weight, childhood growth, and cardiovascular disease risk factors in Japanese aged 20 years. Am J Epidemiol 2001; 153:783–89.[Abstract/Free Full Text]

14 Law CM, de Swiet M, Osmond C et al. Initiation of hypertension in utero and its amplification throughout life. BMJ 1993;306:24–27.[ISI][Medline]

15 Law CM, Sheill AW. Is blood pressure inversely related to birth weight? The strength of evidence from a systematic review of the literature. J Hypertens 1996;14:935–41.[ISI][Medline]

16 Huxley RR, Shiell AW, Law CM. The role of size at birth and postnatal catch-up growth in determining systolic blood pressure: a systematic review of the literature. J Hypertens 2000;18:815–31.[CrossRef][ISI][Medline]

17 Curhan GC, Willett WC, Rimm EB et al. Birth weight and adult hypertension, diatetes mellitus, and obesity in US men. Circulation 1996;94:3246–50.[Abstract/Free Full Text]

18 Curhan GC, Chertow GM, Willett WC et al. Birth weight and adult hypertension and obesity in women. Circulation 1996;94:1310–15.[Abstract/Free Full Text]

19 Andersson SW, Lapidus L, Niklasson A, Hallberg L, Bengtsson C, Hulthen L. Blood pressure and hypertension in middle-aged women in relation to weight and length at birth: a follow-up study. J Hypertens 2000;18:1753–61.[CrossRef][ISI][Medline]

20 Eriksson J, Forsén T, Tuomilehto J, Osmond C, Barker D. Fetal and childhood growth and hypertension in adult life. Hypertension 2000;36:790–94.[Abstract/Free Full Text]

21 Breslow NE, Day NE (eds). Statistical Methods in Cancer Research. Vol. 1. The Analysis of Case-control Studies. Lyon: International Agency for Research on Cancer, 1980.

22 Law CM, Egger P, Dada D et al. Body size at birth and blood pressure among children in developing countries. Int J Epidemiol 2000;29:52–59.

23 Osmond C, Barker DJ, Winter PD, Fall CH, Simmonds SJ. Early growth and death from cardiovascular disease in women. BMJ 1993; 307:1519–24.[ISI][Medline]

24 Leon DA, Lithell HO, Vågerö D et al. Reduced fetal growth rate and increased risk of death from ischaemic heart disease: cohort study of 15 000 Swedish men and women born 1915–29. BMJ 1998;317:241–45.[Abstract/Free Full Text]

25 Bassan H, Trejo LL, Kariv N et al. Experimental intrauterine growth retardation alters renal development. Pediatr Nephrol 2000;15:192–95.[CrossRef][ISI][Medline]

26 Brenner BM, Chertow GM. Congenital oligonephropathy and the etiology of adult hypertension and progressive renal injury. Am J Kidney Dis 1994;23:171–75.[ISI][Medline]

27 Bowlin SJ, Morrill BD, Nafziger AN, Lewis C, Pearson TA. Reliability and changes in validity of self-reported cardiovascular disease risk factors using dual response: The Behavioral Risk Factor Survey. J Clin Epidemiol 1996;49:511–17.[CrossRef][ISI][Medline]

28 Muhajarine N, Mustard C, Roos LL, Young TK, Gelskey DE. Comparison of survey and physician claims data for detection hypertension. J Clin Epidemiol 1997;50:711–18.[CrossRef][ISI][Medline]

29 Vargas CM, Burt VL, Gillum RF, Pamuk ER. Validity of self-reported hypertension in the National Health and Nutrition Examination Survey III, 1988–1991. Prev Med 1997;26:678–85.[CrossRef][ISI][Medline]

30 Colditz GA, Martin P, Stampfer MJ et al. Validation of questionnaire information on risk-factors and disease outcomes in a prospective cohort study of women. Am J Epidemiol 1986;123:894–900.[Abstract]

31 Klungel OH, de Boer A, Peas AHP, Seidell JC, Bakker A. Cardiovascular diseases and risk factors in a population-based study in The Netherlands: agreement between questionnaire information and medical records. The Netherlands Journal of Medicine 1999;55:177–83.[CrossRef][ISI][Medline]

32 Bell AC, Adair LS, Popkin BM. Ethnic differences in the association between body mass index and hypertension. Am J Epidemiol 2002; 155:346–53.[Abstract/Free Full Text]

33 Sanderson M, Williams MA, White E et al. Validity and reliability of subject and mother reporting of perinatal factors. Am J Epidemiol 1998;147:136–40.[Abstract]

34 Troy LM, Michels KB, Hunter DJ et al. Self-reported birthweight and history of having been breastfed among younger women: an assessment of validity. Int J Epidemiol 1996;25:122–27.[Abstract]

35 Allen DS, Ellison GTH, Silva I, Stavola BLD, Fentiman IS. Determinants of the availability and accuracy of self-reported birth weight in middle-aged and elderly women. Am J Epidemiol 2002; 155:379–84.[Abstract/Free Full Text]

36 Must A, Willete WC, Dietz WH. Remote recall of childhood height, weight, and body build by elderly subjects. Am J Epidemiol 1993;138:56–64.[Abstract]