Body Mass Index and Blood Pressure in Adult Type 2 Diabetic Patients in Taiwan

Circ J 2007; 71: 1749 1754 Body Mass Index and Blood Pressure in Adult Type 2 Diabetic Patients in Taiwan Chin-Hsiao Tseng, MD, PhD*, **,,, Background The correlation between body mass index (BMI)/obesity and blood pressure/hypertension has not been confirmed in diabetic patients. This study analyzed the association in Taiwanese adult patients with type 2 diabetes mellitus (T2DM). Methods and Results The National Health Insurance data of 89,857 adult patients (41,398 men and 48,459 women; aged 18 years with mean ± SD of 62.2±11.3 years) interviewed by questionnaire were analyzed, taking into account the potential confounding of age, sex, diabetic duration, smoking and parental hypertension. Prevalence of obesity (BMI 25 kg/m 2 ) was 39.3% and 41.7% in the diabetic men and women, respectively; and hypertension was 54.5% in both sexes. Hypertension increased significantly with increasing age and BMI. The adjusted odds ratio (OR) for hypertension for every 1 kg/m 2 increment of BMI was 1.16 (1.15 1.17) and 1.13 (1.12 1.14) for men and women, respectively. The correlation between BMI and hypertension/blood pressure extended throughout BMI levels to the nonobese range. For men, adjusted OR for BMI <18.5, 18.5 22.9, 23.0 24.9, 25.0 29.9 and 30.0 kg/m 2 were 1.00, 1.23 (1.07 1.41), 1.78 (1.55 2.05), 3.06 (2.66 3.53) and 6.59 (5.56 7.82), respectively; and were 1.00, 1.09 (0.97 1.22), 1.49 (1.32 1.68), 2.32 (2.06 2.60), 4.40 (3.84 5.05), respectively, for women. For every 1 unit BMI increase in patients without a hypertension history, the systolic and diastolic blood pressures significantly increased by 0.618 and 0.447 mmhg for men; and by 0.637 and 0.462 mmhg, respectively, for women. Conclusions BMI/obesity is significantly linked to blood pressure/hypertension throughout the range of BMI in diabetic patients in either sex regardless of a previous hypertension history. (Circ J 2007; 71: 1749 1754) Key Words: Blood pressure; Body mass index; Hypertension; Obesity; Taiwan; Type 2 diabetes mellitus The Framingham Offspring study estimated that approximately 65% and 75% of the hypertension risk in women and men, respectively, could be attributed to being overweight or obese. 1 Although this link was well confirmed in the general populations of the US 2,3 and Norway 4 in more recent epidemiologic studies, the impact of obesity on morbidity and mortality may not be similar among different ethnic groups. 5 10 Furthermore, because the cutoffs for body mass index (BMI) used to define obesity or overweight are different for Asian populations, 11,12 it is necessary to examine whether the recommended cutoffs are associated with hypertension risk in Asian people. Age is an independent risk factor for hypertension, 13 but the impact of body weight on blood pressure could be different between men and women. 4 Family history and smoking are also important risk factors for hypertension, but they were not always considered simultaneously as confounders in the evaluation of the relationship between BMI and hypertension. 2 4 Because diabetic patients are characterized (Received April 19, 2007; revised manuscript received July 14, 2007; accepted July 26, 2007) *National Taiwan University College of Medicine, **Division of Endocrinology and Metabolism, Department of Internal Medicine, National Taiwan University Hospital, Taipei, Department of Medical Research and Development, National Taiwan University Hospital Yun-Lin Branch, Yun-Lin, School of Public Health, Taipei Medical University and Division of Environmental Health and Occupational Medicine of the National Health Research Institutes, Taipei, Taiwan Mailing address: Chin-Hsiao Tseng, MD, PhD, Department of Internal Medicine, National Taiwan University Hospital, No. 7 Chung-Shan South Road, Taipei, Taiwan. E-mail: ccktsh@ms6.hinet.net by various metabolic disturbances that can lead to hypertension, the relative impact of obesity on blood pressure and hypertension might level off in these patients. Therefore, the association between BMI/obesity and blood pressure/ hypertension observed in the general population should not be extrapolated to diabetic patients without evidence derived directly from this unique group of individuals. To the best of our knowledge, the prevalence of obesity in the Taiwanese diabetic patients is not well defined, and whether obesity is associated with hypertension in these patients remains to be confirmed. Therefore, this study evaluated the prevalence of obesity and its correlation with hypertension in adult Taiwanese patients with type 2 diabetes mellitus (T2DM) in the respective sexes, taking into account the potential confounding effects of age, diabetic duration, smoking and parental hypertension. Methods Study Subjects The study was approved by the Ethics Committee of the Department of Health of Taiwan. Because more than 96% (with the exemption of persons involved in military services and those subject to criminal sanction, etc) of the total population of Taiwan has been covered by the compulsory and universal National Health Insurance (NHI) since March 1995, almost all diabetic patients have been using the NHI. 14 18 Therefore, the clinical settings databases claiming for the NHI are appropriate for deriving a national sample of diabetic patients. The assembly of such a national sample has been described in detail elsewhere. 14 18 In brief, the

1750 TSENG C-H Table 1 Sex Percentage Distribution of Body Mass Index by Age and Sex Age Body mass index (kg/m 2 ) (years) <18.5 18.5 22.9 23.0 24.9 25.0 29.9 30 Both sexes <40 4.2 27.4 18.8 34.1 15.4 40 49 2.6 27.4 23.7 36.3 10.0 50 59 2.3 29.0 26.0 34.9 7.8 60 69 2.9 30.9 25.5 34.2 6.5 70 4.2 35.6 24.8 30.1 5.2 All 3.1 31.3 25.0 33.5 7.1 Women <40 6.6 29.0 16.6 31.3 16.5 40 49 2.8 26.9 22.4 35.6 12.3 50 59 2.4 29.0 24.6 34.8 9.3 60 69 3.1 30.4 24.2 34.6 7.7 70 4.5 35.7 23.2 30.0 6.5 All 3.4 31.2 23.7 33.3 8.4 Men <40 2.5 26.3 20.3 36.1 14.7 40 49 2.4 27.8 24.8 36.8 8.2 50 59 2.3 29.0 27.7 35.0 6.0 60 69 2.5 31.5 27.2 33.8 5.0 70 3.9 35.4 26.8 30.2 3.6 All 2.8 31.3 26.6 33.6 5.7 P values for both sexes, men and women are all <0.001 by chi-square test. total number of diabetic patients identified in the NHI database was 536,159, from which a total of 256,036 were identified from 66 hospitals and clinics located evenly throughout Taiwan. To create a cohort of 90,000 patients (approximately one-sixth of the total number of diabetic patients in Taiwan during the period) for long-term followup, 128,572 cases were randomly selected, supposing a response rate of 70%. Telephone Interview From March 1995 to April 30, 2002, well-trained interviewers used a structured questionnaire for telephone interview. Researchers tried up to 3 times to reach subjects before giving up. In order to obtain full information from the patients and keep missing data as minimal as possible, the interviewers handed in the interviewed questionnaires every week and all returned questionnaires were checked by an assistant and then double-checked by the investigator. The information abstracted from the questionnaire for this study included the onset age of diabetes, onset symptoms and treatment modality for the distinction between type 1 diabetes mellitus (T1DM) and T2DM, body height, body weight, personal history of smoking and hypertension and parental hypertension. The ordinary blood pressure of the patients was also requested. Hypertension was defined by self-reported history and also by reported systolic blood pressure (SBP) 140 mmhg and/or diastolic blood pressure (DBP) 90 mmhg. Disease duration was calculated as the age at the time of interview minus the onset age of diabetes. BMI was calculated from the body weight in kg divided by the squared body height in meters. The classification of T1DM was based on either of the following criteria: (1) diabetic ketoacidosis at the onset of diabetes or (2) the patients required insulin injections within 1 year of diagnosis of diabetes. If a patient was not diagnosed as T1DM, he/she was viewed as a patient with T2DM. Patients identified as T1DM or aged <18 years were not included into the present study. A total of 93,484 (response rate: 72.7%) patients completed the interview. After excluding 3,528 with T1DM, there were 89,956 patients identified as T2DM. Exclusion of patients aged <18 years yielded 89,857 cases (41,398 men and 48,459 women). Statistical Analyses BMI was divided into underweight, normal, overweight, and obesity I and II with the following respective cutoffs recommended for Asian populations: <18.5, 18.5 22.9, 23.0 24.9, 25.0 29.9 and 30.0 kg/m 2. 11 Age was divided into the following subgroups: <40, 40 49, 50 59, 60 69 and 70 years. While evaluating the association between BMI/obesity and blood pressure/hypertension, age, sex, diabetic duration, smoking and parental hypertension were treated as potential confounders. A p-value <0.05 was considered as statistically significant. The percentage distribution of the subgroups of BMI by the subgroups of age in both sexes together and in either sex were calculated and tested by chi-square test. The prevalence rates of hypertension with regards to increasing age in each subgroup of BMI, and with regards to increasing BMI in each subgroup of age and in all ages, were test by linear test for trend in both sexes together and in each sex. Logistic regression was used to estimate the odds ratios (OR) and their 95% confidence intervals (CI) for hypertension where BMI was treated as independent variable in either continuous or categorical format. Multiple linear regression models were created to estimate SBP and DBP by BMI. These logistic and multiple linear regression models were created for both sexes together and for separate sexes with and without adjustment for potential confounders. To avoid the potential impact of the use of antihypertensive agents, multiple linear regression models were also created separately for patients with and without a history of hypertension. The SBP and DBP for the 10 th to the 90 th percentiles of BMI after adjustment for age, sex, diabetic duration, smoking and parental hypertension were calculated for patients with and without hypertension history. Results The subjects had a mean (SD) age, diabetic duration and BMI of 62.2 (11.3) years, 7.2 (6.5) years and 24.6 (3.6) kg/m 2, respectively; and the respective prevalence of smoking, hypertension and parental hypertension was 30.6%, 54.5% and 18.9%. Table 1 shows the percentage distribution of BMI by age

Obesity and HT in Taiwanese T2DM 1751 Fig 1. Prevalence of hypertension by body mass index. (A) Comparison of men (closed triangle) and women (closed square) for all ages. (B D) Comparison of the age groups of <40 (closed diamond), 40 49 (closed square), 50 59 (closed triangle), 60 69 (cross) and 70 years (star) for both sexes together (B), for women (C) and for men (D). The figures on the right-hand side are the prevalence ratios of hypertension between the groups with body mass index 30 kg/m 2 and <18.5 kg/m 2. Body mass index was classified according to the recommended cutoffs for Asian populations. 11 A p value for trend test for comparison in each curve is <0.001. Table 2 Odds Ratios for Hypertension for Body Mass Index as Continuous or Categorical Variables by Sex Interpretation Men Women Both sexes Continuous Unadjusted Every 1 kg/m 2 increment 1.12 (1.11 1.13)* 1.11 (1.10 1.12)* 1.11 (1.10 1.12)* Adjusted Every 1 kg/m 2 increment 1.16 (1.15 1.17)* 1.13 (1.12 1.14)* 1.14 (1.13 1.15)* Categorical Unadjusted <18.5 kg/m 2 1.00 1.00 1.00 18.5 22.9 kg/m 2 1.11 (0.98 1.25) 1.04 (0.94 1.15) 1.06 (0.98 1.15) 23.0 24.9 kg/m 2 1.48 (1.31 1.68)* 1.35 (1.22 1.50)* 1.40 (1.29 1.51)* 25.0 29.9 kg/m 2 2.32 (2.05 2.62)* 1.98 (1.79 2.20)* 2.11 (1.95 2.29)* 30 kg/m 2 3.94 (3.39 4.58)* 3.31 (2.93 3.74)* 3.56 (3.24 3.91)* Adjusted <18.5 kg/m 2 1.00 1.00 1.00 18.5 22.9 kg/m 2 1.23 (1.07 1.41)* 1.09 (0.97 1.22) 1.14 (1.04 1.25)* 23.0 24.9 kg/m 2 1.78 (1.55 2.05)* 1.49 (1.32 1.68)* 1.60 (1.46 1.75)* 25.0 29.9 kg/m 2 3.06 (2.66 3.53)* 2.32 (2.06 2.60)* 2.60 (2.38 2.85)* 30 kg/m 2 6.59 (5.56 7.82)* 4.40 (3.84 5.05)* 5.16 (4.64 5.73)* *p<0.01. Adjusted variables include age, sex (for both sexes), diabetes duration, smoking and parental hypertension. and sex. More than 65% of the patients were overweight or obese (>40%) and less than one-third had a normal BMI. Obesity, especially with BMI 30 kg/m 2, was more prevalent in the younger patients; and small percentages of the patients were underweight. Fig 1 shows the prevalence of hypertension by subgroup of BMI and age in the respective sexes. It was evident that hypertension increased with age and simultaneously with BMI. All of the p values for the test of the trend of increasing prevalence of hypertension with increasing BMI or age were <0.001. Table 2 shows the OR and their 95%CI for hypertension. BMI either as a continuous variable or as a categorical variable was significantly associated with hypertension. Table 3 shows the regression coefficients for SBP and DBP estimated by BMI. All of the models were significant and the regression coefficients were larger for patients without a hypertension history when compared with those with such a history. Fig2 shows the estimated SBP and DBP for the percentiles of BMI after adjustment for age, sex, diabetic duration, smoking and parental hypertension. The curves for the patients without a history of hypertension showed a steeper increase of SBP and DBP with increasing BMI than those for patients with a hypertension history.

1752 TSENG C-H Table 3 Regression Coefficient (SE) for Systolic and Diastolic Blood Pressure Estimated by Body Mass Index in Linear Regression Models Blood pressure Men Women Both sexes All patients Systolic Unadjusted 0.589 (0.018) 0.613 (0.016) 0.601 (0.012) Adjusted 0.685 (0.018) 0.693 (0.017) 0.689 (0.012) Diastolic Unadjusted 0.464 (0.012) 0.445 (0.010) 0.451 (0.008) Adjusted 0.495 (0.012) 0.477 (0.011) 0.485 (0.008) Patients without hypertension history Systolic Unadjusted 0.534 (0.019) 0.556 (0.018) 0.540 (0.013) Adjusted 0.618 (0.019) 0.637 (0.019) 0.628 (0.013) Diastolic Unadjusted 0.418 (0.014) 0.430 (0.012) 0.421 (0.009) Adjusted 0.447 (0.014) 0.462 (0.012) 0.456 (0.009) Patients with hypertension history Systolic Unadjusted 0.205 (0.029) 0.285 (0.025) 0.254 (0.019) Adjusted 0.249 (0.032) 0.339 (0.026) 0.305 (0.020) Diastolic Unadjusted 0.283 (0.019) 0.258 (0.015) 0.267 (0.012) Adjusted 0.261 (0.020) 0.270 (0.017) 0.267 (0.013) All p-values are <0.001. Adjusted variables include age, sex (for both sexes), diabetic duration, smoking and parental hypertension. Diastolic BP (mmhg) Systolic BP (mmhg) 141 137 133 129 125 86 84 82 80 78 76 74 20.4 21.7 22.6 23.4 24.2 25.1 26.0 27.3 29.1 Body mass index (kg/m 2 ) 20.4 21.7 22.6 23.4 24.2 25.1 26.0 27.3 29.1 Body mass index (kg/m 2 ) Fig 2. Mean values for systolic and diastolic blood pressure (BP) for the 10 th to the 90 th percentile of body mass index after adjustment for age, sex, diabetic duration, smoking and parental hypertension. (Closed circle) Patients with hypertension history; (closed square) patients without hypertension history. The numbers on the X-axis from left to right are the respective values of body mass index from the 10 th to the 90 th percentile. Discussion The findings of this study suggested that BMI is a major and independent determinant of blood pressure and hypertension in adult Taiwanese patients with T2DM. At the recommended cutoff of BMI 25kg/m 2 for Asian people, 11 the prevalence of obesity was 39.3% for diabetic men and 41.7% for diabetic women (Table1). These figures were much higher than the respective prevalence rates of 24.6% and 23.3% for the adult general population in Taiwan. 19 For BMI 30 kg/m 2, the respective prevalence for men and women was 5.7% and 8.4% in the diabetic patients (Table 1), in contrast to 2.3% and 5.0% in the general population. 19 The correlation between BMI/obesity and blood pressure/hypertension was consistently demonstrated in the different age groups of each sex (Fig 1) and after adjustment for confounders (Tables 2,3, Fig 2). Despite the different distribution of BMI and obesity in the different age groups in each sex (Table 1), the association seemed to be independent (Tables 2,3, Fig 1) and existed in patients with or without a history of hypertension (Table 3, Fig 2). However, as

Obesity and HT in Taiwanese T2DM shown in Table3 and Fig2, the relationship was stronger in those without a hypertension history, probably because of exclusion of the confounding effect caused by the use of antihypertensive agents. In the US, women who are overweight or obese have a 3- or 6-fold higher risk of hypertension, respectively. 20 These were very similar to the findings in the diabetic men of the present study (Table 2) when using BMI 25 kg/m 2 and 30 kg/m 2 as the cutoffs for overweight and obesity, respectively (as recommended for Caucasians). However, the diabetic women showed an increased risk with lesser magnitude than the diabetic men (Table 2). The relative risk for hypertension associated with being overweight declines with age in the general population of the USA. 3 The present study also showed a similar attenuation of the prevalence ratios of hypertension associated with increasing age in the diabetic patients (Fig 1). This result was because of the lower prevalence of hypertension in the younger patients and the relatively higher prevalence in the older ones in the lower BMI categories (Fig1). The relationship between BMI and SBP or DBP seemed not to be limited to the range of obesity, but extended throughout the nonobese range (Table 2, Figs 1,2), though attenuation of the correlation between BMI and prevalence of hypertension was found at lower BMIs, especially in the diabetic women (Fig 1A). In this study, when calculated from patients without hypertension history in the adjusted models (Table 3), a BMI increase of 1 kg/m 2 corresponded to an increase in SBP of 0.62 and 0.64mmHg in men and women, respectively; and a respective increase in DBP of 0.45 and 0.46mmHg. These were comparable to the findings in a previous study that showed a gain of 1 unit BMI corresponded to an increase in SBP of approximately 0.59 and 0.80 mmhg in men and women, respectively. 21 The finding of an increase of BMI of 1 kg/m 2 with an increase of 11% and 13% in the unadjusted and adjusted risk for hypertension, respectively, in the diabetic women of the present study (Table2) also conformed with the finding of a 12% increase in the adjusted risk for hypertension in women in the US. 3 Study Limitations First, the cross-sectional approach did not provide evidence for causality. Second, although a self-reported BMI has been used in many previous epidemiologic studies and correlated well with measured data, 22 24 it is always underestimated because height is always over-reported and weight under-reported, especially at higher BMIs. 25 27 Therefore the prevalence of obesity could be underestimated and the risk for the obese group could also be biased. Third, blood pressure was also based on self-reporting without actual measurement, so the accuracy of the reported blood pressure was unknown. It is also unclear whether the reported blood pressure levels were those before or after treatment with medications in patients taking antihypertensive agents. However, a reported blood pressure that the patients ordinarily had would more actually reflect the average blood pressure of the patients over a period of time rather than the measured value taken at the time of measurement. In conclusion, diabetic patients have a high prevalence of obesity and hypertension, and BMI/obesity is strongly linked with hypertension and blood pressure, with the correlation extending towards the nonobese range. This relationship conforms to findings derived from general populations of other ethnicities. 1753 Acknowledgments This study was partly supported by grants from the Department of Health (DOH89-TD-1035), the National Taiwan University Hospital Yun- Lin Branch (NTUHYL96.G001) and the National Science Council (NSC- 86 2314-B-002 326, NSC-87 2314-B-002 245, NSC88 2621-B-002 030, NSC89 2320-B002 125, NSC-90 2320-B-002 197, NSC-92 2320- B-002 156, NSC-93 2320-B-002 071, NSC-94 2314-B-002 142 and NSC-95 2314-B-002 311), Taiwan. References 1. Hubert HB, Eaker ED, Garrison RJ, Castelli WP. Life-style correlates of risk factor change in young adults: An eight-year study of coronary heart disease risk factors in the Framingham offspring. Am J Epidemiol 1987; 125: 812 831. 2. Must A, Spadano J, Coakley EH, Field AE, Colditz G, Dietz WH. The disease burden associated with overweight and obesity. JAMA 1999; 282: 1523 1529. 3. Huang Z, Willett WC, Manson JE, Rosner B, Stampfer MJ, Speizer FE, et al. Body weight, weight change, and risk for hypertension in women. Ann Intern Med 1998; 128: 81 88. 4. Wilsgaard T, Schirmer H, Arnesen E. Impact of body weight on blood pressure with a focus on sex differences: The Tromso Study, 1986 1995. Arch Intern Med 2000; 160: 2847 2853. 5. Stevens J, Plankey MW, Williamson DF, Thun MJ, Rust PF, Palesch Y, et al. The body mass index-mortality relationship in white and African American women. Obes Res 1998; 6: 268 277. 6. Cossrow N, Falkner B. Race/ethnic issues in obesity and obesity-related comorbidities. J Clin Endocrinol Metab 2004; 89: 2590 2594. 7. Nakamura Y, Turin TC, Kita Y, Tamaki S, Tsujita Y, Kadowaki T, et al. Associations of obesity measures with metabolic risk factors in a community-based population in Japan. Circ J 2007; 71: 776 781. 8. Baba R, Koketsu M, Nagashima M, Inasaka H, Yoshinaga M, Yokota M. Adolescent obesity adversely affects blood pressure and resting heart rate. Circ J 2007; 71: 722 726. 9. Sun Z, Zheng L, Wei Y, Li J, Zhang X, Zhang X, et al. Prevalence and risk factors of the rural adult people prehypertension status in Liaoning Province of China. Circ J 2007; 71: 550 553. 10. Shiraishi J, Kohno Y, Sawada T, Nishizawa S, Arihara M, Hadase M, et al. Relation of obesity to acute myocardial infarction in Japanese patients. Circ J 2006; 70: 1525 1530. 11. World Health Organization. The Asia-Pacific Perspective: Redefining obesity and its treatment. Available: http://www.iotf.org/asiapacific/ (accessed July 15, 2007) 12. Tseng CH. Body mass index and waist circumference as determinants of coronary artery disease in Taiwanese adults with type 2 diabetes mellitus. Int J Obes 2006; 30: 816 821. 13. Lakatta EG. Mechanisms of hypertension in the elderly. J Am Geriatr Soc 1989; 37: 780 790. 14. Tseng CH. Higher risk of hypertension in indigenous type 2 diabetic patients in Taiwan. J Hypertens 2006; 24: 1817 1821. 15. Tseng CH. Mortality and causes of death in a national sample of diabetic patients in Taiwan. Diabetes Care 2004; 27: 1605 1609. 16. Tseng CH. Prevalence of lower-extremity amputation among patients with diabetes mellitus: Is height a factor? Can Med Assoc J 2006; 174: 319 323. 17. Tseng CH, Tseng CP, Chong CK, Huang TP, Song YM, Chou CW, et al. Increasing incidence of diagnosed type 2 diabetes in Taiwan: Analysis of data from a national cohort. Diabetologia 2006; 49: 1755 1760. 18. Tseng CH. Exogenous insulin use and hypertension in adult patients with type 2 diabetes mellitus. Arch Intern Med 2006; 166: 1184 1189. 19. Lin YC, Yen LL, Chen SY, Kao MD, Tzeng MS, Huang PC, et al. Prevalence of overweight and obesity and its associated factors: Findings from National Nutrition and Health Survey in Taiwan, 1993 1996. Prev Med 2003; 37: 233 241. 20. Witteman JC, Willett WC, Stampfer MJ, Colditz GA, Sacks FM, Speizer FE, et al. A prospective study of nutritional factors and hypertension among US women. Circulation 1989; 80: 1320 1327. 21. Kissebah AH, Krakower GR. Regional adiposity and morbidity. Physiol Rev 1994; 74: 761 811. 22. Freedman DS, Khan LK, Serdula MK, Galuska DA, Dietz WH. Trends and correlates of class 3 obesity in the United States from 1990 through 2000. JAMA 2002; 288: 1758 1761. 23. Calle EE, Thun MJ, Petrelli JM, Rodriguez C, Heath CW Jr. Bodymass index and mortality in a prospective cohort of US adults. N Engl J Med 1999; 341: 1097 1105. 24. Lee IM, Manson JE, Hennekens CH, Paffenbarger RS Jr. Body

1754 TSENG C-H weight and mortality: A 27-year follow-up of middle-aged men. JAMA 1993; 270: 2823 2828. 25. Stewart AW, Jackson RT, Ford MA, Beaglehole R. Underestimation of relative weight by use of self-reported height and weight. Am J Epidemiol 1987; 125: 122 126. 26. Rowland ML. Self-reported weight and height. Am J Clin Nutr 1990; 52: 1125 1133. 27. Kuczmarski MF, Kuczmarski RJ, Najjar M. Effects of age on validity of self-reported height, weight, and body mass index: Findings from the Third National Health and Nutrition Examination Survey, 1988 1994. J Am Diet Assoc 2001; 101: 28 34.