NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Whitlock EP, Williams SB, Gold R, et al. Screening and Interventions for Childhood Overweight [Internet]. Rockville (MD): Agency for Healthcare Research and Quality (US); 2005 Jul. (Evidence Syntheses, No. 36.)

  • This publication is provided for historical reference only and the information may be out of date.

This publication is provided for historical reference only and the information may be out of date.

Cover of Screening and Interventions for Childhood Overweight

Screening and Interventions for Childhood Overweight [Internet].

Show details

Appendix K. Childhood Overweight Clinical Screening Tests to Predict Poorer Health Outcomes

Overview of Methods and Prior Reviews

For the relationship between clinical screening measures of overweight and health consequences, we located three overview or review articles1 2, 3, one extensive( but non-systematic) review,4 and one fair- or good-quality systematic review5 at the time we conducted our literature search. The fair-quality systematic evidence review examined the consequences of childhood obesity in childhood and longer term through a comprehensive literature review from January 1981 through December 2001.5 This review used established critical appraisal methodology, including explicit methodological quality rating and a hierarchical study design approach. However, it did not distinguish between cohort and cross-sectional studies in examining the childhood or adult consequences of overweight, nor did it systematically report the quantitative relationship between childhood overweight measures and health consequences. We used this review only as a source of potentially relevant studies. The extensive non-systematic review of long-term health risks of child and adolescent fatness that searched through 19964 located seven U.S. studies addressing childhood overweight and adult health outcomes.612 This review concluded that there were too few studies on which to base firm conclusions about long-term health risks in relation to childhood and adolescent adiposity.4

In addition to screening prognosis trials located in our searches, we retrieved all non-duplicated possibly relevant citations from the five review articles, and from continuing to check bibliographies. Through this process we located another four non-systematic reviews related to health consequences and obesity measures1316; which we reviewed for additional articles not previously located through other sources. None were found.

There were insufficient studies from all of these sources and from our searches to critically appraise the prospective relationship of childhood overweight to childhood health outcomes. We therefore focused on the prospective studies addressing childhood overweight and adult health outcomes.

Longitudinal U.S. Studies of Adult Health Consequences of Childhood Overweight

Table K-1 lists the 11 U.S. studies we examined for this key question.7, 8, 1012, 1722 We excluded a large number of cross-sectional studies identified in previous reviews and other studies for quality reasons (Appendix G). Among the longitudinal studies identified above, we excluded one that had follow-up measures in late adolescence (15–18 years) and not adulthood.9 We excluded another6 for using a non-comparable overweight definition and weight reference standard, for incomplete follow-up of the cohort (700/2,000), and for failing to address loss due to mortality in their morbidity analyses.

Table K-1. Evidence Table: Relationship between Childhood Overweight and Adult Health Measures Other than Overweight.

Table K-1

Evidence Table: Relationship between Childhood Overweight and Adult Health Measures Other than Overweight.

Morbidity and mortality. One often-cited fair- to poor-quality cohort study used a constructed sample (n=508, 27% of the original cohort) from the Third Harvard Growth Study conducted from 1922 to 1935.7 A subsample of this group (n=309, 61%) had adult BMIs measured at mean age of 55 years that could be used in adjusted analyses. For most other analyses it is not clear whether the entire sample (n=508) was used, and if so, what assumptions were made to include those lost to follow-up or who declined participation (166, 32%). The study reported the impact of adolescent BMI (>75th percentile according to NHANES I compared with 25th–50th percentile) between ages 13 and 18 on all-cause mortality, coronary heart disease (CHD) mortality, atherosclerotic heart disease mortality, and colorectal cancer mortality in white males and females over 50 years later. Males in the higher BMI quartile had small increased relative risks for all-cause mortality (RR 1.8, 95% CI 1.2–2.7) and for mortality from CHD (RR 2.3, 95% CI1.4–4.1), atherosclerotic cardiovascular disease (RR 13.2, 95% CI 1.6–108.0), and colorectal cancer (RR 9.1, 95% CI 1.1–77.5). Females in the higher quartile of adolescent BMI did not show significantly elevated risks for mortality from these conditions or from breast cancer. In the subsample (n=309) with measured adult weights at 55 years of age, adjustment for adult BMI slightly decreased the RR for all-cause mortality in males and removed the increased RR for CHD mortality. However, the RR for all-cause mortality cited in the text as “before adult BMI adjustment” in males (2.9, 95% CI 1.5–5.8) does not match that reported for the entire sample in their Table 2, raising the question of the generalizability of the subsample analysis of adjustment for adult BMI on the risk of adult mortality and morbidity from childhood overweight. Morbidity analyses used a further subsample interviewed in 1988 (n=181). Lack of details about the selection and characteristics of this subsample preclude its use.

In another often-cited fair- to poor-quality nested case-control study,8 mortality odds were calculated for measured relative weight (defined internally using the sample) in a population-based study of 13,146 children ages 5 and 18 in 1933-1945. Death certificate information was available for 5,471/13,146 of the population (42%). Pre-pubertal relative weight measures were those before age 10 in girls and age 12 in boys, while post-pubertal measures were those after age 13.5 in girls and 15.5 in boys. A total of 509 deaths were identified, 308 in males (median age at death, 50 years) and 201 in females (median age at death, 51 years). Controlling for sex and year of birth, those in the highest quintile of pre-pubertal and post-pubertal weight had the same, slightly increased mortality odds (OR 1.5, 95% CI 1.0–2.4). The analysis did not control for adult BMI, race, or other sociodemographic factors significantly related to mortality.

Socioeconomic outcomes. A fair- to good-quality longitudinal cohort study using the National Longitudinal Survey of Labor Market Experience Youth Cohort from 1979 examined the risk of lower household income, household poverty, likelihood of marriage, years completed of school, and self-esteem in young adulthood (23–31 years of age) for overweight 16–24 year olds.12 Overweight was defined as those above the 95th percentile of BMI (in NHANES I), and compared with all others in the cohort.12 The cohort was 51% female, 80% white, 14% black, and 6% Hispanic. Between 3.0% and 3.4% were overweight (BMI >95th percentile) at baseline, with black adolescent females significantly more overweight than non-Hispanic whites (5.8% vs. 2.5%, P<.001). At follow-up, 77% of the men and 66% of the women were still overweight. For females, but not males, overweight in adolescence was associated with completing 0.3 mean fewer years of schooling, $6,710 lower household income, and 10% higher rate of poverty. Overweight males and females were less likely to have married (11% and 20% less likely, respectively) but had no differences in self-esteem in young adulthood. Among males only, being 12 inches shorter in height at baseline was independently associated with a 10% higher prevalence of poverty (95% CI 6–13%). The researchers did not control for adult BMI in their analyses but did control for baseline socioeconomic status and aptitude. Redefining overweight as over the 85th percentile increased the prevalence of overweight at baseline and reduced some (and eliminated others) of the reported risk relationships with adult social and economic factors in women and the risk relationship to marriage in men.

Cardiovascular disease and diabetes risk factors. One good-quality cohort study used data from the Bogalusa Heart Study (32% black and 57% female) to examine the longitudinal relationship between childhood BMI or triceps skinfold thickness measured at a mean age of 10 +/- 3 years and adult BMI; lipids (total cholesterol, LDL and HDL cholesterol); insulin; and systolic and diastolic blood pressure after a mean of 17 years of follow-up (adult ages of 18 to 37 years, mean age 26.2 +/- 6.3 years).19 BMI levels in childhood and adulthood were more strongly associated with all adult risk factors than TSF measures in childhood and adulthood. When examined separately, adult BMI levels were moderately correlated with adult cardiovascular risk factors (r=0.21–0.59), as were childhood BMI levels, although the correlations tended to be about 50% as strong (r=0.09–0.26). Controlling for adult BMI eliminated the association of childhood BMI with adult CV risk factors, indicating that the effect of childhood weight status was mediated through its relationship to adult BMI. For all six risk factors, adjusting for adult BMI actually reversed the relationship between greater childhood BMI and greater adult CV risk factors, although these correlations tended to be quite small (<0.15 absolute value). Compared with adults who had normal childhood BMIs (<50th percentile), adults who were overweight in childhood (BMIs ≥95th percentile) had significantly higher adult BMI (34.9 +/- 7 vs. 22.5 +/- 4) and significantly higher CVD risk factor measures (although HDL cholesterol was significantly lower).

However, in an analysis stratifying by both adult and childhood BMI, obese adults who had been overweight as children had similar adult CVD risk factors compared to obese adults who had been normal weight as children. Obese adults who had been overweight children were significantly more obese (mean BMI 38.1) than obese adults who had been normal weight as children (mean BMI 33.2, p<0.05). A similar pattern was seen when comparing normal-weight adults who had been overweight or not as children. The authors did not compare the proportion of adults who were hypertensive or hyperlipidemic, although the proportion diagnosed with diabetes did not differ between those who were overweight vs. normal weight as children, once adult BMI was taken into account. A subset analysis on timing of obesity and adult CV risk factors was reported, but is not considered here due to its limited power and our inability to confirm lack of selection bias.

Seven fair-quality longitudinal cohort studies examined the relationship between childhood overweight and adult CVD risk factors.10, 11, 17, 18, 2022 None of these adjusted for adult BMI in examining this relationship, although four studies included the change from childhood to adult BMI in their analyses.10, 11, 18, 22 Given the potential confounding of adult BMI on the relationship between childhood overweight and adult risk factors, we confine our discussion to the four studies considering change. Three of these studies included racial/ethnic minorities, with about one-third consisting of blacks 11, 22 or blacks and Native Americans.18 The most informative study examined the relationship between elevated childhood lipid levels (according to National Cholesterol Education Program [NCEP] guidelines) at age 5–14 and adult dyslipidemia (elevated total cholesterol, LDL-C, HDL-C, or triglycerides according to NCEP guidelines) at age 20–34 in children from the Bogalusa Heart Study, controlling for race, sex, age, baseline BMI, baseline lipids, and change in BMI from childhood to adulthood.22 Children with baseline LDL-C above 101 had the greatest odds (2.5, 95% CI 2.0–3.1) for adult dyslipidemia. Baseline BMI, change in BMI from childhood to adulthood, and older age all independently raised the odds of adult dyslipidemia (OR 1.7–1.9). Controlling for other factors, females and blacks were significantly less likely to have adult dyslipidemia (61% and 42%, p<0.01). These data are suggestive of a role for each of these factors, including greater weight gain from childhood to adulthood. Without controlling for adult BMI, however, it is not clear that these are independent of adult overweight.

In another study using the Bogalusa Heart study cohort, CVD risk factors at ages 27–31 years were compared in adults who had become overweight as adolescents (above the 75th percentile for age- and sex-specific BMI) and remained overweight as adults with those who were consistently lean (25th–50th percentiles).11 Among 191/783 adolescents ages 13–17 years identified as overweight, 110 (58%) remained overweight as adults, 64 were between the 50th and 75th percentiles, and 17 were below the 25th percentile. The predictive value of overweight status was lowest in black males (52%) and highest in black females (62%). Conversely, the predictive value of lean status was lowest in black males (28%) and highest in white males (52%). Compared with the consistently lean cohort, higher adolescent BMI and change in BMI were associated with increased blood pressure, lipids (decreased HDL-C), glucose, and insulin levels. With the relatively strong tracking of adolescent to adult BMI, it is not possible to understand the independent contribution of these factors beyond adult BMI.

Using the Muscatine cohort, childhood blood pressure, childhood BMI, change in BMI from childhood to adulthood, family history, and behavioral risks at ages 7–18 years were related to adult systolic blood pressure at ages 20–30.10 At age 16, probability of adult systolic blood pressure above the 90th percentile reached about 0.3 in females and 0.4 in males at the 90th percentile for BMI, and increased steeply with greater BMI percentile. This analysis did not consider adult BMI. Across ages and sex, without considering adult BMI, change in BMI from childhood to adulthood and childhood blood pressure level were independent predictors of adult systolic blood pressure, jointly explaining 14%–24% of the variance.

Childhood BMI at age eight and change in BMI in childhood and adolescence were examined in 679 children across a broad range of BMI measures (8 to 34, mean 16.5 in relation to blood pressure, fasting insulin, and lipids at age 24.18 Childhood and adult BMI were strongly correlated (r=0.61). Childhood BMI was not independently related to adult CVD risk factors, after childhood and adolescent weight gain were considered. Weight gain in childhood and adulthood could be surrogates for eventual adult overweight.

References for Appendix K

Serdula MK, Ivery D, Coates RJ, Freedman DS, Williamson DF, Byers T. Do obese children become obese adults? A review of the literature. Prev Med. 1993;22(2):167–177. [PubMed: 8483856]
Dietz WH. Childhood weight affects adult morbidity and mortality. Journal of Nutrition. 1998;128(2 Suppl):411S–414S. [PubMed: 9478038]
Must A. Morbidity and mortality associated with elevated body weight in children and adolescents. Am J Clin Nutr. 1996;63(3 Suppl):445S–447S. [PubMed: 8615339]
Power C, Lake JK, Cole TJ. Measurement and long-term health risks of child and adolescent fatness. Int J Obes Relat Metab Disord. 1997;21(7):507–526. [PubMed: 9226480]
Reilly JJ, Methven E, McDowell ZC, Hacking B, Alexander D, Stewart L. et al. Health consequences of obesity. Arch Dis Child. 2003;88(9):748–752. [PMC free article: PMC1719633] [PubMed: 12937090]
Abraham S, Collins G, Nordsieck M. Relationship of childhood weight status to morbidity in adults. HSMHA Health Rep. 1971;86(3):273–284. [PMC free article: PMC1937073] [PubMed: 5547793]
Must A, Jacques PF, Dallal GE, Bajema CJ, Dietz WH. Long-term morbidity and mortality of overweight adolescents. A follow-up of the Harvard Growth Study of 1922 to 1935. New England Journal of Medicine. 1992;327(19):1350–1355. [PubMed: 1406836]
Nieto FJ, Szklo M, Comstock GW. Childhood weight and growth rate as predictors of adult mortality. American Journal of Epidemiology. 1992;136(2):201–213. [PubMed: 1415142]
Clarke WR, Woolson RF, Lauer RM. Changes in ponderosity and blood pressure in childhood: the Muscatine Study. Am J Epidemiol. 1986;124(2):195–206. [PubMed: 3728436]
Lauer RM, Lee J, Clarke WR. Predicting adult cholesterol levels from measurements in childhood and adolescence: the Muscatine Study. Bulletin of the New York Academy of Medicine. 1989;65(10):1127–1142. [PMC free article: PMC1807893] [PubMed: 2629967]
Srinivasan SR, Bao W, Wattigney WA, Berenson GS. Adolescent overweight is associated with adult overweight and related multiple cardiovascular risk factors: the Bogalusa Heart Study. Metabolism: Clinical & Experimental. 1996;45(2):235–240. [PubMed: 8596496]
Gortmaker SL, Must A, Perrin JM, Sobol AM, Dietz WH. Social and economic consequences of overweight in adolescence and young adulthood. New England Journal of Medicine. 1993;329(14):1008–1012. [PubMed: 8366901]
Zametkin AJ, Zoon CK, Klein HW, Munson S. Psychiatric aspects of child and adolescent obesity: a review of the past 10 years. J Am Acad Child Adolesc Psychiatry. 2004;43(2):134–150. [PubMed: 14726719]
Dietz WH. Health consequences of obesity in youth: childhood predictors of adult disease. Pediatrics. 1998;101(3 Pt 2):518–525. [PubMed: 12224658]
Must A, Strauss RS. Risks and consequences of childhood and adolescent obesity. Int J Obes Relat Metab Disord. 1999;23(Suppl 2):S2–11. S2–11. [PubMed: 10340798]
Must A. Does overweight in childhood have an impact on adult health? Nutr Rev. 2003;61(4):139–142. [PubMed: 12795448]
Freedman DS, Khan LK, Serdula MK, Dietz WH, Srinivasan SR, Berenson GS. Inter-relationships among childhood BMI, childhood height, and adult obesity: the Bogalusa Heart Study. International Journal of Obesity & Related Metabolic Disorders. 2004;28(1):10–16. [PubMed: 14652621]
Sinaiko AR, Donahue RP, Jacobs DR Jr., Prineas RJ. Relation of weight and rate of increase in weight during childhood and adolescence to body size, blood pressure, fasting insulin, and lipids in young adults. The Minneapolis Children's Blood Pressure Study. Circulation. 1999;99(11):1471–1476. [PubMed: 10086972]
Freedman DS, Khan LK, Dietz WH, Srinivasan SR, Berenson GS. Relationship of childhood obesity to coronary heart disease risk factors in adulthood: the Bogalusa Heart Study. Pediatrics. 2001;108(3):712–718. [PubMed: 11533341]
Lauer RM, Clarke WR, Burns TL. Obesity in childhood: the Muscatine Study. Chung-Hua Min Kuo Hsiao Erh Ko i Hsueh Hui Tsa Chih. 1997;38(6):432–437. [PubMed: 9473814]
Srinivasan SR, Myers L, Berenson GS. Predictability of childhood adiposity and insulin for developing insulin resistance syndrome (syndrome X) in young adulthood: the Bogalusa Heart Study. Diabetes. 2002;51(1):204–209. [PubMed: 11756342]
Bao W, Srinivasan SR, Wattigney WA, Bao W, Berenson GS. Usefulness of childhood low-density lipoprotein cholesterol level in predicting adult dyslipidemia and other cardiovascular risks. The Bogalusa Heart Study. Archives of Internal Medicine. 1996;156(12):1315–1320. [PubMed: 8651840]


  • PubReader
  • Print View
  • Cite this Page

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...