Advertising Disclaimer »

Article

The Relation of Overweight to Cardiovascular Risk Factors Among Children and Adolescents: The Bogalusa Heart Study

David S. Freedman, William H. Dietz, Sathanur R. Srinivasan and Gerald S. Berenson
Pediatrics June 1999, 103 (6) 1175-1182; DOI: https://doi.org/10.1542/peds.103.6.1175
Download PDF

Abstract

Background. Although overweight and obesity in childhood are related to dyslipidemia, hyperinsulinemia, and hypertension, most studies have examined levels of these risk factors individually or have used internal cutpoints (eg, quintiles) to classify overweight and risk factors.

Objective. We used cutpoints derived from several national studies to examine the relation of overweight (Quetelet index, >95th percentile) to adverse risk factor levels and risk factor clustering.

Design. The sample consisted of 9167 5- to 17-year-olds examined in seven cross-sectional studies conducted by the Bogalusa Heart Study between 1973 and 1994.

Results. About 11% of examined schoolchildren were considered overweight. Although adverse lipid, insulin, and blood pressure levels did not vary substantially with the Quetelet index at levels <85th percentile, risk factor prevalences increased greatly at higher levels of the Quetelet index. Overweight schoolchildren were 2.4 times as likely as children with a Quetelet index <85th percentile to have an elevated level of total cholesterol. Odds ratios for other associations were 2.4 (diastolic blood pressure), 3.0 (low-density lipoprotein cholesterol), 3.4 (high-density lipoprotein cholesterol), 4.5 (systolic blood pressure), 7.1 (triglycerides), and 12.6 (fasting insulin). Several of these associations differed between whites and blacks, and by age. Of the 813 overweight schoolchildren, 475 (58%) were found to have at least one risk factor. Furthermore, the use of overweight as a screening tool could identify 50% of schoolchildren who had two or more risk factors.

Conclusions. Because overweight is associated with various risk factors even among young children, it is possible that the successful prevention and treatment of obesity in childhood could reduce the adult incidence of cardiovascular disease.overweight, lipids, blood pressure, insulin, children, adolescents.

  • Abbreviations:
    OR =
    odds ratio •
    CVD =
    cardiovascular disease •
    SBP =
    systolic blood pressure •
    P =
    percentile •
    TC =
    total cholesterol •
    TG =
    triglycerides •
    LDLC =
    low-density lipoprotein cholesterol •
    HDLC =
    high-density lipoprotein cholesterol •
    DBP =
    diastolic blood pressure

    • Overweight children and adolescents have adverse levels of lipids, insulin, and blood pressure,1–6 and longitudinal changes in relative weight are associated with changes in these risk factors.7 Furthermore, various measures of adiposity among adolescents and young adults are associated with fatty streaks, raised lesions, and calcifications in the aorta and coronary arteries.8–10 Overweight children are also at increased risk for various chronic diseases in later life,11 and some evidence12 suggests that this association may exist independently of obesity status in adulthood.

    Much of the information on overweight and risk factors among youth, however, has been summarized by using correlation3 ,5 ,13 ,14or regression4 coefficients, techniques that may not reflect the nonlinearity of the underlying associations.15Furthermore, most studies have provided little information on either 1) the probability that an overweight child will have adverse risk factor levels (predictive value), or 2) the proportion of children with adverse risk factors that can be detected by examining overweight children (sensitivity). Although odds ratios (ORs) and other categorical techniques have been used by some investigators,2 14–19 the frequent use of internal cutpoints (eg, quartiles) complicates comparisons across studies. Relatively little attention has been given to the importance of overweight in risk factor clustering6 16–18 ,20 and to differences in these associations by sex, race/ethnicity, or age.3–5 ,15

    The need for additional information on the ability of overweight to detect adverse risk factors has been emphasized by the publication of guidelines that define overweight,21hypercholesterolemia,22 ,23 and hypertension24among children and adolescents. By using recommended cutpoints, the current study examines the relation of overweight to adverse levels of lipids, insulin, and blood pressure among 5- to 17-year-olds. The study population consists of 9167 schoolchildren who participated in the Bogalusa Heart Study between 1973 and 1994.

    METHODS

    Study Population

    The Bogalusa (Louisiana) Heart Study is a community-based study of cardiovascular disease (CVD) risk factors in early life.25 The 1990 population of the biracial (Embedded Image black) community, which is fairly typical of semirural towns in the South, was ∼43 000 (Washington Parish). Seven cross-sectional examinations of schoolchildren in Bogalusa, each with a participation rate of >80%, were conducted between 1973 and 1994.26Approximately 55% of all subjects were examined more than once, and the current analyses are based on measurements obtained at the final screening of each individual.

    Analyses were restricted to fasting subjects who had recorded values for weight, height, total cholesterol, and systolic blood pressure (SBP). These criteria resulted a sample of 9167 5- to 17-year-olds (mean age, 11.9 years); levels of insulin were first measured in the 1981 to 1982 survey, and were available for 6349 subjects. The examinations that contributed the largest number of persons to the current analyses were the 1984 to 1985 (n = 1315), 1987 to 1988 (n = 1736), and 1992 to 1994 (n= 2648) studies.

    Measurements

    Anthropometry

    Height was measured to the nearest 0.1 cm with an Iowa Height Board, and weight to the nearest 0.1 kg, using a balance beam metric scale. No adjustments were made for the weight of the gown, underpants, or socks that were worn during the examination. Although analyses focused on the Quetelet index (kg/m2) as a measure of relative weight, the Rohrer index (kg/m3) was also included in some analyses. (The Quetelet index is often referred to as the body mass index or BMI.) These weight-height indices are highly correlated (r ∼0.9), but the Rohrer index is less strongly correlated with height among schoolchildren than is the Quetelet index (r = −0.12 vs 2.62 in the current study). Triceps and subscapular skinfolds were each measured three times with Lange Skinfold Calipers (Cambridge Scientific Industries, Cambridge, MD).

    Overweight was defined as a Quetelet index above the smoothed 95th percentile (P) of combined data from five national surveys conducted between 1963 and 1994 (Health Examination Survey II and III, and National Health and Nutrition Examination Surveys I, II, and III). Percentiles were estimated for each sex and 6-month age interval, using the LMS technique.27 This method summarizes age trends in 1) the power (the λ in LMS) of the Box-Cox transformation needed to approximate normality, 2) the mean, and 3) coefficient of variation (ς). The resulting smoothed curves allow any percentile (orZ score) of the Quetelet index to be calculated, and this method has recently been used in the development of a research child growth reference.28 Although numerous adjectives, including obese and severely overweight, have been applied to persons with high levels of the Quetelet index,29 we refer to schoolchildren with a Quetelet index >95th P as overweight.

    The 95th P cutpoints for the classification of overweight, along with the prevalence of overweight in the Bogalusa sample, are summarized inTable 1. (It should be noted that the estimated percentiles based on the LMS method are fairly comparable with previously published smoothed percentiles from National Health and Nutrition Examination Survey I using 1-year age intervals.30) Although the prevalence of overweight among both boys and girls was 10.8%, prevalences were highest among 11-year-old girls (17.6%) and 10-year-old boys (15.6%), and lowest among 5-year-olds (4.4% boys, 6.0% girls). Age was positively related to both the level of the Quetelet index and the range of values, with the 95th P of the Quetelet index among 5-year-old boys (18.4 kg/m2) approximately equal to the median value among 12-year-olds. Furthermore, the difference between the 95th P and the median Quetelet index among boys increased with age, ranging from 3 kg/m2 (18.5 to 15.5 kg/m2) among 5-year-olds to 7 kg/m2 among 17-year-olds.

    View this table:
    Table 1.

    Median Levels of Quetelet Index and Proportion of Persons Considered Overweight

    Laboratory Determinations

    Serum concentrations of total cholesterol (TC) and triglycerides (TG) were determined by using enzymatic procedures in a centralized laboratory that is monitored by the Lipid Standardization Program of the National Centers for Disease Control and Prevention (CDC). Determinations of low-density lipoprotein cholesterol (LDLC) and high-density lipoprotein cholesterol (HDLC) used a combination of heparin-calcium precipitation and agar-agarose gel electrophoresis.31 The cholesterol and TG analyses were monitored, using quality control samples, and the accuracy was well within the limits set by the CDC. Plasma insulin determinations were performed, starting with the fourth (of seven) cross-sectional examinations in Bogalusa, using a radioimmunoassay procedure (Phaadebas Insulin Kit). Of the 6543 children examined after 1980 and included in the current analyses, ∼3% were missing data on insulin levels.

    A second blood specimen was drawn on each screening day from 10% of the subjects to assess the measurement error of each laboratory test. Intraclass correlation coefficients for the lipid and lipoprotein determinations ranged from 0.96 HDLC to 0.995 TG, but insulin levels were less reproducible (intraclass correlation coefficient = 0.93).

    High levels of TC and LDLC were defined as those >200 mg/dL and >130 mg/dL, respectively.23 As suggested,22 ,32 a TG level ≥130 mg/dL was considered high, and an HDLC level <35 mg/dL was considered low. (A TG level of 130 mg/dL is approximately the 95th P during the second decade of life.22) Because cutpoints for adverse levels of insulin have not been proposed for children, values ≥95th P (adjusted for race, sex, and age) were viewed as high.

    Blood Pressure

    Right arm, sitting, SBP, and diastolic blood pressure (DBP) were measured six times by trained observers with a mercury sphygmomanometer (Baumanometer); the allocation of subjects to instruments and observers was specified daily. The cuff size, which was based on the length and circumference of the upper arm, was chosen to be as large as possible without having the elbow skin crease obstruct the stethoscope.25 ,33 On each screening day, 10% of the examined children were selected to assess reproducibility, and based on replicate measurements from the 1992 to 1994 examination, intraclass (within-observer) correlation coefficients were 0.88 SBP and 0.80 DBP.

    Elevated levels of SBP and DBP were defined in a manner comparable with that used by the National High Blood Pressure Education Program.24 ,34 Regression models containing race, sex, age, age2, height, height2, and various product terms were used to predict blood pressure, and persons with residuals >95th P from these models were considered to have high levels of SBP or DBP. Blood pressure levels measured in the Bogalusa Heart Study are generally ∼10 mm Hg lower than in other studies of children and adolescents33; this difference is largely caused by the methods used to obtain blood pressure.

    Statistical Analyses

    Quetelet index and the various risk factors were categorized into two groups (high vs not high) in most analyses, and associations were summarized by using cross-tabulations and ORs (adjusted for race, sex, age, and year of study) estimated in logistic regression models. The sensitivity (proportion of persons with specified risk factor who would be detected by screening overweight persons) and positive predictive value (proportion of overweight persons with the risk factor) for each risk factor was also calculated. Analyses that examined the relation of overweight (Quetelet index ≥95th P) to various risk factors excluded subjects with values between the 85th and 94th P to more clearly contrast overweight versus nonoverweight persons.

    We also examined the proportion of persons with various risk factors by categories of the Quetelet index, and nonlinearity (on a logit scale) was assessed by adding higher order polynomial terms for the Quetelet index to regression models. The possibility that associations with risk factors differed by race, sex, or age was assessed by adding interaction (product) terms to various regression models. We then examined the −2·log likelihoods of hierarchical models; the difference in this statistic between two models, the likelihood ratio, has a χ2 distribution.

    To examine risk factor clustering, the total number of risk factors for each person was calculated as the sum of an LDLC level ≥130 mg/dL, a TG level ≥130 mg/dL, an HDLC level <35 mg/dL, an insulin level ≥95 P, and an SBP or DBP level ≥95th P. ORs, based on separate logistic regression models for each risk factor category, were used to summarize the relation of overweight to the number of risk factors; the referent category consisted of persons without any risk factors. These analyses were restricted to 5477 (of the 9167) persons with measured levels of all risk factors; this sample is ∼97% of the eligible persons who participated in one of the last four cross-sectional examinations in Bogalusa. Insulin levels were not measured in the cross-sectional examinations until 1981.

    RESULTS

    Mean levels of various characteristics, as well as the proportion of persons with adverse risk factor levels, are shown in Table 2. The mean age of the 9167 schoolchildren was 11.9 years; 36% were black, 48% were female, and 10.8% (990) were considered overweight. Median levels of the Quetelet index and subscapular skinfold thickness varied substantially by age, but age was less strongly associated with the Rohrer index and triceps skinfold thickness. Overall, the proportion of persons with adverse lipid levels ranged from 7% TG to 10% TC, and the prevalence of several risk factors also varied with age. An HDLC level ≤35 mg/dL, for example, was seen among 6% of the 5- to 10-year-olds vs 11% of the 11- to 17-year-olds. High levels of insulin and blood pressure were defined as values > 95th (race-, sex-, and age-specific) P.

    View this table:
    Table 2.

    Description of Study Population*

    Table 3 shows the prevalence of the examined risk factors within seven categories of the Quetelet index among children (top) and adolescents (bottom). In general, the prevalences of most risk factors varied only slightly at Quetelet indices <75th or 85th P; this was particularly evident among the 5- to 10-year-olds. The prevalence of a high LDLC level, for example, increased from 8% to only 10% between the thinnest (<25th P) children and those with a Quetelet index between the 75th and 84th P, and the prevalence of adverse HDLC levels did not increase over these categories. Substantial increases in the prevalence of risk factors, however, were consistently seen at Quetelet index levels >85th P. Insulin levels showed the strongest association with the Quetelet index, with the prevalence of high levels increasing from 2% to 27% (5- to 10-year-olds) and from 1% to 25% (11- to 17-year-olds) across the seven categories; corresponding correlation coefficients were ∼0.50. Furthermore, the prevalence of all risk factors increased substantially between the 95th to 97th P and >97th P categories, indicating that even among overweight schoolchildren, higher levels of the Quetelet index were associated with increases in risk factors.

    View this table:
    Table 3.

    Relation of Quetelet Index to Adverse Risk Factors, by Age Group

    The relation of overweight to adverse risk factor levels is shown inTable 4. (With the exception of the final column in Table 4, subjects with a Quetelet index between the 85th P and 94th P have been excluded from subsequent analyses to more clearly assess associations with overweight, defined as a Quetelet index ≥95th P.) Although the sensitivity (23%–62%) and positive predictive value (9%–24%) of overweight was generally low for each risk factor, many of the ORs were substantial. Compared with other schoolchildren, an overweight youth was 2.4 times as likely to have a high TC level, and 7.1 times as likely to have a high TG level. In agreement with previous analyses, the largest ORs were seen with levels of insulin (OR = 12.6), and 170 of the 273 persons with an insulin level ≥95th P were also overweight (sensitivity = 62%). Although most associations tended to be stronger among whites than among blacks, the only significant differences (P < .01, as assessed by product terms) were for insulin and DBP. Observed ORs with DBP were >3 among whites, but were close to 1.0 among blacks; ORs with insulin were >15 (whites) versus ≤7.5 (blacks). Only small differences in the magnitudes of the associations were seen between boys and girls.

    View this table:
    Table 4.

    Estimated ORs Between Overweight (Quetelet Index >95 P) and Adverse Risk Factor Levels Among 5- to 17-Year-Olds

    ORs based on other characteristics to classify overweight are also shown in Table 4 (right columns). Rohrer index (kg/m3) and triceps skinfold thickness were dichotomized so that 10% of the sample with the highest (race-, sex-, and age-specific) levels would be contrasted with levels ≤80th P. (In previous analyses of the Quetelet index, 11% of the children were classified as overweight.) Estimated ORs based on the Rohrer index were similar to those for the Quetelet index, but associations with triceps skinfold thickness were generally weaker, particularly for levels of TG and insulin. Additional analyses also indicated that the triceps skinfold thickness provided very little additional information on risk factor levels if the Quetelet index was known. ORs contrasting Quetelet index levels between the 85th P and 94th P with lower levels are also shown in Table 4. The magnitudes of most associations were generally much smaller than those seen in analyses based on ≥95th P of the Quetelet index, and no association was seen with levels of DBP. ORs for levels of TC and LDLC, however, decreased only slightly.

    Although associations with overweight did not vary linearly with age, younger children showed the weakest associations with some risk factors (TC, LDLC, and insulin) and the strongest associations with others (TG, SBP, and DBP) (Table 5). As assessed by the addition of product terms to logistic regression models, associations with levels of LDLC, SBP, and DBP varied significantly (P < .01) by age. For levels of LDLC, ORs ranged from 0.7 (5- to 6-year-olds) to 6.6 (13- to 14-year-olds), and for SBP, from 16.0 (5- to 6-year-olds) to 2.5 (13- to 14-year-olds). Additional analyses indicated that within each race-sex group, overweight was most strongly related to SBP among younger children.

    View this table:
    Table 5.

    Estimated ORs* Between Overweight (Quetelet Index >95th P) and Adverse Risk Factor Levels, by Age

    We then examined the relation of overweight to risk factor clustering (Table 6). In both age groups, the proportion of children who were overweight increased as the number of risk factors increased. Among the 5- to 10-year-olds, for example, overweight children represented 7% (119 children) of the 1670 children with no risk factors, 22% (103 of 463) of the children with 1 risk factor, and 80% (32/40) of the children with ≥3 risk factors. Estimated ORs also indicated that compared with other children, overweight children were 9.7 times as likely to have 2 risk factors and 43.5 times as likely to have 3 risk factors. Comparable associations were also seen among the 11- to 17-year-olds, with 50 (71%) of the 71 children who had ≥3 risk factors being overweight. In general, risk factor clustering was less prevalent among black children than among whites, but the ORs between overweight and multiple risk factors were fairly similar in both groups (data not shown).

    View this table:
    Table 6.

    Relation of Overweight to Clustering of Adverse Risk Factor Levels

    Overweight was fairly effective in screening for any risk factor. Of the 302 overweight 5- to 10-year-olds, 183 (61%) had at least one risk factor, and the corresponding positive predictive value among 11- to 17-year-olds was 58% (295 of 511). Of the individual risk factors, overweight subjects were most likely to have high TG levels, but low levels of HDLC were also very prevalent among overweight 11- to 17-year-olds. It is interesting that 90% of the 80 children with high levels of both insulin and TG were also overweight.

    DISCUSSION

    The results of the current study indicate that overweight children and adolescents are at a substantially increased risk for adverse levels of several CVD risk factors. Of the 9167 schoolchildren in the current study, 11% were considered overweight with a Quetelet index ≥95th P of aggregated data from several national studies. Compared with children and adolescents with a Quetelet index <85th P, ORs among overweight schoolchildren ranged from 2.4 (TC and DBP) to 12.6 (insulin). There was little difference in the magnitude of these associations between boys and girls, but overweight was more strongly related to elevated levels of insulin and DBP among white than among black children, and some associations varied by age. Although the accuracy of overweight in identifying children and adolescents with adverse levels of individual risk factors was fairly low, it performed well as a screening tool for multiple risk factors. Of the overweight 5- to 10-year-olds, for example, 61% (183 of 302) had ≥1 of the 5 risk factors, and the comparable positive predictive value among 11- to 17-year-olds was 58% (295 of 511). Furthermore, of the 111 schoolchildren who had ≥3 risk factors, 82 (74%) were overweight.

    Although the underlying associations may be curvilinear,4 ,15 most studies of children and adolescents have used regression (or correlation) analyses to examine the relation of overweight to risk factor levels. In general, various measures of overweight and obesity have been found to be related weakly to levels of TC and LDLC (r ∼0.05–0.15),3–6 ,14 ,20 but more strongly to levels of HDLC and TG,3 ,4 ,6 ,20 blood pressure,5 ,6 ,14 ,18 ,20 and insulin.35 Although our results generally agree well these previous investigations, comparisons across studies can be influenced by differences in the 1) age and race distribution of the sample, 2) the specific weight-height index or skinfold thickness used, and 3) the statistical techniques used. Because levels of the Quetelet index increase with age among youth, for example, simple correlations may overestimate the magnitude of the associations. In the current study, adjustment for age reduced the correlation with DBP from 0.47 to 0.25.

    The study of overweight and risk factors as dichotomous characteristics, particularly when age-appropriate cutpoints have been recommended, has several advantages. ORs are frequently easier to interpret than are correlation or regression coefficients, and the positive predictive value would be useful in the evaluation of overweight children. Furthermore, it would be important to know the sensitivity of overweight for various risk factors when planning screening and intervention studies. Although relatively few studies of schoolchildren have presented the predictive value or sensitivity, ORs between triceps skinfold thickness (≥95th P) and elevated blood pressure levels among schoolchildren have ranged from 3.4 to 6.1 SBP, and from 1.9 to 5.3 DBP19; corresponding ORs in the current study were 4.6 SBP and 2.6 DBP. The ORs that we observed between overweight and high TC levels among 15- to 17-year-olds (OR = 2.8) were also fairly similar to those reported among young adults in National Health and Nutrition Examination Survey II36 (OR = 2.1), and between various quantiles of the Rohrer index and TC levels among adolescents.15

    Based in part on the relation of lipids to the early stages of atherosclerosis,8–10 LDLC screening has been recommended for children with a family history of hypercholesterolemia (parental TC ≥240 mg/dL) or premature CHD.23 Our results suggest that overweight should also be considered as an initial screening tool. It has been estimated22 that the current guidelines would trigger lipid testing among 25% of all children (80% of whom would be tested because of parental TC levels), and would identify 40% of those with an LDLC level ≥130 mg/dL. Additional estimates suggest that the use of a parental TC level of ≥300 mg/dL would achieve a sensitivity of 28% by screening 14% of all children,22 and our results (see Table 4) indicate that overweight would produce similar results. In additional analyses, we found that screening all children with a Quetelet index >85th P would yield a sensitivity of 43% by screening 23% of all schoolchildren, estimates that are nearly identical to those for the current lipid guidelines.

    Because of the reliance on reported (rather than recorded) parental TC levels, the recommended guidelines may be less accurate than expected. For example, a reported parental TC level ≥240 mg/dL has been found to result in 6% of children undergoing lipid screening, with a low sensitivity (6.5%) and positive predictive value (13.5%) for high LDLC levels.37 Compared with parental TC levels, it is likely that overweight would also be simpler to use as a screening trigger, and would more accurately identify adverse levels of other risk factors. In the current analyses, LDLC levels showed one of the weaker associations with overweight.

    Relatively little is known about the clustering of risk factors among children and adolescents, but previous analyses from the Bogalusa Heart Study showed that up to 6 times as many children and adolescents as expected had relatively high levels of the Quetelet index (or subscapular skinfold thickness) and various risk factors.16 ,17 Furthermore, a recent report of adolescents in Taiwan found that 55% (girls) to 70% (boys) of overweight children had ≥1 risk factors.18 Because of the large increase in fibrous plaques in the coronary arteries of children and young adults who have 3 or 4 CVD risk factors,9 the use of overweight as a screening tool for risk-factor clustering may be particularly effective in the early identification of persons likely to develop CVD.

    Several of the associations that we observed with overweight differed by race (DBP and insulin) or age (LDLC, SBP, and DBP). Many, but not all,38 investigators have also reported that relative weight is more strongly associated with levels of lipids and blood pressure5 ,15 ,36 ,39 among whites than among blacks. Previous reports have also noted that relative weight is most strongly related to lipid levels among older children,3 and to blood pressure levels among younger children.39 Although age differences in the relation of overweight to various risk factors have also been noted in studies of adults,36 ,38differences in the current study may reflect the statistical properties of the Quetelet index and the examined risk factors. For example, differences between the median and 95th P of the Quetelet index increased greatly with age (see Table 1), and the TG cutpoint of 130 mg/dL is the 97th P among blacks, but only the 89th P among whites.

    It is possible, however, that these contrasting associations reflect differences in the relation of the Quetelet index to body fat and fat distribution across race/ethnicity groups or age.40Correlations between the Quetelet index and more direct estimates of adiposity, such as dual radiograph absorptiometry, range from 0.50 to 0.80 among children and adolescents,41 and at comparable levels of the Quetelet index, white children have more total and intraabdominal body fat than do black children.42 We found, however, that if weight and height were already known, the triceps skinfold thickness provided little additional information on risk factor levels. Weight-height indices are also predictive of morbidity and mortality,43 are very reproducible, and appear to be as strongly correlated with risk factor levels as are densitometric estimates of fat mass.44 It is important, however, to be aware of the limitations of weight-height indices as surrogate measures for adiposity.

    The association between overweight and SBP among 5- to 6-year-olds was particularly striking (OR = 16), and it has been noted39 that this association is often strongest among younger children. Although the correlation coefficient that we observed with SBP levels in this age group (r = 0.43) is similar to other results,39 ,45 the estimated OR should be interpreted with caution as only 18 of the 913 children in this age group were overweight and had a high SBP (95% CI, 7.7–34). Although an inappropriately small bladder cuff can increase the recorded SBP of overweight children,46 the cuff was selected to be as large as possible during the examinations.25 ,33 ,47Furthermore, analyses that excluded observations made with a toddler-sized cuff yielded an OR of 14.

    The prevalence of overweight and obesity is increasing among youth,26 ,48 and our results demonstrate that even among 7- to 8-year-olds, overweight is consistently related to several CVD risk factors. Furthermore, as adults with multiple risk factors will likely be at high risk for CVD in later life,49 it is of particular interest that a large proportion of these persons be identified in early life through weight and height measurements. These findings emphasize the importance of the prevention and treatment of obesity in children and adolescents.

    ACKNOWLEDGMENTS

    This work was supported by Grant HL 15103 and Grant HL 32194 from the National Heart, Lung, and Blood Institute, National Institutes of Health; and by funds from the CDC, and the Robert W. Woodruff Foundation.

    Footnotes

      • Received August 31, 1998.
      • Accepted December 17, 1998.

    • Reprint requests to (D.S.F.) CDC Mailstop K-26, 4770 Buford Hwy, Atlanta, GA 30341-3717.

    REFERENCES

    1. ↵
      1. Berenson GS,
      2. Srinivasan SR,
      3. Wattigney WA,
      4. Harsha DW
      (1993) Obesity and cardiovascular risk in children. Ann N Y Acad Sci. 699:93–103.
      OpenUrlCrossRefPubMed
    2. ↵
      1. Williams DP,
      2. Going SB,
      3. Lohman TG,
      4. Harsha DW,
      5. Srinivasan SR,
      6. Webber LS,
      7. et al.
      (1992) Body fatness and risk for elevated blood pressure, total cholesterol, and serum lipoprotein ratios in children and adolescents. Am J Public Health. 82:358–363.
      OpenUrlCrossRefPubMed
    3. ↵
      1. Laskarzewski P,
      2. Morrison JA,
      3. Mellies MJ,
      4. et al.
      (1980) Relationships of measurements of body mass to plasma lipoprotein in schoolchildren and adults. Am J Epidemiol. 111:395–406.
      OpenUrlPubMed
    4. ↵
      1. Frerichs RR,
      2. Webber LS,
      3. Srinivasan SR,
      4. Berenson GS
      (1978) Relation of serum lipids and lipoproteins to obesity and sexual maturity in white and black children. Am J Epidemiol. 108:486–496.
      OpenUrlPubMed
    5. ↵
      1. Killeen J,
      2. Vanderburg, Harlan WR
      (1978) Application of weight-height ratios and body indices to juvenile populations: the National Health Examination Survey data. J Chron Dis. 31:529–537.
      OpenUrlCrossRefPubMed
    6. ↵
      1. Raitakari OT,
      2. Porkka KVK,
      3. Rönnemaa T,
      4. Åkerblom HK
      (1994) Clustering of risk factors for coronary heart disease in children and adolescents: the Cardiovascular Risk in Young Finns Study. Acta Paediatr. 83:935–940.
      OpenUrlPubMed
    7. ↵
      1. Freedman DS,
      2. Burke GL,
      3. Harsha DW,
      4. Srinivasan SR,
      5. Cresanta JL,
      6. Webber LS,
      7. Berenson GS
      (1985) Relationship of changes in obesity to serum lipid and lipoprotein changes in childhood and adolescence. JAMA. 254:515–520.
      OpenUrlCrossRefPubMed
    8. ↵
      1. Mahoney LT,
      2. Burns TL,
      3. Stanford W
      (1996) Coronary risk factors measured in childhood and young adult life are associated with coronary artery calcification in young adults: the Muscatine Study. J Am Coll Cardiol. 27:277–284.
      OpenUrlPubMed
    9. ↵
      1. Berenson GS,
      2. Srinivasan SR,
      3. Bao W,
      4. et al.
      (1998) Association between multiple cardiovascular risk factors and atherosclerosis in children and young adults. N Engl J Med. 338:1650–1656.
      OpenUrlCrossRefPubMed
    10. ↵
      1. McGill HC,
      2. McMahan A,
      3. Malcom GE,
      4. et al.
      (1995) Relation of glycohemoglobin and adiposity to atherosclerosis in youth. Arterioscler Thromb Vasc Biol. 15:431–440.
      OpenUrlAbstract/FREE Full Text
    11. ↵
      1. Power C,
      2. Lake JK,
      3. Cole TJ
      (1997) Measurement and long-term health risks of child and adolescent fatness. Int J Obes. 21:507–526.
      OpenUrlCrossRefPubMed
    12. ↵
      1. Must A,
      2. Jacques PF,
      3. Dallal GE,
      4. Bajema CJ,
      5. Dietz WH
      (1992) Long-term morbidity and mortality of overweight adolescents: a follow-up of the Harvard Growth Study of 1922 to 1935. N Engl J Med. 327:1350–1355.
      OpenUrlCrossRefPubMed
    13. ↵
      1. Tershakovec AM,
      2. Jawad SF,
      3. Stallings VA,
      4. Cortner JA,
      5. Zemel BS,
      6. Shannon BM
      (1998) Age-related changes in cardiovascular disease risk factors of hypercholesterolemic children. J Pediatr. 132:414–420.
      OpenUrlCrossRefPubMed
    14. ↵
      1. Lauer RM,
      2. Connor WE,
      3. Leaverton PE,
      4. Reiter MA,
      5. Clarke WR
      (1975) Coronary heart disease risk factors in school children: the Muscatine Study. J Pediatr 86:697–706.
      OpenUrlCrossRefPubMed
    15. ↵
      Resnicow K, Morabia A. The relation between body mass index and plasma total cholesterol in a multiracial sample of US schoolchildren. Am J Epidemiol. 1990;132;1083–1089
    16. ↵
      1. Smoak CG,
      2. Burke GL,
      3. Webber LS,
      4. Harsha DW,
      5. Srinivasan SR,
      6. Berenson GS
      (1987) Relation of obesity to clustering of cardiovascular disease risk factors in children and young adults: the Bogalusa Heart Study. Am J Epidemiol. 125:364–372.
      OpenUrlPubMed
    17. ↵
      1. Srinivasan SR,
      2. Bao W,
      3. Wattigney WA,
      4. Berenson GS
      (1996) Adolescent overweight is associated with adult overweight and related multiple cardiovascular risk factors: the Bogalusa Heart Study. Metabolism. 45:235–240.
      OpenUrlCrossRefPubMed
    18. ↵
      1. Chu NF,
      2. Rimm EB,
      3. Wang DJ,
      4. Liou HS,
      5. Shieh SM
      (1998) Clustering of cardiovascular disease risk factors among obese schoolchildren: the Taipei Children's Heart Study. Am J Clin Nutr. 67:1141–1146.
      OpenUrlAbstract
    19. ↵
      1. Gortmaker SL,
      2. Dietz WH,
      3. Sobol AM,
      4. Wehler CA
      (1987) Increasing pediatric obesity in the United States. Am J Dis Child. 141:535–540.
      OpenUrlCrossRefPubMed
    20. ↵
      1. Khoury PR,
      2. Morrison JA,
      3. Kelly K,
      4. Mellies M,
      5. Horvitz R,
      6. Glueck C
      (1980) Clustering and interrelationships of coronary heart disease risk factors in schoolchildren, ages 6–19. Am J Epidemiol. 112:524–538.
      OpenUrlPubMed
    21. ↵
      1. Himes JH,
      2. Dietz WH
      (1994) Guidelines for overweight in adolescent preventive services: recommendations from an expert committee. Am J Clin Nutr. 59:307–316.
      OpenUrlAbstract/FREE Full Text
    22. ↵
      National Cholesterol Education Program. Report of the Expert Panel on Blood Cholesterol Levels in Children and Adolescents. National Institutes of Health publication no. 91–2732. Washington, DC: US Department of Health and Human Services; September 1991
    23. ↵
      1. American Academy of Pediatrics, Committee on Nutrition
      (1998) Cholesterol in childhood. Pediatrics. 101:141–147.
      OpenUrlAbstract/FREE Full Text
    24. ↵
      1. National High Blood Pressure Education Program Working Group on Hypertension Control in Children and Adolescents
      (1996) Update on the 1987 Task Force report on high blood pressure in children and adolescents. Pediatrics. 98:649–658.
      OpenUrlAbstract/FREE Full Text
    25. ↵
      Berenson GS. Cardiovascular Risk Factors in Children. New York, NY: Oxford University Press; 1980:240–257
    26. ↵
      1. Freedman DS,
      2. Srinivasan SR,
      3. Valdez RA,
      4. Williamson DF,
      5. Berenson GS
      (1997) Secular increases in relative weight and adiposity among children over two decades: the Bogalusa Heart Study. Pediatrics. 99:420–425.
      OpenUrlAbstract/FREE Full Text
    27. ↵
      1. Cole TG
      (1990) The LMS method for constructing normalized growth standards. Eur J Clin Nutr. 44:45–60.
      OpenUrlPubMed
    28. ↵
      1. Mei Z,
      2. Yip R,
      3. Grummer-Strawn LM,
      4. Trowbridge FL
      (1998) Development of a research child growth reference and its comparison with the current international growth reference. Arch Pediatr Adolesc Med. 152:471–479.
      OpenUrlCrossRefPubMed
    29. ↵
      1. Kuczmarski RJ
      (1992) Prevalence of overweight and weight gain in the United States. Am J Clin Nutr. 55:495s–502s.
      OpenUrlPubMed
    30. ↵
      1. Must A,
      2. Dallal GE,
      3. Dietz WH
      (1991) Reference data for obesity: 85th and 95th percentiles of body mass index (wt/ht2)—a correction. Am J Clin Nutr. 54:773.
      OpenUrlFREE Full Text
    31. ↵
      1. Srinivasan SR,
      2. Frerichs RR,
      3. Webber LS,
      4. Berenson GS
      (1976) Serum lipoprotein profile in children from a biracial community: the Bogalusa Heart Study. Circulation. 54:309–318.
      OpenUrlAbstract/FREE Full Text
    32. ↵
      1. Kwiterovich PO Jr
      (1991) Plasma lipid and lipoprotein levels in childhood. Ann N Y Acad Sci. 623:90–107.
      OpenUrlCrossRefPubMed
    33. ↵
      1. Berenson GS,
      2. Cresanta JL,
      3. Webber LS
      (1984) High blood pressure in the young. Annu Rev Med. 35:535–560.
      OpenUrlCrossRefPubMed
    34. ↵
      1. Rosner B,
      2. Prineas J,
      3. Loggie MH,
      4. Daniels SR
      (1993) Blood pressure nomograms for children and adolescents, by height, sex, and age, in the United States. J Pediatr. 123:871–886.
      OpenUrlCrossRefPubMed
    35. ↵
      1. Freedman DS,
      2. Srinivasan SR,
      3. Burke GL,
      4. et al.
      (1987) Relation of body fat distribution to hyperinsulinemia in children and adolescents: the Bogalusa Heart Study. Am J Clin Nutr. 46:403–410.
      OpenUrlAbstract/FREE Full Text
    36. ↵
      Van Itallie TB. Health implications of overweight and obesity in the United States. Ann Intern Med. 1985;103(suppl 6, pt 2):983–988
    37. ↵
      1. Resnicow K,
      2. Cross D
      (1993) Are parents' self-reported total cholesterol levels useful in identifying children with hyperlipidemia? An examination of current guidelines. Pediatrics. 92:347–353.
      OpenUrlAbstract/FREE Full Text
    38. ↵
      1. Folsom AR,
      2. Burke GL,
      3. Byers CL,
      4. et al.
      (1991) Implications of obesity for cardiovascular disease in blacks: the CARDIA and ARIC studies. Am J Clin Nutr. 53:1604S–1611S.
      OpenUrlAbstract/FREE Full Text
    39. ↵
      1. Resnicow K,
      2. Futterman R,
      3. Vaughan RD
      (1993) Body mass index as a predictor of systolic blood pressure in a multiracial sample of US schoolchildren. Ethn Dis. 3:351–361.
      OpenUrlPubMed
    40. ↵
      1. Goran MI
      (1998) Measurement issues related to studies of childhood obesity: assessment of body composition, body fat distribution, physical activity, and food intake. Pediatrics. 101:505–518.
      OpenUrlCrossRefPubMed
    41. ↵
      1. Pietrobelli A,
      2. Faith MS,
      3. Allison DB,
      4. Gallagher D,
      5. Chiumello G,
      6. Heymsfeld SB
      (1998) Body mass index as a measure of adiposity among children and adolescents: a validation study. J Pediatr. 132:204–210.
      OpenUrlCrossRefPubMed
    42. ↵
      1. Daniels SR,
      2. Khoury PR,
      3. Morrison JA
      (1997) The utility of body mass index as a measure of body fatness in children and adolescents: differences by race and gender. Pediatrics. 99:804–807.
      OpenUrlAbstract/FREE Full Text
    43. ↵
      1. Dietz WH,
      2. Robinson TN
      (1998) Use of body mass index as a measure of overweight in children and adolescents. J Pediatr. 132:191–193.
      OpenUrlCrossRefPubMed
    44. ↵
      1. Spiegelman D,
      2. Israel RG,
      3. Bouchard C,
      4. Willett WC
      (1992) Absolute fat mass, percent body fat, and body-fat distribution: which is the real determinant of blood pressure and serum glucose? Am J Clin Nutr. 55:1033–1044.
      OpenUrlAbstract/FREE Full Text
    45. ↵
      1. Gutin B,
      2. Basch C,
      3. Shea S,
      4. et al.
      (1990) Blood pressure, fitness, and fatness in 5- and 6-year-old children. JAMA. 264:1123–1127.
      OpenUrlCrossRefPubMed
    46. ↵
      1. Thomas PW,
      2. Peters TJ,
      3. Golding J,
      4. Haslum MN
      (1989) Height, weight, and blood pressures in ten-year-old children. Hum Biol. 61:213–225.
      OpenUrlPubMed
    47. ↵
      1. Voors AW
      (1975) Cuff bladder size in a blood pressure survey of children. Am J Epidemiol. 101:489–494.
      OpenUrlPubMed
    48. ↵
      1. Troiano RP,
      2. Flegal KM,
      3. Kuczmarski RJ,
      4. Campbell SM,
      5. Johnson CL
      (1995) Overweight prevalence and trends for children and adolescents: the National Health and Nutrition Examination Surveys, 1963 to 1991. Arch Pediatr Adolesc Med. 149:1085–1091.
      OpenUrlCrossRefPubMed
    49. ↵
      1. Genest J Jr,
      2. Cohn JS
      (1995) Clustering of cardiovascular risk factors: targeting high-risk individuals. Am J Cardiol. 76:8A–20A.
      OpenUrlCrossRefPubMed
    View Abstract
    Back to top

    In this issue

    Pediatrics
    Vol. 103, Issue 6
    1 Jun 1999
    View this article with LENS
    Email Article
    Request Permissions
    Article Alerts
    Citation Tools
    The Relation of Overweight to Cardiovascular Risk Factors Among Children and Adolescents: The Bogalusa Heart Study
    David S. Freedman, William H. Dietz, Sathanur R. Srinivasan, Gerald S. Berenson
    Pediatrics Jun 1999, 103 (6) 1175-1182; DOI: 10.1542/peds.103.6.1175
    del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo
    Print
    Download PDF
    Insight Alerts

    Jump to section

    Related Articles

    • No related articles found.

    Cited By...

    More in this TOC Section

    Similar Articles

    Subjects