Demographic features and the prevalence of the cardiovascular risk factors for the overall study population are presented in . Overall, 10.3% of the study population was overweight. Metabolic syndrome was detected for 3.6% of the study population. Microalbuminuria was detected for 8.9% of the study population.
Population Characteristics of Study Sample and Projected US Population of Adolescents 12 to 19 Years of Age (NHANES, 1999–2004)
The median albumin/creatinine ratio for girls was significantly greater than that for boys (7.9 vs 4.9 mg/g; P < .0001). The median albumin/creatinine ratio for boys 12 to 15 years of age did not differ significantly from that for boys 16 to 19 years of age (5.0 vs 4.6 mg/g; P = .09). The median albumin/creatinine ratio for girls 12 to 15 years of age did not differ significantly from that for girls 16 to 19 years of age (8.2 mg/g vs 7.6 mg/g; P = .15).
The median urinary albumin concentration for girls was significantly higher than that for boys (10.3 vs 7.7 μg/mL; P < .0001). Boys 16 to 19 years of age had a higher median urinary albumin concentration than did boys 12 to 15 years of age (8.8 vs 6.9 μg/mL; P = .03). The median urinary albumin concentration for girls 12 to 15 years of age did not differ significantly from that for girls 16 to 19 years of age (9.8 vs 10.9 μg/mL; P = .16).
Nonoverweight and overweight adolescents did not differ with respect to age (P = .31) or gender (P = .22). The prevalence of overweight among all non-Hispanic white adolescents was 8.78%, whereas the prevalence of overweight was 14.8% for non-Hispanic black adolescents, 13.14% for Hispanic adolescents, and 8.24% for adolescents of other races.
Cardiovascular risk factors stratified according to BMI category are described in . As anticipated, the prevalence of adverse cardiovascular risk factors was greater among overweight adolescents, compared with nonover-weight adolescents. Specifically, overweight adolescents had greater prevalence rates of abdominal obesity (96.6% vs 9.2%; P < .0001), impaired fasting glucose (16.5% vs 9.1%; P = .04), insulin resistance (62.5% vs 9.4%; P < .0001), hypertension (10.7% vs 2%; P < .0001), high triglyceride levels (26.4% vs 9.5%; P < .0001), low HDL cholesterol levels (24.7% vs 7.3%; P < .0001), and metabolic syndrome (22.0% vs 1.5%; P < .0001), compared with nonoverweight adolescents. Diabetes mellitus was more prevalent among overweight adolescents than among nonoverweight adolescents, but the difference was not statistically significant (1.7% vs 0.7%; P = .29). Fewer overweight adolescents smoked, compared with nonover-weight adolescents, but the difference was not statistically significant (11.8% vs 15.2%; P = .31).
FIGURE 1 Prevalence of microalbuminuria and cardiovascular risk factors in adolescents according to BMI category. MA indicates microalbuminuria; TG, triglyceride level; LDL, low-density lipoprotein level; IFG, impaired fasting glucose level; IR, insulin resistance; (more ...)
Factors Associated With Microalbuminuria
Despite the clustering of adverse cardiovascular risk factors among overweight adolescents, microalbuminuria was more prevalent among nonoverweight adolescents than among overweight adolescents (8.7% vs 0.3%; P = .005) (). The median albumin/creatinine ratio decreased with increasing BMI z scores for boys (overall test of trend, P = .03) and girls (overall test of trend, P < .0001) ().
FIGURE 2 Box plot of ACR of adolescents according to gender and BMI z-score quartiles. Shown are the 25th, 50th, and 75th percentiles of ACR distribution for boys and girls without microalbuminuria according to age group. The boxes enclose the 25th to 75th percentile, (more ...)
Similarly, microalbuminuria was more prevalent among adolescents without abdominal obesity, compared with adolescents with abdominal obesity (7.9% vs 1.0%; P = .03), and among adolescents without insulin resistance, compared with adolescents with insulin resistance (7.8% vs 0.7%; P = .02). Similar results were seen when urinary albumin concentration alone (urinary albumin concentration of ≥30 μg/mL) was used to define microalbuminuria (data not shown).
In classification of all adolescents according to fasting glucose levels and albuminuria status, 81.4% had normal fasting glucose and normal albuminuria, 7.9% had normal fasting glucose and microalbuminuria, 9.3% had impaired fasting glucose and normal albuminuria, 0.5% had impaired fasting glucose and microalbuminuria, 0.3% had diabetes mellitus and normal albuminuria, and 0.5% had diabetes mellitus and microalbuminuria. In analyses of the entire sample, hypertension, high triglyceride levels, low HDL cholesterol levels, smoking, and the metabolic syndrome were not significantly associated with microalbuminuria. Significantly more girls than boys had microalbuminuria (5.8% vs 3.1%; P = .0002). The prevalence of microalbuminuria did not differ with respect to race/ethnicity (P = .85) or age (P = .15).
Modification of Associations Between Cardiovascular Risk Factors and Microalbuminuria by Being Overweight
Our main results were then stratified according to the presence or absence of overweight (BMI of ≥95th percentile) and were adjusted for age, gender, and race/ethnicity. Among nonoverweight adolescents, microalbuminuria was not associated with any cardiovascular disease risk factor except for overt diabetes mellitus ().
Unadjusted and Adjusted ORs for Microalbuminuria in Adolescents, Stratified According to BMI Category
A different picture emerged among overweight adolescents. Among overweight adolescents, in addition to diabetes mellitus (adjusted OR: 1373; 95% confidence interval [CI]: 15–123 757), cardiovascular risk factors such as impaired fasting glucose (adjusted OR: 13; 95% CI: 1.51–118), insulin resistance (adjusted OR: 75; 95% CI: 1.58–3573), hypertension (adjusted OR: 11; 95% CI: 2.47–53), and smoking (adjusted OR: 25; 95% CI: 1.29–484) were associated with microalbuminuria (). Furthermore, other cardiovascular risk factors examined (high triglyceride levels, low HDL cholesterol levels, and the metabolic syndrome) were also associated with an increased risk of microalbuminuria, although these associations did not reach conventional levels of statistical significance ().
Among nonoverweight adolescents, girls were more likely than boys to have microalbuminuria (OR: 2.07; 95% CI: 1.45–2.96). However, among overweight adolescents, this association was attenuated and no longer statistically significantly (OR: 1.71; 95% CI: 0.19–15.34). Neither race/ethnicity nor age was significantly associated with microalbuminuria among nonoverweight or overweight adolescents.
We observed similar associations between the cardiovascular risk factors and microalbuminuria when we used alternative definitions for the metabolic syndrome and its individual criteria, as published by Cook et al,40
de Ferranti et al,41
and Invitti et al23
(which changed the prevalence of the metabolic syndrome to 6.3%, 12.9%, and 6.7%, respectively). Similar results were also observed when we included adolescents with macroalbuminuria and when we excluded adolescents with diabetes mellitus (except that hypertension could not be assessed, because all adolescents with hypertension and microalbuminuria also had diabetes mellitus).
We also used urinary albumin concentrations of ≥30 μg/mL as the outcome, to avoid any potential confounding attributable to differences in generation of creatinine among overweight and nonoverweight adolescents; 15.1% of adolescents had urinary albumin concentrations of ≥30 μg/mL. Urinary albumin concentrations of ≥30 μg/mL were more prevalent among nonoverweight adolescents than among overweight adolescents (14.6% vs 0.5%; P = .0001). We observed associations between the cardiovascular risk factors and urinary albumin concentrations of ≥30 μg/mL similar to those our original analysis (data not shown).