Results
After excluding 4143 individuals missing data on any of the parameters used to define the metabolic syndrome, 318 individuals who had a BMI<18.5 kg/m, 3536 individuals who did not fast or had missing data on fasting status prior to collection of the blood samples, 319 individuals lost to follow-up, and 83 individuals who were receiving hemodialysis at the time of study enrollment, 21,840 participants were included in the final analyzed sample.
Table 1 depicts the baseline characteristics of the study sample by categories of BMI and metabolic health. As compared with individuals with normal weight, individuals in the overweight and obese categories were more likely to be younger, black, have a history of hypertension, diabetes, coronary heart disease (CHD), and stroke, have lower socioeconomic status, have albuminuria (urine albumin to creatinine ratio (ACR)≥30 mg/g), and have a lower estimated glomerular filtration rate (eGFR) at baseline. Participants with the metabolic syndrome were more likely to be female, black, have lower socioeconomic status, and have a history of hypertension, diabetes, CHD, and stroke at baseline as compared with individuals without the metabolic syndrome. Those with the metabolic syndrome were also more likely to have albuminuria and lower eGFR compared with those without the metabolic syndrome.
Associations of BMI and Metabolic Health With ESRD
A total of 247 participants developed ESRD over a mean 6.3±1.3 years of follow-up. Metabolic health modified the association of BMI with ESRD in a Cox regression model adjusted for age, race, sex, geographic region of residence, sociodemographic status, lifestyle factors, and comorbidities (Pinteraction=0.01); therefore, we stratified all subsequent analyses by the presence or absence of the metabolic syndrome. The association of BMI and ESRD was not modified by age, gender, or race (P for interaction >0.10 for all). When stratified by the presence or absence of the metabolic syndrome, among participants without the metabolic syndrome, higher BMI was associated with lower risk of ESRD in models adjusted for age, race, sex, and geographic region of residence (hazard ratio (HR) per 5 kg/m increase in BMI 0.67, 95% confidence interval (CI) 0.49, 0.91) and in models further adjusted for education, income, physical activity, cigarette smoking, systolic blood pressure, and a history of CHD and stroke (HR per 5 kg/m increase in BMI 0.70, 95% CI 0.52, 0.95). Among participants with the metabolic syndrome, there were no statistically significant associations of increasing BMI with risk of incident ESRD (HR per 5 kg/m increase in BMI 1.06, 95% CI 0.93, 1.21).
Figure 1 depicts the association of BMI with ESRD risk in the full study sample and stratified by the absence or presence of the metabolic syndrome. Increasing BMI was associated with increased risk of developing ESRD in the full study sample (Figure 1a). However, when stratified by the absence or presence of the metabolic syndrome, increasing BMI was associated with a lower risk of developing ESRD in individuals without the metabolic syndrome (Figure 1b), whereas no association of BMI with risk of ESRD was observed in individuals with the metabolic syndrome (Figure 1c).
(Enlarge Image)
Figure 1.
Associations of body mass index with risk of end-stage renal disease. Hazard ratios for incident end-stage renal disease (ESRD) as a function of body mass index (BMI) in the full study sample (a) and stratified by the absence (b) or presence (c) of the metabolic syndrome. BMI was modeled as a continuous variable and fitted in a Cox proportional hazard model using restricted quadratic spline regression adjusted for age, race, sex, geographic region of residence, education, income, physical activity, current smoking, history of coronary heart disease, and history of stroke. Knots for the spline were placed at a BMI of 25 and 30 kg/m, and the reference point was a BMI of 22 kg/m. Dashed horizontal lines correspond to reference values. Shaded areas represent 95% confidence intervals for hazard ratios. Histograms present distributions of BMI in study participants.
Table 2 depicts the incidence rates for ESRD per 1000 person-years of follow-up by categories of weight and metabolic syndrome status. ESRD incidence rates were higher in those with the metabolic syndrome as compared with those without the metabolic syndrome within each weight category. Figure 2 reports the HRs for ESRD as a function of weight and metabolic syndrome categories, with individuals who were normal weight and without the metabolic syndrome serving as the referent group. In models adjusted for age, race, sex, geographic region of residence, educational achievement, annual family income, physical activity, cigarette smoking, and a history of CHD and stroke, the HR for ESRD was higher in overweight and obese participants with the metabolic syndrome than in normal weight participants without the metabolic syndrome (HR 2.03 95% CI, 1.26, 3.17 and HR 2.29 95% CI 1.51, 3.48, respectively). Among all the individual metabolic risk factors, higher triglycerides, higher blood pressure, and higher fasting glucose were significantly associated with higher risk of ESRD, with the magnitude of the association being the greatest for higher blood pressure and higher fasting glucose (data not shown). In contrast, when comparing overweight or obese without the metabolic syndrome to normal weight participants without the metabolic syndrome, the HRs for ESRD were 0.65 (95% CI 0.39, 1.11) and 0.47 (95% CI 0.23, 0.95), respectively.
(Enlarge Image)
Figure 2.
Hazard ratios (95% confidence intervals) for incident end-stage renal disease (ESRD) among Reasons for Geographic and Racial Differences in Stroke (REGARDS) Study participants by weight and metabolic subtype categories. Model was adjusted for age, race, sex, geographic region of residence, education, income, physical activity, current smoking, history of coronary heart disease, and history of stroke. Ptrend for the association of weight categories (normal, overweight, obese) with ESRD risk in participants without the metabolic syndrome was 0.02; Ptrend for the association of weight categories (normal, overweight, obese) with ESRD risk in participants with the metabolic syndrome was 0.16.