The effects of weight loss by exercise or by dieting on plasma high-density lipoprotein (HDL) levels in men with low, intermediate, and normal-to-high HDL at baseline

Abstract

To assess whether baseline plasma high-density lipoprotein (HDL) cholesterol levels affected the HDL response to weight loss, we examined lipoprotein changes in overweight men aged 30 to 59 years who were randomized to lose weight by exercise training (primarily running, n = 46) or by caloric restriction (ie, dieting, n = 42) or to remain sedentary, nondieting controls (n = 42) in a 1-year study. In exercisers, absolute increases in HDL (mg/dL) were greatest in men with normal-to-high baseline HDL and least in men with low baseline HDL. Specifically, when divided into groups of low (< or = 37 mg/dL), intermediate (38 to 47 mg/dL), and normal-to-high HDL cholesterol (> or = 48 mg/dL) at baseline, the exercisers increased HDL cholesterol by 2.3 +/- 1.9, 4.9 +/- 1.1, and 7.0 +/- 1.3 mg/dL, respectively; HDL2 cholesterol by 0.8 +/- 1.6, 2.3 +/- 1.2, and 5.1 +/- 1.3 mg/dL; and HDL2 mass by 2.8 +/- 5.1, 9.5 +/- 8.9, and 31.7 +/- 11.0 mg/dL. Relative increases in HDL cholesterol were more similar in the low (7.1% +/- 6.1%), intermediate (12.4% +/- 3.9%), and normal-to-high men (13.2% +/- 4.0%). Regression analyses were performed to assess whether baseline HDL cholesterol was related to the amount of absolute HDL change per unit of weight loss. In exercisers, the increase in HDL3 cholesterol concentrations was significantly greater in men with low HDL than in those with normal-to-high HDL at entry (2.0 +/- 0.8 v 0.2 +/- 0.8 mg/dL per kg/m2 lost).(ABSTRACT TRUNCATED AT 250 WORDS)

Fig 1

Relationship between baseline HDL cholesterol levels and l-year change in (A) HDL cholesterol, (B) HDL₂ cholesterol, and (C) HDL₂ mass in exercisers (squares) and controls (diamonds). Significance levels are given for the difference in the regression slopes. Independent HDL cholesterol measurements are used for baseline HDL and ΔHDL. Baseline HDL cholesterol was not significantly related to changes in the exercisers' HDL₃ cholesterol (β_E − β_C = −0.01 ± 0.08, P = 0.88) and HDL₃ mass (β_E − β_C = 0.11 ± 0.67, P = 0.88: results not displayed).

Fig 2

Effects of weight loss (body mass index) on HDL cholesterol, HDL₂ cholesterol, HDL₃ cholesterol, and HDL₃ mass in exercisers with initially low (diamonds) and normal-to-high HDL (squares). Significance levels are given for the difference in the regression slopes between the exercisers with low and normal-to-high HDL at entry. The regression slopes for exercisers with intermediate HDL at entry fell between the low and the normal-to-high HDL men (not displayed, see text).

Fig 3

Relationship between baseline HDL cholesterol levels and l-year change in HDL cholesterol in dieters (squares) and controls (diamonds). Significance level is given for the difference in the regression slopes. Independent HDL cholesterol measurements are used for baseline HDL and ΔHDL.