RESUMO
BACKGROUND: While heavier weight is known to increase the incidence of dyslipidemia, limited data are available on the relationship between weight gain and its development. METHODS: A total of 2647 males were categorized into the following four groups according to the difference between their self-reported weight at 20 years of age and their measured weight in 1994-95: a loss of ≥5% (decrease), loss of <5% or gain of <5% (no change), gain of ≥5 to <15% (increase) and gain of ≥15% (sizable increase). They were followed up until their 2002-03 health examination. Using the 'no change' group as reference, the multivariable-adjusted odds ratio (adjusted for age, body mass index at 20 years of age, physical activity, smoking and alcohol intake) and 95% confidence interval (95% CI) for the incidence of dyslipidemia were determined using logistic regression models. RESULTS: A total of 1342 participants developed dyslipidemia during the follow-up period. The 'increase' and 'sizable increase' groups had odds ratios for the incidence of dyslipidemia of 1.97 (95% CI, 1.59-2.45) and 2.68 (2.15-3.34), respectively, demonstrating that there was a significant dose-response association between weight gain since 20 years of age and the incidence of dyslipidemia (P < 0.001 for trend). CONCLUSION: These results suggest that dyslipidemia could be prevented by avoiding weight gain in adulthood.