Diabetes & Metabolism Journal (original) (raw)

Fig. 1 Association between change in alcohol consumption and change in waist circumference or fasting serum glucose. Adjusted mean values of change in (A) waist circumference or (B) fasting serum glucose were calculated for initial non-drinkers, light drinkers, moderate drinkers, and heavy drinkers according to follow-up alcohol consumption: non-drinkers (0.0 g/day), light drinkers (0.1 to 19.9 g/day for males, 0.1 to 9.9 g/day for females), moderate drinkers (20.0 to 39.9 g/day for males, 10.0 to 19.9 g/day for females), heavy drinkers (≥40.0 g/day for males, ≥20.0 g/day for females). Adjusted mean values determined by linear regression analysis after adjustments for age, sex, household income, education level, physical activity, smoking, total energy intake, body mass index (excluded for analyses on change in waist circumference), past history of cardiovascular disease, family history of hypertension, diabetes, and dyslipidemia, and follow-up duration. Groups without change in alcohol consumption were considered reference groups (e.g., initial non-drinkers to follow-up non-drinkers). P values for statistical significance: aP<0.05, bP<0.01, cP<0.001.

Fig. 2 Association between change in alcohol consumption and change systolic or diastolic blood pressure. Adjusted mean values of change in (A) systolic or (B) diastolic blood pressure were calculated for initial non-drinkers, light drinkers, moderate drinkers, and heavy drinkers according to follow-up alcohol consumption: non-drinkers (0.0 g/day), light drinkers (0.1 to 19.9 g/day for males, 0.1 to 9.9 g/day for females), moderate drinkers (20.0 to 39.9 g/day for males, 10.0 to 19.9 g/day for females), heavy drinkers (≥40.0 g/day for males, ≥20.0 g/day for females). Adjusted mean values determined by linear regression analysis after adjustments for age, sex, household income, education level, physical activity, smoking, total energy intake, body mass index, past history of cardiovascular disease, family history of hypertension, diabetes, and dyslipidemia, and follow-up duration. Groups without change in alcohol consumption were considered reference groups (e.g., initial non-drinkers to follow-up non-drinkers). P values for statistical significance: aP<0.05, bP<0.01, cP<0.001.

Fig. 3 Association between change in alcohol consumption and change triglycerides or high density lipoprotein cholesterol (HDL-C). Adjusted mean values of change in (A) triglycerides or (B) HDL-C were calculated for initial non-drinkers, light drinkers, moderate drinkers, and heavy drinkers according to follow-up alcohol consumption: non-drinkers (0.0 g/day), light drinkers (0.1 to 19.9 g/day for males, 0.1 to 9.9 g/day for females), moderate drinkers (20.0 to 39.9 g/day for males, 10.0–19.9 g/day for females), heavy drinkers (≥40.0 g/day for males, ≥20.0 g/day for females). Adjusted mean values determined by linear regression analysis after adjustments for age, sex, household income, education level, physical activity, smoking, total energy intake, body mass index, past history of cardiovascular disease, family history of hypertension, diabetes, and dyslipidemia, and follow-up duration. Groups without change in alcohol consumption were considered reference groups (e.g., initial non-drinkers to follow-up non-drinkers). P values for statistical significance: aP<0.05, bP<0.01, cP<0.001.