METHODS: Data were derived from the Global School-Based Student Health Survey (GSHS). Data from 71176 adolescents aged 12-15 years residing in 23 countries were analyzed. The Centers for Disease Control and Prevention (CDC) 2000 growth charts were used to identify underweight, normal weight, and overweight/ obesity. Weighted age- and gender-adjusted prevalence of weight categories and tobacco use was calculated. Multivariate logistic regression analysis was performed to estimate the association between weight categories and tobacco use for each country, controlling for covariates. Pooled odds ratios and confidence intervals were computed using random- or fixed-effects meta-analyses.
RESULTS: A significant association between weight categories and tobacco use was evident in only a few countries. Adolescents reporting tobacco use in French Polynesia, Suriname, and Indonesia, had 72% (95% CI: 0.15-0.56), 55% (95% CI: 0.24-0.84), and 24% (95% CI: 0.61-0.94) lower odds of being underweight, respectively. Adolescents reporting tobacco use in Uganda, Algeria, and Namibia, had 2.30 (95% CI: 1.04-5.09), 1.71 (95% CI: 1.25-2.34), and 1.45 (95% CI: 1.00-2.12) times greater odds of being overweight/obese, but those in Indonesia and Malaysia had 33% (95% CI: 0.50-0.91) and 16% (95% CI: 0.73-0.98) lower odds of being overweight/obese.
CONCLUSIONS: The association between tobacco use and BMI categories is likely to be different among adolescents versus adults. Associating tobacco use with being thin may be more myth than fact and should be emphasized in tobacco prevention programs targeting adolescents.
Methods: Analyses were performed on 243 women (mean body mass index 31.27 ± 4.14 kg/m2) who completed a 12-month lifestyle intervention in low socioeconomic communities in Klang Valley, Malaysia. Analysis of covariance (ANCOVA) was used to compare changes of cardiometabolic risk factors across weight change categories (2% gain, ±2% maintain, >2 to <5% loss, and 5 to 20% loss) within intervention and control group.
Results: A graded association for changes in waist circumference, fasting insulin, and total cholesterol (p=0.002, for all variables) across the weight change categories were observed within the intervention group at six months postintervention. Participants who lost 5 to 20% of weight had the greatest improvements in those risk markers (-5.67 cm CI: -7.98 to -3.36, -4.27 μU/mL CI: -7.35, -1.19, and -0.59 mmol/L CI: -.99, -0.19, respectively) compared to those who did not. Those who lost >2% to <5% weight reduced more waist circumference (-4.24 cm CI: -5.44 to -3.04) and fasting insulin (-0.36 μU/mL CI: -1.95 to 1.24) than those who maintained or gained weight. No significant association was detected in changes of risk markers across the weight change categories within the control group except for waist circumference and adiponectin.
Conclusion: Weight loss of >2 to <5% obtained through lifestyle intervention may represent a reasonable initial weight loss target for women in the low socioeconomic community as it led to improvements in selected risk markers, particularly of diabetes risk.