Self-reported substantial 1-year weight change among men and women in the United States.

Population-based data have not been readily available on relatively short-term changes in weight. Therefore, we sought to determine the nature of self-reported substantial (> 10%) weight change over one year in a representative sample of the US population which participated in the 1989 National Health Interview Survey (NHIS). Across all ages, a larger proportion of women than men reported both weight loss as well as weight gain of any amount (18.9% vs. 16.1% for weight loss and 20.0% vs. 16.1% for weight gain). In sex-specific logistic regression analyses, significant risk factors common to both sexes for substantial weight loss included divorced/separated marital status, smoking, increased number of blood pressure checks, increased BMI (body mass index) and increased number of bed days. Black race reduced the risk of weight loss for both men and women. Sex-specific risk factors for weight loss in men only were widowhood or never married marital status, while increasing age was a protective factor in women only. Concerning weight gain > 10% over the past year, increased number of blood pressure checks and having one or more diabetic parents were significant risk factors among both men and women; while never being married, increased age, BMI, and education exerted a protective effect in both sexes. For women only, risk factors for weight gain included black race, increased number of contacts with a health professional, and being unemployed. Intention to lose weight was associated with both weight gain and weight loss in both sexes, although it did not serve as a confounder in any of these relationships. A greater likelihood of substantial weight loss among women relative to men was diminished for persons with higher BMI, higher number of blood pressure checks, being widowed, divorced or separated, and intention to lose weight. A greater likelihood of substantial weight gain among women relative to men was diminished for persons with low BMI. The results of this cross-sectional study of weight change, involving a one-year follow-up period, generally correspond with the results obtained by longitudinal studies involving a longer follow-up.