A 4-year prospective study of soft drink consumption and weight gain: the role of calorie intake and physical activity

Abstract

Purpose: Examine the association between soft drink consumption and risk of weight gain over 4 years. Also, determine if the relationship between soft drink intake and weight gain is a result of differences in calorie intake or physical activity (PA), or other potential mediating factors.

Design: Four-year prospective cohort.

Setting: Approximately 20 cities in Utah and Wyoming.

Subjects: One hundred seventy nonsmoking, apparently healthy women.

Measures: At baseline, soft drink consumption and menopause status were measured using a questionnaire; calorie intake was estimated using 7-day, weighed food records; and PA was assessed using 7-day accelerometer data. Weight was assessed using an electronic scale at baseline and follow-up.

Analysis: Multiple regression and partial correlation.

Results: Four-year weight gain in participants who, at baseline, consumed sugar-sweetened soft drinks (2.7 ± 5.1 kg) was greater than in participants who consumed artificially sweetened soft drinks (-.1 ± 4.4 kg) or no soft drinks (.5 ± 5.1 kg) (F = 5.4, p = .022). Adjusting for objectively measured PA had no effect on risk of weight gain. However, controlling statistically for differences in calorie intake significantly weakened the relationship between soft drink consumption and weight gain by 28%.

Conclusion: Consuming artificially sweetened soft drinks or no soft drinks instead of sugar-sweetened soft drinks may help to reduce risk of weight gain in women. The relationship appears to be partly a function of differences in calorie intake, but not differences in PA.

Keywords: Beverage; Body Weight; Health focus: weight control; Manuscript format: research; Obesity; Outcome measure: biometric; Overweight; Prevention Research; Research purpose: relationship testing; Setting: community; Soft Drink; Strategy: behavior change; Study design: prospective cohort; Sugar; Target population age: adults; Target population circumstances: education/income level, race/ethnicity.