Purpose: The high prevalence of obesity in Western societies has been attributed in part to high-fat low-CHO food consumption. However, people have also become less active, and inactivity may have increased the risk for weight gain from poor dietary choices. Analyses were performed to test whether diet-weight relationships were attenuated by vigorous exercise.
Methods: Age- and education-adjusted cross-sectional regression analyses of 62,042 men and 44,695 women recruited for the National Runners' Health Study were conducted. Reported meat and fruit intakes were analyzed separately and as indicators of high-risk diets.
Results: The runners were generally lean (mean ± SD: males = 24.15 ± 2.81 kg·m−2, females = 21.63 ± 2.70 kg·m−2) as measured by body mass index (BMI), educated (males = 16.42 ± 2.47 yr, females = 16.04 ± 2.32 yr), and middle-aged (males = 44.40 ± 10.83 yr, females = 38.21 ± 10.08 yr), who ran 5.30 ± 3.23 km·d−1 if male and 4.79 ± 3.00 km·d−1 if female. Running significantly attenuated BMI's relationship to reported meat and fruit intakes in men (P < 10−8 and P < 10−12, respectively) and women (P < 10−15 and P < 10−6, respectively). Specifically, compared with running <2 km·d−1, running >8 km·d−1 reduced the apparent BMI increase per serving of meat by 43% in men (slope ± SE = from 0.74 ± 0.10 to 0.42 ± 0.06) and 55% in women (from 1.26 ± 0.13 to 0.57 ± 0.09) and reduced the apparent BMI reduction per serving of fruit by 75% in men (from −0.28 ± 0.04 to −0.07 ± 0.02) and 94% in women (from −0.16 ± 0.05 to −0.01 ± 0.02). Running also significantly attenuated the concordant relationship between reported meat intake and waist and chest circumferences in men (P < 10−9 and P = 0.0002, respectively) and women (P = 0.0004 and P < 10−5, respectively) and the concordant relationship between meat intake and hip circumference in women (P < 10−6).
Conclusions: Vigorous exercise may mitigate diet-induced weight gain, albeit not guaranteeing protection from poor dietary choices.
Lawrence Berkeley National Laboratory, Berkeley, CA
Address for correspondence: Paul T. Williams, Ph.D., Lawrence Berkeley National Laboratory, Donner 464, 1 Cyclotron Road, Berkeley, CA 94556; E-mail: email@example.com.
Submitted for publication January 2011.
Accepted for publication March 2011.