Objectively measured physical activity predicts subsequent energy intake in 300 women

To determine if 7 d physical activity (PA) predicts energy intake over the same days and if PA on a given day predicts energy intake on the same day, the next day or subsequent days.

A 7 d prospective investigation. PA was measured using accelerometers worn for seven consecutive days. During the same week, total energy intake (including alcohol) was assessed using weighed food records.

Twenty cities in the USA.

Three hundred middle-aged women.

After controlling for the covariates, energy intake was 36 (se 8) kJ higher for each additional 100000 activity counts (F=19·0, P<0·0001), based on the entire week of monitoring. Women with Low PA (n 75) had a mean energy intake of 8364 (sd 1235) kJ/d (1998 (sd 295) kcal/d), those with Moderate PA (n 150) consumed 8523 (sd 1264) kJ/d (2036 (sd 302) kcal/d) and those with High PA (n 75) consumed 9079 (sd 1473) kJ/d (2169 (sd 352) kcal/d; F=6·4, P=0·0019), a 5023 kJ/d (1200 kcal/d) difference between the High and Low PA groups. PA on a given day was predictive of energy consumption on the same day on four of the seven monitored days, and was predictive of energy intake the next day, with five of six associations significant. PA was not consistently predictive of energy intake on days beyond the next day.

Over 7 d, PA and energy consumption tend to be positively related. Considered day to day, increased PA is weakly associated with increased energy intake the following day and, to a lesser extent, the same day. Therefore, managing weight by increasing PA may not result in the energy deficit expected.