The effectiveness of exercise in reducing body weight is individual and may be related to variability in compensatory responses⁽ Footnote ^{1 )}. The aim of the study was to identify the underlying mechanisms opposing exercise-induced perturbations in energy balance in overweight women.

Thirty-one healthy sedentary women (mean age 31.9 (sd 8.3) years; BMI 28.8 (sd 3.6) kg/m²; percentage body fat 39.1 (sd 4.7) participated in a supervised 8-week exercise programme, which consisted of 150 min cycling per week at an intensity of 90–95% lactate threshold. Body composition, activity energy expenditure (AEE; calculated as energy expended during all active activities including exercise sessions), inactivity energy expenditure (IEE; calculated as energy expenditure of sedentary activities), sleeping energy expenditure (SEE) and energy intake (EI) were assessed at baseline and during week 8 of the exercise programme. AEE and IEE were estimated from individual relationships between V_O2 and V_CO2 and heart rate (HR) and records of HR and physical-activity diaries obtained during waking hours of week 0 and week 8⁽ Footnote ^{2 )}. SEE was obtained from basal metabolic rate (BMR) measurements (Deltatrac System; Datex Instrumentation Inc., Baldwin Park, CA, USA). EI was assessed by 7 d self-recorded weighed intake. Measures of body mass and its components were taken using bioelectrical impedance scales (Tanita TBF-300; Tanita UK Ltd, Yiewsley, Middlesex, UK). Nine of the subjects achieved ≥75% predicted body-fat loss and were classified as responders (R), while another twenty-two subjects fell into the group of non-responders (NR). Changes in all variables from baseline to post-intervention assessment were compared by two-way ANOVA (group×time) with repeated measures on the ‘time’ factor and post hoc Tukey test was used to identify changes within a group.

There was a significant interaction (P<0.05) for the change in AEE between groups R and NR, with AEE during week 8 being significantly higher (P<0.05) in comparison with baseline only in group R. Changes in IEE, SEE and EI from week 0 to week 8 were not significant within both groups and not different between groups R and NR. Reported daily EI of 41% of participants was below BMR×1.3.

In conclusion, in overweight women reduction in physical activity in the non-exercise time can be expected during supervised exercise programmes and thus explain a lower-than-predicted weight loss.