The relationship between total energy expenditure and physical activity level in women living in an impoverished Brazilian urban area.

Objective: This study aimed to describe how strong is the relationship between TEE and PAL in women living in an impoverished Brazilian urban area.

Methods: Anthropometric, hormonal (insulin, TSH, FT , and FT ), body composition (deuterium), TEE (doubly labeled water) and PAL (metabolic equivalent task [MET]-7-day triaxial accelerometer, ActivPAL®) data were collected from 55 women (mean age: 31y, mean BMI: 27.4 kg/m ). Adjusted-TEE models were calculated incorporating the residuals of anthropometric, hormonal, and body composition variables in the TEE, to assess the relation between MET and adjusted-TEE, through three different analyzes: linear regression, nonlinear regression and change-point regression.

Results: Most participants (89.1%, n = 49) were classified as low-active. There was no association between crude TEE and MET.h/d (R  = 0.05; p = 0.09). There was a positive, although weak, linear relationship between adjusted-TEE and MET.h/d (β = 2705.26 kcal; 95% CI: 436.25; 4974.27; adj-R  = 0.08). A change point of this relationship was identified in the MET.h/d value of 1.53 (SD = 0.02; adj-R = 0.13). The association between adjusted-TEE and MET.h/d before the change point (β = 5101.46 kcal; 95%CI: 978.84; 9224.08; adj-R  = 0.11) was stronger than the association in the full linear model, nevertheless, it was still weak. This association stop existing when it is assessed in individuals after the change point (β = -6609.92 kcal; 95%CI: -16773.24; 3553.40; adj-R  = 0.08).

Conclusions: These results suggest that the relationship between TEE and PAL is weak and it is not completely linear in women living in an impoverished Brazilian urban area.