Abstract
Background
We aimed to assess whether distinct lifestyle strategies can differentially affect specific body adipose depots.
Methods
We performed an 18-month randomized controlled trial among 278 sedentary adults with abdominal obesity (75%) or dyslipidemia in an isolated workplace with a monitored provided lunch. Participants were randomized to isocaloric low-fat or Mediterranean/low-carbohydrate (MED/LC) diet+28 g walnuts/day with/without added moderate physical activity (PA; 80% aerobic; supervised/free gym membership). Overall primary outcome was body fat redistribution, and the main specific end point was visceral adipose tissue (VAT). We further followed the dynamics of different fat depots (deep and superficial subcutaneous, liver, pericardial, muscle, pancreas, and renal sinus) by magnetic resonance imaging.
Results
Of 278 participants (age, 48 years, 89% men, body mass index, 30.8 kg/m2), 86% completed the trial with good adherence. The low-fat group preferentially decreased reported fat intake (-21.0% versus -11.5% for the MED/LC; P<0.001), and the MED/LC group decreased reported carbohydrates intake (-39.5% versus -21.3% for the low-fat group; P<0.001). The PA+ groups significantly increased the metabolic equivalents per week versus the PA- groups (19.0 versus 2.1; P=0.009). Whereas final moderate weight loss was indifferent, exercise attenuated the waist circumference rebound with the greatest effect in the MED/LCPA+ group (P<0.05). VAT (-22%), intrahepatic (-29%), and intrapericardial (-11%) fats declines were higher than pancreatic and femur intermuscular fats (1% to 2%) loss. Independent of weight loss, PA+ with either diet had a significantly greater effect on decreasing VAT (mean of difference, -6.67cm2; 95% confidence interval, -14.8 to -0.45) compared with PA-. The MED/LC diet was superior to the low-fat diet in decreasing intrahepatic, intrapericardial, and pancreatic fats (P<0.05 for all). In contrast, renal sinus and femoral intermuscular fats were not differentially altered by lifestyle interventions but by weight loss per se. In multivariate models further adjusted for weight loss, losing VAT or intrahepatic fat was independently associated with improved lipid profile, losing deep subcutaneous adipose tissue with improved insulin sensitivity, and losing superficial subcutaneous adipose tissue remained neutral except for an association with decreased leptin.
Conclusions
Moderate weight loss alone inadequately reflects the significant lifestyle effects on atherogenic and diabetogenic fat depots. The MED/LC diet mobilizes specific ectopic fat depots, and exercise has an independent contribution to VAT loss. Fat depots exhibit diverse responsiveness and are differentially related to cardiometabolic markers. Distinct lifestyle protocols may uniquely induce fat mobilization from specific anatomic sites.