The present sub-project of the SPeED study aimed to investigate the difference in circulating leptin levels between depressed patients and healthy controls and to examine the impact of acute and long-term endurance exercise on leptin levels.
Baseline characteristics - Within the presented sub-project of the SPeED study, 105 patients with major depressive disorder (MDD) were compared with 34 healthy controls (HC) as reported in the flow chart in Fig. 1. At baseline, the HC and MDD group had comparable anthropometric characteristics (Fig. 2A) including the phenotypic traits body weight (72.2 ± 2.4 vs. 75.4 ± 1.6; p = 0.296), body fat (18.9 ± 0.9 vs. 17.9 ± 0.8; p = 0.589), percent body fat (25.6 ± 1.7 vs. 23.1 ± 0.8; p = 0.163) as well as a similar body mass index (BMI) (24.6 ± 0.7 vs. 24.7 ± 0.4; p = 0.929). Depressive symptoms (BDI-2) were significantly higher in the MDD group (0.82 ± 0.27 vs. 27.08 ± 0.77; p < 0.001) (Fig. 2B).
In the current study, bivariate correlation analyses also showed a positive correlation between leptin levels and body weight (HC: r = 0.474, p = 0.005; MDD: r = 0.198, p = 0.043). Given the numerical difference in the magnitude of the association between depressed patients and healthy controls, we calculated a z-test to determine whether the association between body weight and leptin levels was attenuated in depressed patients, but this difference proved not be significant (z = 1.46, p = 0.14). For the percentage of body fat, we found an even stronger correlation of equal magnitude in both groups, HC and MDD (HC: r = 0.755, p < 0.001; MDD: r = 0.675, p < 0.001) (Fig. 2C). Similarly, no difference in leptin levels was detected between patients with depression and healthy controls (HC: 12.43 ± 1.89 vs. MDD: 10.91 ± 1.05) (Fig. 2D, left panel). The subdivision of the individual groups according to gender showed, as expected, significantly higher values for women than for men, but again independent of the depression status (Fig. 2D). In conclusion, circulating leptin levels were not altered in patients with major depression compared to healthy controls.
Effects of acute exercise - To test whether leptin levels change after a single bout of exercise and were dependent on the depression status, all participants performed a bicycle ergometer test. As shown in Fig. 3A, the change of leptin levels (Time: F(1,137) = 1.678, p = 0.197, partial η² = 0.012, Time*Group: F(1,137) = 0.534, p = 0.466, partial η² = 0.004) was not significantly different between pre- and post- acute exercise and not different between MDD patients and healthy controls, showing that the bicycle ergometer test had no effect on the change in leptin levels, nor that the response was altered in depression.
Long-term effects of endurance exercise - All MDD patients were randomly allocated to one of three groups; a low-intensity (LEX) or a high-intensity (HEX) endurance exercise group (EX) or a waiting list control group (WL) (Fig. 1). Overall, the comparison of EX and WL revealed a marginally significant main and significant interaction effect (Time: F(1,65) = 3.176, p = 0.079, partial η² = 0.047; Time*Group: F(1,65) = 4.474, p = 0.038, partial η² = 0.064). In contrast to the exercise group with similar leptin levels pre- and post-exercise intervention, leptin levels of the waiting group, which were higher at baseline, decreased after 12 weeks (Fig. 3B). Body weight (Time: F(1,65) = 0.381, p = 0.144, partial η² = 0.033; Group*Time: F(1,65) = 0.381, p = 0.539, partial η² = 0.006) and percent of body fat (Time: F(1,65) = 0.497, p = 0.483, partial η² = 0.008; Group*Time: F(1,65) = 0.004, p = 0.948, partial η² = 0.000) were not different between the groups pre and post exercise intervention (Fig. 3C), but body fat content was numerically higher in the WL than in the exercise groups (Fig. 3C). Furthermore, the WL group showed a sizeable decrease in body fat at baseline (t1, Fig. 1C) and this initial difference in body fat between WL and EX groups may explain the higher leptin levels in the WL at baseline as well as the subsequent stronger decrease in leptin levels over time. An independent sample t test revealed a statistically significant difference in percent of body fat at baseline between EX and WL, t(28.05) = -2.155, p = .040*. Since a considerable interaction effect of group membership on the leptin levels was observed, a possible impact of percent of body fat on the change in leptin levels was investigated by using an ANCOVA, taking leptin levels before and after the intervention as dependent variables, the group’s membership as the fixed factor and body fat as a covariate. The ANCOVA revealed that the interaction of the change in leptin levels and group membership was no longer significant when taking body fat as covariate, Time: F(1,64) = 0.044, p = 0.834; partial η² = 0.001; Time*Body fat: F(1,64) = 0.164, p = 0.687 partial η² = 0.003; Time*Group: F(1,64) = 3.587, p = 0.063, partial η² = 0.053.
-
In terms of training intensity, results of the rmANOVA considering leptin levels showed no significant changes indicating that there was no overall effect of exercise on leptin levels and this is independent of the exercise intensity (Time: F(1,45) = 0.150, p = 0.701, partial η² = 0.003; Time*Group: F(1,45) = 2.117, p = 0.153, partial η² = 0.045). However, the concentration of leptin levels decreased numerically in the HEX group, while the levels increased in the LEX group (Fig. 3D, left panel).