The Impact of Exercise Intensity and Sex on Endogenous Ghrelin Levels and Appetite in Healthy Humans

Abstract

Context: Ghrelin circulates in acylated (AG) and deacylated (DAG) forms, which are known to affect appetite. Although acute exercise has been shown to modulate ghrelin levels, data on the impact of exercise intensity on AG and DAG levels and their effects on appetite are sparse and primarily limited to males.

Objective: To investigate the effect of exercise intensity and sex on ghrelin levels and appetite in untrained humans.

Methods: Eight males (age: 43.1 ± 10.9 years; body mass index [BMI]: 22.2 ± 1.7 kg/m²; peak oxygen consumption [VO_2peak]: 36.3 ± 6.4 mL/kg/min) and 6 females (age: 32.2 ± 11.1 years; BMI: 22.7 ± 1.0 kg/m²; VO_2peak: 29.2 ± 4.0 mL/kg/min) completed a maximal graded cycle ergometer lactate threshold (LT)/VO_2peak test. These data were used to determine the exercise intensity on 3 subsequent randomized control or calorically matched cycle exercise bouts: (1) CON, no exercise; (2) MOD, the power output at LT; (3) HIGH, the power output associated with 75% of the difference between LT and VO_2peak. Perception of appetite was analyzed using visual analog scales.

Results: Females had higher levels of total ghrelin (TG) (P = .03) and DAG (P = .01) at baseline than males. Both groups exhibited reduced DAG levels in HIGH compared with MOD and CON (P < .0001-.004); however, only females had significantly reduced AG in HIGH (P < .0001). Hunger scores were higher in MOD than in CON (P < .01).

Conclusion: High-intensity may be superior to moderate-intensity exercise for reducing ghrelin levels and modifying hunger, and sex may impact this response.

Keywords: acute exercise; appetite regulation; gut hormones; lactate kinetics; sex differences.

Figure 1.

Outline of testing visit measurements.

Figure 2.

Effect of condition on (A) TG_TIMEPOINT and (B) TG_AUC. Data are mean ± SE. The statistical analysis revealed a main effect of condition (P < .0001), where levels of TG_TIMEPOINT were lower in HIGH than in MOD and CON. TG_TIMEPOINT levels were also significantly higher in MOD than in CON (P < .0001). There was a main effect of time (P < .0001), where TG_TIMEPOINT levels were lower during 40 to 180 minutes compared with baseline. There were significant interaction effects between condition and time and condition and sex (both P < .0001). HIGH had reduced TG levels at 40 to 120 minutes compared with baseline. Males had higher levels of TG_TIMEPOINT in MOD compared to HIGH. Females had significantly lower levels of TG in HIGH than in CON and MOD. Females also had lower TG in CON than MOD and lower TG levels in HIGH compared with males in HIGH (P = .03). For TG_AUC, there was a main effect for condition (P = .003), where levels of TG_AUC were significantly lower in HIGH than in MOD and CON. *Significantly different from MOD in males, P = .006. †Significantly different from CON in females, P < .0001. ††Significantly different from CON and MOD in females, P < .0001. $Significantly different from CON, P < .0001. $$Significantly different from MOD and CON, P ≤ .01.

Figure 3.

Effect of condition on (A) AG_TIMEPOINT and (B) AG_AUC. Data are mean ± SE. The statistical analysis revealed there was a main effect of condition (P < .0001), where levels of AG_TIMEPOINT were lower in HIGH than in MOD and CON. AG_TIMEPOINT levels were also significantly higher in MOD than in CON. There was a main effect of time (P < .0001), where AG_TIMEPOINT levels were lower during 40 to 180 minutes compared with baseline. There were significant interaction effects between condition and time and condition and sex (both P < .0001). HIGH had significantly decreased AG levels at 30 to 90 minutes compared with baseline. Females had lower AG_TIMEPOINT levels in HIGH than in MOD and CON. AG levels in MOD were also significantly higher than in CON. There was no significant difference across any condition in males. There was a significant main effect of condition for AG_AUC (P < .005), where levels of AG_AUC were significantly lower in HIGH than in MOD and CON. There was also an interaction effect between condition and sex (P < .0001), where female AG_AUC in HIGH was significantly lower than MOD. *Significantly different within females, P < .0001. †Significantly different from CON in females, P < .006. ††Significantly different from CON and MOD in females, P < .0001. $Significantly different from CON, P < .0006. $$Significantly different from MOD and CON, P ≤ .005.

Figure 4.

Effect of condition on (A) DAG_TIMEPOINT and (B) DAG_AUC. Data are mean ± SE. The statistical analysis revealed there was a main effect of condition (P < .0001), where levels of DAG_TIMEPOINT were lower in HIGH than in MOD and CON. DAG_TIMEPOINT levels were also significantly higher in MOD than in CON. There was a main effect of time (P < .01), where DAG levels were lower during 60 minutes compared with baseline. A main effect for sex indicated females had lower DAG_TIMEPOINT levels regardless of condition than males. There was a significant interaction effect between condition and time (P < .0001) and condition and sex (P = .007). DAG was significantly lower at 50 and 60 minutes compared with baseline in HIGH. Males had lower levels of DAG_TIMEPOINT in HIGH than in MOD and CON. Females had lower DAG_TIMEPOINT levels in HIGH than in MOD and CON. CON in females was lower than CON in males. For DAG_AUC, there was a significant main effect of condition (P = .01), where levels of DAG_AUC were significantly lower in HIGH than in MOD. *Significantly different from CON and MOD in males, P ≤ .004. ††Significantly different from CON and MOD in females, P ≤ .0002. $Significantly different from CON, P = .001. $$Significantly different from MOD and CON, P ≤ .002. ^Significantly different from MOD, P = .008.

Figure 5.

Effect of condition on (A) LAC_TIMEPOINT and (B) LAC_AUC. Data are mean ± SE. The statistical analysis revealed there was a main effect of condition (P < .0001), where levels of LAC_TIMEPOINT were elevated in HIGH compared with MOD and CON. Levels were higher in MOD than in CON (P < .0001). There was a main effect of time (P < .0001), where LAC levels were higher during 10 to 90 minutes compared with baseline. There was significant condition by time and condition by sex interactions (both P < .0001). LAC_TIMEPOINT was elevated from 10 to 90 minutes compared with baseline in HIGH. LAC_TIMEPOINT was higher from 20 to 60 minutes compared with baseline during MOD. Females and males had higher levels of LAC during HIGH compared with MOD and CON and MOD compared with CON. For LAC_AUC, there was a main effect of condition (P < .001), where HIGH had elevated levels compared with MOD and CON. *Significantly different from the other 2 conditions P < .0001.

Figure 6.

Effect of condition on (A) HUN_TIMEPOINT and (B) HUN_AUC. Data are mean ± SE. The statistical analysis revealed there was a main effect of condition (P < .0001), where HUN_TIMEPOINT was higher in MOD than in CON. There was a main effect of time (P < .0001), where scores were lower during 90 to 180 minutes compared with baseline. There was a significant interaction between condition and sex (P = .01). Males had significantly higher HUN_TIMEPOINT in MOD than in HIGH. Females had significantly higher HUN_TIMEPOINT in MOD than in CON. *Significantly different from HIGH in males P = .05. †Significantly different from CON in females, P = .05.

Figure 7.

Effect of condition on (A) APP_TIMEPOINT and (B) APP_AUC. Data are mean ± SE. The statistical analysis revealed there were significant main effects for condition (P < .001) and time (P < .0001). Appetite scores were significantly lower in CON than in MOD and HIGH. APP scores were significantly lower at 120, 150, and 180 minutes compared with baseline (P < .0001). *Significantly different from CON, P < .05.