From the publication:

Twenty-seven studies were included, with a total of 309 participants. Short-term (<3 weeks), low- versus high-carbohydrate diets moderately increased resting cortisol. Whereas, long-term (≥3 weeks), low-carbohydrate diets had no consistent effect on resting cortisol. Low- versus high-carbohydrate diets resulted in much higher post-exercise cortisol, after long-duration exercise (≥20 min). Moderate-protein (<35%), low-carbohydrate diets had no consistent effect on resting total testosterone, however high-protein (≥35%), low-carbohydrate [≤35%] diets greatly decreased resting [∼5.23 nmol/L = 150.72 ng/dL] and post-exercise total testosterone.