Is a fasting testosterone level really necessary for the determination of androgen status in men?

As circulating testosterone may be suppressed in the post-prandial state, it has been recommended that measurements are carried out with the patient fasted.

In this regard, we assessed the effect of fasting/non-fasting status on total testosterone (T) levels in men.

Data was collected in a single UK Hospital in men with two serum T requests taken within a 6-month period of each other and sampled at a time of day ≤ 2 h apart. Three groups were established, with T levels compared via signed-rank test in men with both a fasting and non-fasting sample (Group 1; n = 69), and in men with paired non-fasting (Group 2; n = 126) and paired fasting (Group 3; n = 18) samples. The differences in T levels between the paired samples was compared between the three groups using the rank-sum test and also via multiple regression analysis with the groups factorised.

Median (Interquartile Range, IQR) age did not vary significantly between Groups 1, 2 and 3 at 49 (38-56), 51.5 (42-60) and 51.5 (40-59) years, respectively. No significant difference (p = 0.89) was found between the T levels in Group 1 with non-fasting (median (IQR) T = 11.1 (9.3-13.6) nmol/L) versus fasting samples T = 10.8 (8.9-14.1) nmol/L). Paired T levels did not significantly differ in each of the other two groups (2 and 3). There was no significant association between the differences in paired T levels between the three groups, even when the model was adjusted for age and time, with Group 1 (as reference) versus Group 2 (p = 0.79) and versus Group 3 (p = 0.63).

We found no significant differences between fasting and non-fasting T levels. A definitive confirmatory study is required to determine whether fasting status is necessary to diagnose hypogonadism.

Non-requirement of fasting status when checking testosterone levels would remove a major hurdle in the diagnosis of hypogonadism.