Skip to main content
. 2018 Nov 14;76(1):31–40. doi: 10.1001/jamapsychiatry.2018.2734

Table. Meta-regression of the Effectiveness of Testosterone Treatment on Various Study-Level Moderators.

Characteristic Estimate (SE) [95% CI] NHST
χ2 (df) P Value
Baseline
Age, y
40 0.223 (0.198) [−0.166 to 0.612] 0.835 (1) .36
60 0.393 (0.116) [0.165 to 0.621]
80 0.562 (0.238) [0.096 to 1.029]
Testosterone status
Eugonadal 0.208 (0.221) [−0.225 to 0.641] 1.021 (1) .31
Hypogonadal 0.475 (0.144) [0.192 to 0.758]
HIV infection
Yes 0.276 (0.331) [−0.373 to 0.926] 0.096 (1) .76
No 0.386 (0.123) [0.144 to 0.627]
Symptom level
Severe 0.385 (0.232) [−0.070 to 0.840] 1.194 (2) .55
Mild 0.277 (0.155) [−0.027 to 0.582]
Subclinical 0.595 (0.247) [0.112 to 1.078]
Symptom variability, CV, %
20 0.615 (0.180) [0.262 to 0.968] 2.935 (1) .09
50 0.439 (0.116) [0.211 to 0.667]
100 0.146 (0.169) [−0.185 to 0.478]
Treatment
Treatment dosage, g/wk
0.1 0.166 (0.137) [−0.103 to 0.434] 5.298 (1) .02
0.3 0.312 (0.108) [0.100 to 0.524]
1.0 0.823 (0.224) [0.383 to 1.263]
Treatment duration, wk
5 0.363 (0.141) [0.086 to 0.640] 0.012 (1) .91
20 0.370 (0.117) [0.142 to 0.599]
100 0.408 (0.345) [−0.268 to 1.083]
Administration
Intramuscular 0.243 (0.141) [−0.033 to 0.520] 3.352 (2) .19
Oral 0.851 (0.306) [0.252 to 1.450]
Transdermal 0.430 (0.222) [−0.006 to 0.865]

Abbreviations: CV, coefficient of variation; NHST, null hypothesis significance test.