Table 3.
Changes in sexual function in circumcised and uncircumcised men over 12 and 24 months of follow-up: results from adjusted random-intercept logistic regression models∗
| Outcome and parameters | Estimate (β) | Standard error | Pr > |t| |
|---|---|---|---|
| Lack of interest in sex | |||
| 12 mo | −0.8 | 0.08 | <0.001†‖ |
| 24 mo | −1.2 | 0.1 | <0.001†‖ |
| VMMC | 0.0 | 0.1 | 0.622 |
| Employment | −0.1 | 0.1 | 0.033 |
| Erectile dysfunction | |||
| 12 mo | −0.7 | 0.1 | <0.001†‖ |
| 24 mo | −1.3 | 0.1 | <0.001†‖ |
| VMMC | 0.0 | 0.1 | 0.741 |
| Age 25–29 y | 0.3 | 0.1 | 0.007‖ |
| Age 30–35 y | 0.4 | 0.2 | 0.007‖ |
| Secondary education | −0.4 | 0.1 | <0.001‖ |
| Marital status | 0.3 | 0.1 | <0.001‖ |
| Condom use at most recent sexual encounter | −0.2 | 0.1 | 0.018§‖ |
| Orgasm difficulties | |||
| 12 mo | −0.7 | 0.1 | <0.001†‖ |
| 24 mo | −1.2 | 0.1 | <0.001†‖ |
| VMMC | 0.1 | 0.1 | 0.485 |
| Secondary education | −0.5 | 0.1 | <0.001‖ |
| Marital status | 0.2 | 0.1 | 0.001‖ |
| Premature ejaculation | |||
| 12 mo | −0.3 | 0.1 | <0.001†‖ |
| 24 mo | −0.7 | 0.1 | <0.001†‖ |
| VMMC | −0.2 | 0.1 | 0.001‖ |
| Age 25–29 y | 0.3 | 0.1 | 0.001‖ |
| Secondary education | 0.2 | 0.1 | 0.023‖ |
| Employment | 0.1 | 0.1 | 0.016‖ |
| Marital status | 0.3 | 0.1 | <0.001‖ |
| Lack of pleasure during sex | |||
| 12 mo | −0.9 | 0.1 | <0.001†‖ |
| 24 mo | −1.5 | 0.1 | <0.001†‖ |
| VMMC | 0.0 | 0.1 | 0.587 |
| Employment | −0.2 | 0.1 | 0.010‖ |
| Marital status | 0.4 | 0.1 | <0.001‖ |
| Pain during intercourse (dyspareunia) | |||
| 12 mo | −0.1 | 0.1 | 0.329 |
| 24 mo | 0.1 | 0.1 | 0.711 |
| VMMC | −0.1 | 0.2 | 0.623 |
| 12 mo × VMMC | −0.9 | 0.2 | <0.001‡‖ |
| 24 mo × VMMC | −1.7 | 0.2 | <0.001‡‖ |
| Employment | 0.4 | 0.1 | <0.001‖ |
| Condom use at most recent sexual encounter | −0.3 | 0.1 | 0.003§‖ |
VMMC = voluntary medical male circumcision.
Main effects for time (12 and 24 months), group (VMMC), and time-by-group interaction using baseline visit and control group as reference. Time-by-group interaction and other covariates are listed only for a significant α value equal to 0.05. Age categories based on intervals used for recruitment of participants (18–24, 25–29, and 30–35 years) use the youngest age category as reference. All models were adjusted for age, education, employment, and marital status regardless of statistical significance.
Significant decreases over time in the two study groups.
Significant decrease over time in the VMMC group only.
No significant decreases over time in the effect of condom use at the most recent sexual encounter for erectile dysfunction and pain during intercourse. For pain during intercourse, condom use at the most recent sexual encounter was significant only for the VMMC group in stratified models (interaction β = −0.770, P < .001).
P values less than or equal to .05.