The Relationship Between Erection and Satisfaction in ED
The Relationship Between Erection and Satisfaction in ED
Of the 209 men who were randomized and received treatment with sildenafil (n=104; mean±s.d. age, 53±12 years; mean ED duration, 4.6 years) or placebo (n=105; mean±s.d. age, 51±12 years; mean ED duration, 3.8 years), 167 had sufficient DBPC and open-label data for the current analysis.
When the model was applied to the data, the value obtained for the Bentler Comparative Fit Index was 0.98, indicating that the model fits the data well.
Direct effects of treatment (sildenafil vs placebo) on erection, but not on individual satisfaction or couple satisfaction, were statistically significant at week 4 and week 10 (t-value >1.96 in absolute terms). The direct effects of treatment on individual satisfaction and on couple satisfaction were close to zero and were virtually completely mediated via erection. Correlations between disturbances for the same domain over time were larger for adjacent times than for non-adjacent times, as expected.
Results from performing bootstrap simulations to estimate the 95% CIs showed that direct treatment effects on individual satisfaction and couple satisfaction were not statistically significant. The indirect effects of the treatment (sildenafil vs placebo) on individual satisfaction and couple satisfaction via erection were 97% (95% CI: 78%, 125%) and 108% (95% CI: 91%, 132%), respectively, at week 4 and 100% (95% CI: 82%, 124%) and 92% (95% CI: 76%, 113%), respectively, at week 10 (Table 1).
The difference between direct effects at week 4 vs week 10 and the difference between indirect effects at week 4 vs week 10 were not statistically significant, indicating that the mediation effects are longitudinally invariant (Table 1).
Results
Of the 209 men who were randomized and received treatment with sildenafil (n=104; mean±s.d. age, 53±12 years; mean ED duration, 4.6 years) or placebo (n=105; mean±s.d. age, 51±12 years; mean ED duration, 3.8 years), 167 had sufficient DBPC and open-label data for the current analysis.
Mediation Modeling
When the model was applied to the data, the value obtained for the Bentler Comparative Fit Index was 0.98, indicating that the model fits the data well.
Direct effects of treatment (sildenafil vs placebo) on erection, but not on individual satisfaction or couple satisfaction, were statistically significant at week 4 and week 10 (t-value >1.96 in absolute terms). The direct effects of treatment on individual satisfaction and on couple satisfaction were close to zero and were virtually completely mediated via erection. Correlations between disturbances for the same domain over time were larger for adjacent times than for non-adjacent times, as expected.
Results from performing bootstrap simulations to estimate the 95% CIs showed that direct treatment effects on individual satisfaction and couple satisfaction were not statistically significant. The indirect effects of the treatment (sildenafil vs placebo) on individual satisfaction and couple satisfaction via erection were 97% (95% CI: 78%, 125%) and 108% (95% CI: 91%, 132%), respectively, at week 4 and 100% (95% CI: 82%, 124%) and 92% (95% CI: 76%, 113%), respectively, at week 10 (Table 1).
The difference between direct effects at week 4 vs week 10 and the difference between indirect effects at week 4 vs week 10 were not statistically significant, indicating that the mediation effects are longitudinally invariant (Table 1).
