Efficacy of Lifestyle Interventions in Treating Erectile Dysfunction: A Systematic Review and Meta-Analysis of Randomized Controlled Trials

16 RCTs, 1,477 men: lifestyle interventions improved IIEF scores by 2.35 points — statistically significant, clinically borderline.

Journal: The Journal of Sexual Medicine | Published: 2026-05-11 | Type: Systematic Review, Meta-Analysis | PMID: 42143598 Authors: Li Tingxin, He Xinlei, Huang Wenrui, Liu Guangyi, Feng Qiang (Sichuan Academy of Medical Sciences & Sichuan Provincial People's Hospital; Shenzhen Hospital of Traditional Chinese Medicine; First Hospital of China Medical University) Funding/COI: Sichuan Science and Technology Program; Health Science Research Project of Sichuan Province. No COI declared.

Summary

This meta-analysis pooled 16 RCTs covering 1,477 men and found that diet, exercise, or combined lifestyle interventions improved erectile function scores by a mean of 2.35 points on the International Index of Erectile Function (IIEF) compared to controls (95% CI: 1.68–3.01). The effect held across all intervention subtypes — no single approach outperformed the others. GRADE-rated certainty of evidence was moderate, which is unusually honest for this literature.

The number matters. The minimally clinically important difference (MCID) on the IIEF erectile function domain is generally cited around 4 points. A 2.35-point improvement is statistically real but may land below the threshold most patients would notice.

Claims

• Lifestyle interventions improved IIEF scores by MD 2.35 (95% CI 1.68–3.01) vs. controls
• Heterogeneity was moderate: I² = 53%
• Effect was significant in all three subgroups: exercise-only, diet-only, combined diet-and-exercise
• No statistically significant differences between subgroups — no modality won
• Effects held across intervention duration, age, BMI, and supervision categories
• Sensitivity analyses (leave-one-out, exclusion of diet-only trials) did not materially change results
• GRADE certainty rated moderate

Study Quality

This is methodologically above average for this topic. The authors searched four databases through December 2025, used both fixed- and random-effects models, assessed risk of bias with the Cochrane tool, applied GRADE to certainty ratings, and pre-registered subgroup analyses by intervention type, duration, age, BMI, and supervision status. The willingness to rate their own evidence as "moderate" rather than inflating it is a point in their favor.

The heterogeneity (I² = 53%) is moderate and expected given that "lifestyle intervention" covers everything from supervised aerobic exercise three times a week to a Mediterranean diet pamphlet. The authors acknowledge they cannot identify which intervention design is optimal — an honest admission that the pooled effect obscures a lot of variation in what was actually done to participants.

Red Flags

• 2.35-point IIEF improvement is below the ~4-point MCID — the pooled effect may be statistically significant but clinically marginal for individual patients
• "Lifestyle intervention" is heterogeneous to the point of near-incoherence: the meta-analysis cannot tell you what kind of exercise, how much, or for how long
• Inconsistent reporting of ED etiology and baseline severity across trials limits interpretation — a man with mild psychogenic ED and a man with severe vasculogenic ED are being averaged together
• Adherence data varied widely and was incompletely reported across trials
• 16 trials and 1,477 participants is modest for a meta-analysis covering a condition this prevalent
• No long-term follow-up data — unclear whether improvements persist after intervention ends
• Mixed clinical populations (some trials included men on concurrent PDE5 inhibitors or other treatments)

Strengths

• Pre-specified subgroup analyses reduce post-hoc fishing
• Multiple sensitivity analyses support robustness of the primary finding
• Authors explicitly rate GRADE certainty as moderate rather than claiming high-quality evidence
• Broad database search through late 2025 captures recent trials
• Restricted to RCTs — no observational studies inflating the effect estimate

Verdict

This paper does what a good meta-analysis should: it gives you a number, tells you how confident to be in it, and admits what it can't tell you. The finding — lifestyle interventions move the IIEF by about 2.35 points — is real but should be contextualized against the MCID before anyone gets excited. The 53% heterogeneity means the pooled effect is an average across a wide range of interventions of varying intensity, duration, and design, which limits practical application considerably. Worth reading for the subgroup tables; do not cite the headline figure without noting the clinical significance question.