Association between improved erectile function and dietary patterns: a systematic review and meta-analysis

Abstract

Erectile dysfunction (ED) is prevalent among men, but its relationship with dietary habits is uncertain. The aim of our study was to assess whether dietary patterns enhance erectile function by reviewing the literature published before August 1, 2022, via PubMed, Web of Science, and EMBASE databases. The data compiled included author details; publication dates, countries, treatments, patient numbers, ages, follow-ups, and clinical trial outcomes, such as ED cases, odds ratios (ORs), confidence intervals (CIs), and International Index of Erectile Function-5 (IIEF-5) scores with means and standard deviations. An analysis of 14 studies with 27 389 participants revealed that plant-based diets (OR = 0.71, 95% CI: 0.66–0.75; P < 0.00001), low-fat diets (OR = 0.27, 95% CI: 0.13–0.53; P = 0.0002), and alternative diets such as intermittent fasting and organic diets (OR = 0.54, 95% CI: 0.36–0.80; P = 0.002) significantly reduced ED risk. High-protein low-fat diets (hazard ratio [HR] = 1.38, 95% CI: 1.12–1.64; P < 0.00001) and high-carb low-fat diets (HR = 0.79, 95% CI: 0.55–1.04; P < 0.00001) improved IIEF-5 scores. Combined diet and exercise interventions decreased the likelihood of ED (OR = 0.49, 95% CI: 0.28–0.85; P = 0.01) and increased the IIEF-5 score (OR = 3.40, 95% CI: 1.69–5.11; P < 0.0001). Diets abundant in fruits and vegetables (OR = 0.97, 95% CI: 0.96–0.98; P < 0.00001) and nuts (OR = 0.54, 95% CI: 0.37–0.80; P = 0.002) were also correlated with lower ED risk. Our meta-analysis underscores a strong dietary-ED association, suggesting that low-fat/Mediterranean diets rich in produce and nuts could benefit ED management.

INTRODUCTION

Erectile dysfunction (ED) refers to the condition where a man cannot get or keep a firm enough erection for a satisfying sexual experience.1 It is possible that ED has a negative impact on men’s mental health and has a significant effect on patients’ and their partners’ lives.1 Several common risk factors are associated with ED, both modifiable and unmodifiable, such as older age, the presence of diabetes mellitus, dyslipidemia, hypertension, metabolic syndrome (MetS), increased body mass index (BMI) and waist circumference, obesity, a lack of exercise, and smoking.2,3,4,5

ED is a prevalent medical issue that predominantly affects men aged 40 years and above. The International Consultation Committee for Sexual Medicine’s comprehensive analysis revealed that the incidence of ED in men under 40 years is between 1% and 10%, rising to 2%–9% for those between 40 years and 49 years of age and sharply increasing to 20%–40% for those between 60 years and 69 years of age. For individuals over 70 years old, the incidence rate soars to 50%–100%.6,7,8,9,10 The Massachusetts Male Aging Study’s longitudinal investigation reported an ED incidence rate of 26 cases per 1000 man yearly, with a significant increase to 46 cases for the 60–69-year age bracket. Among men aged 40 years to 70 years, the combined prevalence of mild, moderate, and complete ED is 52%.11

In the treatment of ED, in addition to medications and surgery, lifestyle interventions could be implemented for the aforementioned risk factors. For example, quitting smoking, reducing alcohol consumption, and engaging in more physical exercise significantly and favorably affect ED.12,13 In clinical trials, a low-fat or replacement diet may relieve sexual dysfunction or, at least in part, prevent it from worsening.14,15,16 We carried out a thorough review and meta-analysis of pertinent studies to investigate the associations between diet and ED.

PATIENTS AND METHODS

This study was conducted in accordance with the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) and was registered with the International Prospective Register of Systematic Reviews (PROSPERO) prior to the screening of studies for inclusion (ID: CRD42023409561).17

Data search

To find pertinent studies published before August 12, 2022, we conducted a literature search in the PubMed, Web of Science, and EMBASE databases. The following terms and their combinations were employed: (“diet”) AND (“erectile dysfunction” OR “sexual dysfunction”).

Study selection

Two reviewers independently evaluated each study (DLM and YCZ) and comprehensively assessed it via the following inclusion criteria: (1) erectile function was assessed via appropriate methods, including International Index of Erectile Function-5 (IIEF-5) and IIEF-15 scores and professional medical diagnosis; (2) the associations between ED status and patients’ dietary characteristics were assessed; (3) studies directly provided the odds ratios (ORs) with corresponding 95% confidence intervals (CIs) or means with standard deviations (s.d.) for ED, which was evaluated via the IIEF-5 or IIEF-15; and (4) studies were published in English. Studies with the following conditions were excluded: (1) case studies, letters, editorials, reviews, comments, or meeting abstracts; (2) investigations carried out on animals or in vitro; (3) multiple studies with similar data; or (4) studies that supplied data that could not be combined.

Data extraction

The data were separately extracted and compiled by two authors (BY and CW). If there were any disagreements, they could be worked out through conversation. The following fundamental traits were gathered: name of the first author, year of publication, nation, treatment, number of patients, age, and follow-up months. In terms of data from clinical trials, the number of patients with ED or without ED before and after the intervention, especially OR and 95% CI, and the IIEF-5 score with mean ± s.d. were collected.

Quality assessment

Two authors (BY and CW) independently conducted the quality assessment. The quality of the case-control studies and cohort studies was assessed according to the Newcastle–Ottawa Scale (NOS) criteria.18 The NOS standard includes the following three quality parameters: (1) selection 0–4; (2) comparability 0–1; and (3) outcome 0–3.18 Case-control studies and cohort studies with 5 or more points were deemed to be of excellent quality. The quality assessment of cross-sectional studies was performed according to criteria provided by the Agency for Healthcare Research and Quality (AHRQ).19 The AHRQ standard includes the following five parameters: (1) selection bias; (2) implementation bias; (3) follow-up bias; (4) measurement bias; and (5) reporting bias.19 Cross-sectional studies with 4–7 points were deemed of medium quality, and studies with 8 or more points were deemed of excellent quality. Quality assessment for randomized controlled trials (RCTs) was performed according to the Cochrane Collaboration tool for assessing the risk of bias. The Cochrane Collaboration tool standard includes the following six parameters: (1) selection bias; (2) performance bias; (3) detection bias; (4) attrition bias; (5) reporting bias; and (6) other bias. Studies with scores of 4 points or higher were considered to be of high quality.

Data synthesis and analyses

We used Review Manager software version 5.4 (The Nordic Cochrane Centre, Copenhagen, Denmark) to calculate the ORs with 95% CIs. Heterogeneity was assessed by the Chi-square test and I² statistic. When the P values of the Chi-square test were ≥0.05, fixed-effects models were used; otherwise, random-effects models were used. The statistical tests were two-sided, and P < 0.05 was considered to indicate statistical significance. If there were more than 10 included studies in a forest plot, a funnel plot was used to evaluate publication bias.

RESULTS

Study characteristics

We identified 1413 records from the PubMed, Web of Science, and EMBASE databases and included 14 articles14,15,16,20,21,22,23,24,25,26,27,28,29 in the final qualitative and quantitative analysis (Figure 1). The characteristics of the included studies are shown in Supplementary Table 1. A total of 27 389 patients from seven countries, including the USA, Australia, Canada, China, Italy, Singapore, and Spain, were included in these studies, which were published between 2004 and 2021. Notably, Khoo et al.,14 Maiorino et al.,15 and Moran et al.16 reported two cohorts with different intervention measures, and we adopted the method of self-comparison before and after the statistical analysis of the articles. Furthermore, as shown in Figure 1, we divided all the studies into three parts. The first part investigated the effect of diet on ED, and the results of this part were either ED or not ED.14,20,21,22,23,24,25 In the second part, diet was examined as a potential factor in ED, and the IIEF-5 score was calculated.15,16,26,27 In the last part, we discussed the differences in diet between people with ED and those without ED.28,29,30 Different scoring methods exist for different research types. According to the NOS score, AHRQ score, and Cochrane score, the quality of all included studies was middle or high (Supplementary Figure 1 ^{(95.6KB, tif)} and Supplementary Table 2–4). Sexual function was assessed via the IIEF-5.31 ED was categorized according to the overall score: mild ED (17–21 points), mild to moderate ED (12–16 points), moderate ED (8–11 points), and severe ED (1–7 points).

Figure 1.

Preferred reporting items for systemic reviews and meta-analysis flow diagram of literature screening. ED: erectile dysfunction.

Supplementary Table 1.

Characteristics of studies included in the meta-analysis

Study	Year*	Nation or location	Participant (n)	Age (years)	Diabetes, n (%)	BMI (kg m−2)
Lu et al.20	2021	China	184	45.0±13.2a	NM	24.9±6.2a
Bauer et al.21	2020	USA	21 469	62±8a	0 (0)	25.9±3.7a
Huynh et al.23	2020	USA and Irvine	271	53.3±17.6a	35 (12.9)	27.6±5.1a
Punjani et al.22	2018	Canada	2000	48 (34–60)b	204 (10.2)	NM
Mykoniatis et al.28	2018	USA	350	28 (23–34)b	0 (0)	25.0 (23.5–27.7)b
Maiorino et al.15	2016	Italy	104	53.0±9.4a	106 (100.0)	>25d
Moran et al.16	2016	Australia	118	49.7±1.2a	0 (0)	33.2±0.5a
Ramírez et al.29	2016	Spain	440	54.7±9.1a	0 (0)	28.4±3.8a
Wang et al.30	2013	Canada	1466	65.1c	1257 (85.7)	NM
Khoo et al.14	2014	Singapore	48	40.5±1.1a	0 (0)	32.7±0.5a
Giugliano et al.24	2010	Italy	555	57.0±6.7a	555 (100.0)	28.9±5.0a
Esposito et al.25	2009	Italy	209	45.5±6.9a	0 (0)	31.7±4.7a
Esposito et al.26	2006	Italy	65	43.9±6.2a	10 (15.4)	28.0±3.7a
Esposito et al.27	2004	Italy	110	43.3±5.0a	0 (0)	36.7±2.4a

*The literature published year. ^aMean±standard deviation. ^bMedian (IQR). ^cMean. ^dRange. NM: not mentioned; IQR: interquartile range; BMI: body mass index

Supplementary Table 2.

The quality assessment of cross-sectional studies by using Agency for Healthcare Research and Quality

Evaluation item	Huynh et al.23	Punjani et al.22	Ramírez et al.29	Wang et al.30
Define the source of information	Yes	Yes	Yes	Yes
List inclusion and exclusion criteria for exposed and unexposed subjects or refer to previous publications	Unclear	Yes	Yes	Yes
Indicate time period used for identifying patients	Yes	Yes	Yes	Yes
Indicate whether or not subjects were consecutive if not population based	No	No	Yes	No
Indicate if evaluators of subjective components of study were masked to other aspects of the status of the participants	Yes	No	Unclear	Unclear
Describe any assessments undertaken for quality assurance purposes	No	No	No	No
Explain any patient exclusion from analysis	Yes	Yes	Yes	Yes
Describe how confounding was assessed and/or controlled	No	Yes	Yes	Yes
If applicable, explain how missing data were handled in the analysis	No	No	No	No
Summarize patient response rates and completeness of data collection	No	No	No	Yes
Clarify what follow-up, if any, was expected and the percentage of patients for which incomplete data or follow-up was obtained	No	No	No	No

Supplementary Table 4.

The quality assessment of cohort studies by using the Newcastle–Ottawa Scale

Study	Selection				Comparability Comparability of cohorts on the basis of the design or analysis	Outcome			Total score
	Representative of exposed cohort	Selection of nonexposed cohort	Ascertainment of exposure	Demonstration that outcome of interest was not present at start of study		Assessment of outcome	Was follow-up long enough for outcomes to occur	Adequacy of follow-up of cohorts
Bauer et al.21	0	1	1	1	1	1	1	0	6
Giugliano et al.24	0	1	1	1	1	0	1	1	6

Supplementary Table 3.

The quality assessment of case–control studies by using the Newcastle–Ottawa Scale

Study	Selection				Comparability Comparability of cases and controls on the basis of the design or analysis	Outcome			Total score
	Is the case definition adequate?	Representativeness of the cases	Selection of controls	Definition of controls		Ascertainment of exposure	Same method of ascertainment for cases and controls	Nonresponse rate
Lu et al.20	1	1	1	1	1	1	1	1	8
Mykoniatis et al.28	0	0	1	1	1	1	1	1	6

The effect of diet on ED

Seven articles14,20,21,22,23,24,25 regarding the association between diet and ED are presented in this section (Figure 2). A plant-based diet (OR = 0.71, 95% CI: 0.66–0.75; P < 0.00001), a low-fat diet (OR = 0.27, 95% CI: 0.13–0.53; P = 0.0002), and other dietary patterns, including an intermittent fasting diet and an organic diet (OR = 0.54, 95% CI: 0.36–0.80; P = 0.002) were significantly associated with a decreased risk of developing ED. The Mediterranean diet did not significantly decrease the risk of developing ED (OR = 0.96, 95% CI: 0.91–1.01; P = 0.12). There was significant heterogeneity (P < 0.001; I² = 90%), so a random-effects model was used. Notably, detailed recommendations, including diet and exercise, decreased the risk of developing ED (OR = 0.49, 95% CI: 0.28–0.85; P = 0.01).

Figure 2.

Forest plot of dietary patterns on the risk of erectile dysfunction. CI: confidence interval; s.e.: standard error; df: degree of freedom; IV: inverse-variance; PDI: plant-diet index; hPDI: healthy PDI.

The effect of diet on IIEF-5

Four articles15,16,26,27 are discussed in this section (Figure 3). High-protein and, low-fat diet (mean difference [MD] = 1.38, 95% CI: 1.12–1.64; P < 0.00001); and a high-carbohydrate and, low-fat diet (MD = 0.79, 95% CI: 0.55–1.04; P < 0.00001) were significantly associated with higher IIEF-5 scores, than those of patients consuming other diets. However, there was significant heterogeneity (P < 0.00001; I² = 93%); as a result, we used a random-effects model. Notably, individuals following the detailed recommendations, including those for diet and exercise, would increase IIEF-5 score (MD = 3.40, 95% CI: 1.69–5.11; P < 0.0001).

Figure 3.

Forest plot of International Index of Erectile Function-5 score in individuals with different dietary patterns. CI: confidence interval; s.d.: standard deviation; df: degree of freedom; IV: inverse-variance.

Differences in diet between patients with ED and those without ED

Three articles28,29,30 investigated the differences in the consumption tendency of diet between patients with ED and those without ED (Figure 4). Non-ED patients tended to consume a diet rich in vegetables and fruits (OR = 0.97, 95% CI: 0.96–0.98; P < 0.00001) or a nut-rich diet (OR = 0.54, 95% CI: 0.37–0.80; P = 0.002).

Figure 4.

Forest plot of the differences in the consumption tendency of diet between individuals with erectile dysfunction and those without ED. CI: confidence interval; s.e.: standard error; df: degree of freedom; IV: inverse-variance; ED: erectile dysfunction.

DISCUSSION

Fruits, nuts, legumes, grains, and seafood make up the majority of the Mediterranean diet. In the Mediterranean diet, the major source of monounsaturated fat is olive oil, whereas the consumption of alcohol and red meat is moderate to low. This diet is linked to several positive health outcomes, such as lower risks of cardiovascular disease, cancer, and death.32 The Mediterranean diet is advantageous for treating metabolic diseases, such as type 2 diabetes and metabolic syndrome.33 At present, the anti-inflammatory, antithrombotic, and antioxidant benefits of the Mediterranean diet are described as part of this protective mechanism.34 The Mediterranean diet can stop the ED that develops over time in patients with diabetes.15

Diabetes is recognized as a significant risk factor for ED. However, dietary fiber has been linked to a reduced risk of type 2 diabetes,35 suggesting that increased fiber consumption could indirectly mitigate the risk of developing ED.36 Furthermore, the cholesterol-lowering properties of dietary fiber and its role in enhancing cardiovascular health are noteworthy. These benefits may exert a positive influence on ED by improving vascular function, which is integral to achieving and maintaining erections. These findings highlight the importance of a fiber-rich diet not only in diabetes management but also in the broader context of men’s health, particularly with respect to sexual function.

Emerging research points to a significant correlation between the intake of polyunsaturated fatty acids (PUFAs) and the prevalence of ED. A cross-sectional analysis leveraging data from the National Health and Nutrition Examination Survey (NHANES) revealed that increased consumption of arachidonic acid (AA) was linked to a reduced incidence of ED.37 Furthermore, certain fatty acids, notably omega-3 fatty acids, have been implicated in lowering the risk of developing ED by enhancing endothelial function and mitigating inflammation.38 Notably, dietary patterns such as the Mediterranean diet and the Alternative Healthy Eating Index (AHEI)-2010, which advocate for increased consumption of fruits, vegetables, whole grains, nuts, and legumes, along with polyunsaturated fats and omega-3 fatty acids, and decreased intake of red meat, processed meat, sugary beverages, and trans fats, have demonstrated efficacy in reducing the risk of male ED.39

Intermittent fasting has been shown to increase testosterone levels, improve nutritional metabolism, and manage blood sugar levels.40,41,42 In addition, eating organically decreases exposure to pesticides and fertilizers, which are both known to contribute to poor development and neurological diseases.43,44,45 The benefits of intermittent fasting and an organic diet, which were previously discussed, may not only help improve overall health but also help with ED.

A diet characterized by a low-fat content, whether complemented by high protein or high complex carbohydrate contents, has been shown to ameliorate ED. Generally, a low-fat diet significantly diminished the risk of developing ED (OR = 0.27, 95% CI: 0.13–0.53; P = 0.0002), indicating a statistically robust association. Subsequently, we found that both low-fat and high-protein diets, and low-fat and high-complex-carbohydrate diets positively influence IIEF-5 scores. These findings are similar, indicating that both dietary patterns are beneficial. Currently, relevant studies comparing high-protein and low-fat diets, and high-carbohydrate and low-fat diets are limited and should be conducted in the future.

This meta-analysis suggests that the non-ED group eats more fruits, vegetables, and nuts than the ED population does. Those who follow a plant-based diet also have a lower chance of developing ED. According to short-term interventional trials, fruits and vegetables and other meals high in antioxidants are associated with positive postprandial effects on endothelial function and blood pressure.46,47,48,49

Nuts have high vitamin, mineral, and heart-healthy fat contents, which are essential for sustaining cardiovascular health, a key factor in sexual function. Therefore, the inclusion of nuts in one’s diet may contribute to the management of ED. For example, one study suggested that pistachios may improve ED,50 although further research is needed to confirm this effect. In addition, in the study conducted by Ramírez et al.,29 the specific types and processing methods of nuts were not detailed, which should be noted in future studies.

A plant-based diet (PBD) may offer a preventative and therapeutic approach to treating ED by enhancing vascular function. Rich in bioactive polyphenolic compounds, plant-based foods can increase the bioavailability of nitric oxide, a key molecule in the vasodilation process that facilitates erections. This mechanism could underlie the benefits observed in both Mediterranean diet and PBD for improving erectile function. Moreover, PBDs are reported to mitigate the risk of conditions such as type 2 diabetes, hypercholesterolemia, hypertension, atherosclerosis, and coronary artery disease, which are recognized as contributors to ED.20,51,52,53 However, it is important to consider the case reported by Siepmann et al.,54 where a 19-year-old man with type 1 diabetes experienced a decline in sexual function after consuming a large amount of soy products as part of a vegan diet. His sexual function fully recovered 1 year after the diet was discontinued, suggesting that while PBDs can be beneficial, they may not be universally suitable for all individuals. The relationship between PBDs and ED is an area of ongoing research.

In 2018, La et al.55 conducted a literature review that dissected the intricate interplay between nutrition and a spectrum of male sexual health issues, including ED, male gonadal dysfunction, and male infertility. The review underscored the promising role of the Mediterranean diet in enhancing male sexual health and offered concrete dietary guidance. It also highlighted the need for further research, especially prospective studies and randomized controlled trials, to clarify the influences of nutrition on erectile function. In our research, we have advanced the discourse initiated by La et al.,55 conducting a more substantial and expansive examination of the associations between dietary habits and erectile function. By employing systematic review methodologies and meta-analytic techniques, we have integrated the latest research insights and adhered to rigorous quality control standards, thereby achieving a more nuanced comprehension of the dietary elements that could ameliorate ED. Our study serves to augment and diversify the existing body of knowledge, potentially enriching the perspectives of health-care providers and the informed public on the therapeutic potential of dietary patterns in ED management.

According to the NOS score, AHRQ score, and Cochrane score, the methodological quality of the included studies was rated as moderate to high, indicating the reliability and robustness of the evidence presented, which is essential for informing clinical practice and guiding future research directions in the field. However, our meta-analysis had the following limitations that must be considered. The quality of the present meta-analysis was limited by several factors that may lead to seemingly opposite results from the reports included in the study. (1) Most of the included articles were retrospective studies. As a result, it was impossible to completely exclude confounding factors, which caused the results to deviate. (2) Due to publication bias, the limited sample sizes of some studies could provide widely divergent results. (3) Some studies did not provide adequate data, which could introduce bias. (4) Some studies implemented two interventions at the same time, and there was no control group; therefore, we used a statistical method of comparison for pre-post interventions, which could have led to some bias. (5) Given the limitations of the included studies, we did not adjust for some confounders, such as diabetes status, that could influence ED. (6) In our study, there were no publications regarding Japanese or Buddhist vegan diets, which are similar to the Mediterranean diet. (7) Regional sample size imbalances might bias the data, affecting the study’s representativeness. We took several significant actions to lessen these biases. First, we carried out a methodical and thorough search through online databases. Second, the inclusion criteria were quite specific; we removed some bias caused by possible confounders, and two independent reviewers independently collected or summarized the data. Finally, we carried out subgroup analyses for various dietary patterns.

CONCLUSION

Our meta-analysis indicated that ED status and dietary patterns are strongly associated. ED patients should adopt a healthy diet, such as a low-fat or Mediterranean diet, and consume more fruits, vegetables, and nuts, thus improving the ED state.

AUTHOR CONTRIBUTION

JHL, XYY, and WMY conceptualized the study and supervised the project. BY and CW curated the data, performed the formal analysis, managed the software tasks, validated the findings, prepared the original draft, and were in charge of visualization. DLM and YCZ conducted the investigations. DLM and JB developed the methodology. JB, ZL, XYY, and WMY administered the project. ZL and XML provided the resources. YCZ, XML, XYY, and WMY contributed to the review and editing of the manuscript. All authors read and approved the final manuscript.

COMPETING INTERESTS

All authors declare no competing interests.

Supplementary Figure 1

Risk of bias of included randomized controlled trials. Risk of (a) bias graph and (b) bias summary.