Antioxidant Treatment and the Chance to Conceive in Men Seeking Fertility Care: The SUMMER Randomized Clinical Trial.

This investigator-initiated, double-blind, placebo-controlled randomized clinical trial or SUMMER trial was conducted from May 2018 to February 2024 in 21 centers in the Netherlands. All participants provided written informed consent before randomization. The regional medical-ethical commission CMO Arnhem-Nijmegen approved the trial ( Supplement 1 ). We followed the Consolidated Standards of Reporting Trials ( CONSORT ) reporting guideline. Those eligible for inclusion were men aged 18 to 50 years, with female partners aged 18 to 43 years, who were unable to conceive after 12 months of unprotected intercourse and who were advised to undergo expectant management (EM); treatment with first, second, or third cycle of intrauterine insemination (IUI) with or without mild ovarian stimulation; in vitro fertilization (IVF); or ICSI. Exclusion criteria were as follows: (1) treatment with testicular biopsy (testicular sperm extraction) or percutaneous epididymal sperm aspiration, ovulation induction without IUI, or IVF for bilateral tubal pathology; (2) treatment with donor sperm or oocytes, cryopreserved, or electroejaculated semen; (3) known chromosomal, urological, or endocrine abnormalities related to infertility; (4) male use of fertility-enhancing drugs; and (5) use of antioxidant supplements within 3 months before the study. Randomization was performed in a 1:1 ratio as block randomization with flexible block sizes of 2 and 4, using a web-based application, stratified for fertility treatment (EM, IUI, and IVF or ICSI) and participating center. Participants were randomly assigned to receive an antioxidant supplement or a placebo ( Figure ). The study was double-blinded; all medical and laboratory personnel, researchers, and patients were blinded to the treatment allocation, except for personnel responsible for randomization and distribution of medication. These persons were neither involved in clinical treatment nor the processing or analysis of study data. 20 IUI indicates intrauterine insemination. a One patient’s partner achieved an ongoing pregnancy between 0 and 3 months and a clinical pregnancy at 6 months. All participating couples received standard infertility care according to the guidelines of the Dutch Society of Obstetrics and Gynecology. Men in both treatment groups took 1 tablet of study medication daily for a total duration of 6 months. The antioxidant supplement (Impryl) and placebo tablets were produced by Labomar SRL according to Good Manufacturing Practices guidelines. The antioxidant supplement was chosen based on results from nonrandomized studies on its preceding formulation. 21 One tablet of the supplement contains betaine (200 mg), L-cystine (200 mg), niacin (16 mg), zinc (10 mg), vitamin B 6 (1.4 mg), vitamin B 2 (1.4 mg), folic acid (400 µg), and vitamin B 12 (2.5 µg). The placebo tablet and its packaging were identical in appearance to the tablet and packaging of the antioxidant supplement. Participants received online questionnaires at baseline, each month during the 6-month treatment period, and 9 months after completion of the treatment. Patients reported on relevant demographics, medical history, lifestyle factors, adverse events, and whether their partner had become pregnant. 22 Patients were asked to report their race and ethnicity, with options defined by the investigators (African, Asian, Mediterranean, White, or other [not further defined]), because there are known racial and ethnic variations in semen parameters. 23 Treatment adherence was determined by requesting participants to record the number of tablets left each month. To assess the impact of the antioxidant supplement on fertility, we established the primary outcome of the study as ongoing pregnancy, conceived within 6 months after randomization. Ongoing pregnancy was defined as a viable pregnancy at 12 weeks of gestation confirmed by ultrasonography. Follow-up of the primary outcome was completed in December 2024. Secondary short-term clinical outcomes were biochemical pregnancy rate (defined as a positive home urinary pregnancy test result), clinical pregnancy rate (defined as the presence of an intrauterine gestational sac on transvaginal ultrasonography), first-trimester pregnancy loss (defined as the loss of a pregnancy before 12 weeks’ gestation), and ectopic pregnancy rate. The prespecified outcome of ongoing pregnancy rate within 4 to 6 months after randomization was also assessed. Considering a spermatogenesis cycle of approximately 72 days, we regarded 4 to 6 months as the window of optimal treatment effect. Secondary long-term clinical outcomes were live birth rate and total pregnancy loss (defined as a first-trimester pregnancy loss and nonviable intrauterine pregnancy after 12 weeks, termination of pregnancy, and stillbirth) both within 6 months after randomization as well as cumulative number of pregnancies and time to pregnancy leading to ongoing pregnancy (defined as the time in months between randomization and a positive home urinary pregnancy test result) both within 9 months after randomization. Obstetric and neonatal outcomes included mode of delivery, gestational age at delivery, birth weight, and congenital malformations. The secondary outcomes live birth rate, total pregnancy loss, and obstetric and neonatal outcomes were not yet available and therefore not included in this article. In addition to the clinical outcomes, fertility diagnostic tests were assessed as follows. Semen parameters were measured in accordance with World Health Organization guidelines at baseline and after 3 to 6 months, when part of standard care, measuring volume, concentration, progressive motility, and total motile sperm count (TMSC). 24 We defined a binary outcome of a change of TMSC classification, potentially leading to a less invasive fertility treatment category according to Dutch guidelines. 25 In a patient subgroup of 1 participating center (Jeroen Bosch Hospital), we performed an additional semen analysis at baseline and after at least 12 weeks. In these samples, we assessed sperm DNA fragmentation (SDF) and sperm vitality as described by Punjabi et al 26 (eMethods in Supplement 2 ). Outcomes of patients treated with IVF or ICSI included fertilization rate (defined as the number of zygotes [2 pronuclei] divided by the number of inseminated or injected oocytes); embryo utilization rate (defined as the number of transferred and frozen embryos divided by the total number of zygotes); and pregnancy rate after fresh embryo transfer (ET), frozen-thawed ET, and ovum pickup (OPU). To evaluate the effect of an antioxidant supplement on ongoing pregnancy rate, we determined a sample size of 600 men per treatment group (1200 in total) was needed. We assumed differences in ongoing pregnancy rates of 6.5% in favor of the antioxidant supplement and an expected ongoing pregnancy rate of 20% in the placebo group, a 2-sided α = 5% and 76% power. A separate sample size calculation was performed for the outcome of SDF. This calculation was based on a study reporting a difference in pre- and posttreatment SDF of 6.6% with an SD of 14.2. 21 At 80% power, a 2-sided α = 5%, and with adjustment for 10% loss to follow-up, the sample size goal was 80 men. Binary clinical outcomes in the 2 treatment groups, odds ratios (ORs), and corresponding 95% CIs were estimated using a fixed-effects logistic regression model with binomial distribution, adjusting for randomization strata and female age. Couples with an ongoing pregnancy within 3 months of study treatment were excluded from the analysis of ongoing pregnancy conceived between 4 and 6 months because they were no longer at risk for achieving pregnancy. Due to small numbers, ectopic pregnancy rate was assessed with the Pearson χ 2 test. Prespecified subgroup analyses were performed for ongoing pregnancy per fertility treatment stratum. Time to pregnancy was evaluated with Kaplan-Meier estimates, which were compared with log-rank tests adjusted for fertility treatment stratum. Change in prewash TMSC, potentially leading to a less invasive treatment, was defined for each patient and compared between treatment groups using a fixed-effects binomial regression model with an identity link. Continuous semen parameters were compared using a linear regression model (when normally distributed) or a γ regression model with a log link and fixed effects (when skewed). All analyses were adjusted for randomization strata. Pretreatment and posttreatment SDF and sperm vitality were compared within the 2 groups using the Wilcoxon signed-rank test. All statistical tests were conducted using appropriate methods based on variable type. Comparisons of continuous variables were performed using unpaired, 2-tailed t tests or analysis of variance, while categorical variables were analyzed using Fisher exact test. A 2-sided hypothesis testing approach was applied, with statistical significance levels set at P < .05. We performed an intention-to-treat analysis for all outcomes, using SPSS version 29 (IBM Corp). Missing data were not imputed for any variables. To evaluate the effect of nonadherence to the study protocol, we assessed clinical outcomes in the per-protocol population. Patients were excluded if they used other antioxidant supplements at randomization or consumed less than 75% of the study medication. Patients who discontinued treatment for clinical reasons were not regarded as nonadherent. 27

We randomized 1217 patients and included 1171 patients in the primary analysis, of whom 591 were in the antioxidant supplement group and 580 were in the placebo group ( Figure ). Participants had a median (IQR) age of 34 (31-38) years, and their female partners had a median (IQR) age of 32 (30-35) years. Among these men, 10 (0.9%) self-identified as African, 28 (2.4%) as Asian, 25 (2.1%) as Mediterranean, 993 (84.8%) as White, and 74 (6.3%) as other race and ethnicity. Demographics of participants and their female partners were comparable between the treatment groups ( Table 1 ). Baseline semen analysis was available for more than 90% of participants and was also comparable between groups (eTable 1 in Supplement 2 ). Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); ICSI, intracytoplasmic sperm injection; IVF, in vitro fertilization. Numbers may not add up due to missing values: 0.3% for fertility diagnosis and ranging from 3.6% to 3.8% for all other variables. Self-reported by participants using options defined by the investigators. Not further defined. Finasteride, selective serotonin reuptake inhibitors, and valproic acid. Includes sexual disorders (vaginism or erectile dysfunction) and unknown. The primary outcome of ongoing pregnancy rate within 6 months after randomization was not significantly different between the antioxidant supplement and placebo groups (193 of 571 patients [33.8%] vs 208 of 555 patients [37.5%]; adjusted OR [AOR], 0.85 [95% CI, 0.66-1.09]; P  = .20). Rates of biochemical pregnancy, clinical pregnancy, first-trimester pregnancy loss and ectopic pregnancy within 6 months, cumulative pregnancy number within 9 months, and time to pregnancy leading to ongoing pregnancy were also similar between groups ( Table 2 ). Ongoing pregnancy rate within the window of optimal treatment effect of 4 to 6 months was significantly lower in the antioxidant supplement group compared with placebo group (69 of 446 [15.5%] vs 95 of 442 [21.5%]; AOR, 0.66 [95% CI, 0.47-0.94]; P  = .02) ( Table 2 ). Abbreviations: AOR, adjusted odds ratio; HR, hazard ratio; NA, not applicable; RR, risk ratio. Reasons for missing values for each analysis are presented in the Figure . Calculated with Pearson χ 2 due to small numbers. Considered the window of optimal treatment effect given the spermatogenesis cycle of approximately 72 days. For couples treated with IVF or ICSI, analysis revealed no significant between-group differences in fertilization rate and embryo utilization rate ( Table 3 ). Pregnancy rate was significantly lower in the antioxidant supplement group vs placebo group after fresh ET (67 of 156 [42.9%] vs 82 of 146 [56.2%]; AOR, 0.53 [95% CI, 0.33-0.86]; P  = .009) and after OPU (100 of 173 [57.8%] vs 109 of 159 [68.6%]; AOR, 0.58 [95% CI, 0.36-0.94]; P  = .03). There were no differences in pregnancy rates after frozen-thawed ET. Abbreviations: AOR, adjusted odds ratio; ET, embryo transfer; MD, mean difference; OPU, ovum pickup. Denominator does not equal the total number of patients due to missing values for female age (covariate in fixed-effects binomial model) (n = 7), frozen-thawed embryo transfer not yet performed (n = 6), vitrification of oocytes due to azoospermia (n = 1). Subgroup analyses of ongoing pregnancy rate within 6 months after randomization in the fertility treatment strata (EM, IUI, and IVF or ICSI) showed no significant between-group differences (eTable 2 in Supplement 2 ). The analyses in the per-protocol population showed no differences between treatment groups (eTable 3 in Supplement 2 ). Follow-up semen parameters were available for 191 to 209 participants (33%-34%) in the different treatment groups and for different parameters ( Table 4 ). There were no baseline imbalances between treatment groups, except for a higher percentage of female partners with unilateral tubal pathology in the antioxidant supplement group compared with the placebo group (12 [5.7%] vs 4 [1.7%]; P  = .03) (eTables 4 and 5 in Supplement 2 ). We found no significant differences in semen parameters nor in the proportion of patients that could potentially be assigned to a less invasive treatment based on TMSC classification ( Table 4 ). Abbreviations: NA, not applicable; TMSC, total motile sperm count. A total of 680 patients (58.0%) adequately adhered to treatment. Adherence did not differ between the antioxidant supplement and placebo groups (333 of 591 [56.3%] vs 347 of 580 [59.8%]; P  = .23). The most frequently patient-reported reasons for treatment discontinuation were (temporary) stop or change of fertility treatment (44 of 264 [16.7%]) and no trust in study treatment (46 of 264 [17.4%]) (eTable 6 in Supplement 2 ). Adverse event rate was low and comparable between the antioxidant supplement and placebo groups (16 of 591 [2.7%] vs 10 of 580 [1.7%]; P  = .29). The most frequently reported adverse events were gastrointestinal discomfort, fatigue, and headache (eTable 7 in Supplement 2 ). Four serious adverse events were reported, 2 in each study group. Two serious adverse events were exacerbations of a known disease. One patient experienced a temporary neurological disorder and recovered completely. Another patient was diagnosed with a hematological cancer during the trial. For assessment of sperm DNA fragmentation, we randomized 127 patients: 63 in the antioxidant supplement group and 64 in the placebo group. Sixty patients were included in the analysis after 3 months. The reasons for exclusion are presented in the eFigure in Supplement 2 . Baseline characteristics were similar in both groups (eTable 8 in Supplement 2 ). Conventional semen parameters, SDF, and sperm vitality were not significantly different between the 2 groups at baseline or after 3 months of treatment (eTable 9 in Supplement 2 ). Pretreatment and posttreatment SDF were comparable within treatment groups. Median (IQR) sperm vitality was significantly lower after treatment with the antioxidant supplement compared with the baseline value (54.9% [36.3%-72.7%] vs 62.7% [49.5%-77.3%]; P  = .03). This effect was not observed in the placebo group.

Treatment of men seeking fertility care with the combined antioxidant supplement did not improve pregnancy outcomes, semen parameters, fertilization and embryo utilization rate after IVF or ICSI, or sperm DNA fragmentation. Treatment with the antioxidant supplement resulted in a significantly lower ongoing pregnancy rate within 4 to 6 months after randomization, which was considered the window of optimal treatment effect. This effect was confirmed in patients treated with IVF or ICSI, among whom lower pregnancy rates were observed after OPU and fresh ET in the antioxidant supplement group. Our findings are consistent with those of a large randomized trial in men seeking fertility care, which showed no benefit of treatment with folic acid and zinc to pregnancy outcomes within 6 months. 28 However, our results contradict 2 meta-analyses that reported higher pregnancy rates following antioxidant treatment. 17 , 18 These meta-analyses included only men with abnormal semen parameters and were based on low-quality studies, leading the authors to call for more rigorous, high-quality evidence. To our knowledge, our trial was the first to observe a significantly lower ongoing pregnancy rate within the optimal treatment window of 4 to 6 months. An explanation for the reduced pregnancy rate in the antioxidant supplement group could be reductive stress induced by excess antioxidants. 29 , 30 Given that the antioxidant supplement contained no more than the recommended daily antioxidant dose, overdosing was unlikely. We did not measure baseline ROS or antioxidant levels or select patients based on these factors. It is, therefore, possible that our population had low oxidative stress and no actual need for antioxidant supplementation. The low baseline SDF in our patient subgroup supports this hypothesis. 24 Differences in lifestyle or male age do not appear to explain a decreased level of oxidative stress, as these parameters are consistent with other studies. 28 , 31 , 32 The findings of this study discourage the use of the combined antioxidant supplement in men seeking fertility care, as it does not improve pregnancy rates in their partner. The significantly lower ongoing pregnancy rate within 4 to 6 months of treatment might indicate an adverse effect of the antioxidant supplement in a general group of men who are part of a subfertile couple. The risk of antioxidant treatment has been repeatedly observed for cancer and cardiovascular diseases. 15 , 33 Patients and clinicians are often not aware of potential adverse effects and tend to assume that if it does not help, it does not hurt. Information on the efficacy and risks of nutritional supplements is currently not evidence-based and remains unclear to users due to complex regulation and limited surveillance. 34 , 35 We decided to publish these results before the collection and analysis of the secondary follow-up outcomes were completed for several reasons. Most important of these reasons is it does not seem ethically appropriate to delay the publication of results that show no benefit or even an adverse effect of antioxidant treatment on pregnancy outcomes. Furthermore, it is unlikely that live birth rate will differ substantially from ongoing pregnancy rate. 36 This study has several limitations. First, a heterogenous group of men seeking fertility care was included along with men with normal semen quality. A subgroup analysis of subfertile men was not conducted due to insufficient sample size. Second, treatment adherence in our study was low (58.0%) compared with the common benchmark of 80%. 37 Even though adherence thresholds differ between medication and outcomes, the limited adherence in this study might have led to an underestimation of treatment effect and adverse events. 38 Third, IVF or ICSI outcomes within 3 months after randomization, when the supplement’s effect may have been suboptimal, were also included in the results. Fourth, 46 patients were excluded from the data analysis because they did not meet the eligibility criteria. In some of the analyses, missing data reduced the sample sizes. The results of the per-protocol analysis should be interpreted with care, as many patients had to be excluded. We included only half of the randomized patients in the analysis of the SDF. Although we assessed the impact of incubation and transportation before the study, many samples still lacked a valid positive control in which DNA fragmentation was induced before assessment. Since a cause could not be identified, we continued the tests and excluded invalid samples from the analysis.

The SUMMER trial showed that treatment of men seeking fertility care with a combined antioxidant supplement did not improve ongoing pregnancy rate compared with placebo. Therefore, we do not support the use of this antioxidant supplement in these patients.

The increasing prevalence of infertility, currently estimated at 8% to 12%, leads to a high psychological burden for patients, greater need for assisted reproductive technology, and increased health care costs. 1 , 2 , 3 In approximately half of infertile couples, a male factor is partly or entirely the cause of the infertility. 4 , 5 Available treatments for infertile men, such as intracytoplasmic sperm injection (ICSI) and testicular sperm extraction, are invasive and expensive. Oxidative stress, defined by an imbalance between damaging reactive oxygen species (ROS) and antioxidant defense, is believed to play a role in the pathophysiologic processes of cancer, diabetes, cardiovascular diseases, and neurodegenerative diseases. 6 It has also been proposed as a key factor in male infertility. 7 , 8 , 9 Excessive ROS can cause lipid peroxidation of the sperm membrane, deterioration of sperm DNA, and eventually apoptosis. 10 Factors such as air pollution, advanced male age, obesity, and cigarette smoking are known to induce oxidative stress in sperm. 11 , 12 Antioxidants can neutralize or scavenge ROS and thereby prevent further damage. 13 Building on this role of antioxidants, pharmaceutical companies have marketed over-the-counter supplements that claim to improve male fertility, targeting patients seeking alternative ways to improve their chances to conceive. The use of these supplements has increased in the past decade, driving the need for robust evidence on their efficacy and safety. 14 In other health conditions, such as cancer and cardiovascular diseases, antioxidant use has been associated with adverse effects. 15 , 16 Mechanisms underlying these adverse effects include reductive stress, which occurs when the redox balance shifts due to excessive intake of antioxidants, and the potential for antioxidants to act as pro-oxidants at high doses. The effect of antioxidants on male fertility is also subject to debate, with clinical trials reporting either a favorable effect or no effect on pregnancy rates. 17 , 18 , 19 Moreover, most available studies have substantial methodological limitations. In this multicenter randomized clinical trial, we aimed to assess whether treatment of men seeking fertility care with an antioxidant supplement can improve semen quality and pregnancy rates compared with a placebo.