The effects of male anxiety and depression on IVF outcomes

Abstract

STUDY QUESTION

What are the effects of male anxiety and depression on IVF outcomes?

SUMMARY ANSWER

Men with anxiety had lower final total motile sperm counts (fTMSC) during IVF compared to men without anxiety; however, there were no differences in live birth rates (LBRs).

WHAT IS KNOWN ALREADY

Studies have shown that male anxiety causes low sperm motility, worse sperm morphology, and increased DNA fragmentation, which are known to be influential factors on fertilization rates and embryo quality during IVF. However, data are lacking on whether there is a direct association between male anxiety and/or depression and IVF outcomes.

STUDY DESIGN, SIZE, DURATION

This was a survey-based, retrospective cohort study completed at a single, large hospital-affiliated fertility center with 222 respondents who underwent IVF with or without ICSI. The study was conducted between 6 September 2018 and 27 December 2022.

PARTICIPANTS/MATERIALS, SETTING, METHODS

Male partners of couples who underwent IVF or IVF/ICSI completed a Hospital Anxiety and Depression Scale (HADS) questionnaire. They were separated into two groups for both anxiety (HADS-A ≥ 8 or HADS-A < 8) and depression (HADS-D ≥ 8 or HADS-D < 8). Men with an elevated HADS-A or HADS-D score ≥8 were considered to have anxiety or depression, respectively. The primary outcome was LBR. Secondary outcomes included semen parameters at the time of IVF, cycle outcomes, pregnancy outcomes, and prevalence of erectile dysfunction and low libido.

MAIN RESULTS AND THE ROLE OF CHANCE

There were a total of 222 respondents, of whom 22.5% had a HADS-A ≥ 8 and 6.5% had a HADS-D ≥ 8. The average age of respondents was 37.38 ± 4.90 years old. Antidepressant use was higher in the respondents with a HADS-A or HADS-D ≥ 8 (P < 0.05). Smoking use was similar between groups for both HADS-A and HADS-D (P > 0.05). When adjusted for male BMI, antidepressant use and smoking, men with a HADS-A or HADS-D ≥ 8 had similar rates of erectile dysfunction (adjusted relative risk (aRR) = 1.12 (95% CI 0.60, 2.06)) and low libido (aRR = 1.70 (95% CI 0.91, 3.15)) compared to those with a HADS-A or HADS-D ≤ 8. Men with a HADS-A ≥ 8 were more likely to have a lower fTMSC on the day of oocyte retrieval (11.8 ≥ 8 vs 20.1 < 8, adjusted ß = −0.66 (95% CI −1.22, −0.10)). However, the LBR per embryo transfer (ET) was similar between the HADS-A groups (43.2% ≥8 vs 45.1% <8, adjusted relative risk = 0.90 (95% CI 0.65, 1.06)). Although depression was uncommon in the entire cohort, the HADS-D groups were clinically similar for fTMSC (18.7 ≥ 8 vs 16.0 < 8) and LBR per ET (46.7% ≥8 vs 45.4% <8).

LIMITATIONS, REASONS FOR CAUTION

Limitations of our study are the survey-based design, the lack of sperm morphology assessment at the time of IVF, our inability to fully assess the HADS-D ≥ 8 cohort due to the small sample size and the large Caucasian demographic.

WIDER IMPLICATIONS OF THE FINDINGS

Couples undergoing IVF have an increased likelihood of suffering from anxiety and/or depression. There is currently a debate on whether or not men should be treated with antidepressants while attempting to conceive due to potential detrimental effects on sperm quality. Our study shows that, regardless of antidepressant use, couples with men who did or did not report anxiety and/or depression have similar LBRs when undergoing IVF. Therefore, it is important to assess both partners for mental health and to not withhold treatment due to a concern about a potential impact of antidepressants or anxiety/depression on sperm quality.

STUDY FUNDING/COMPETING INTEREST(S)

There was no funding to report for this study. Z.W. is a contributing author for UptoDate. S.S.S. is on the advisory board for Ferring Pharmaceuticals. E.G. was a medical consultant for Hall-Matson Esq, Teladoc, and CRICO and is a contributing author for UptoDate. The remaining authors have nothing to report.

TRIAL REGISTRATION NUMBER

N/A.

anxiety, depression, HADS, IVF, live birth rates, semen parameters

Introduction

Male factor infertility is estimated to be a cause of infertility in up to 50% of infertile couples and is the sole diagnosis in 20–30% of cases (Agarwal et al., 2015). Several factors may contribute to male factor infertility, including smoking (Harlev et al., 2015), alcohol use (Ricci et al., 2017), marijuana use (Gundersen et al., 2015), diet (Danielewicz et al., 2018), and obesity (Sermondade et al., 2013). The impact of psychological stress, such as depression and anxiety, on male fertility has also gained increasing attention. Occupational and life stressors are associated with worse semen parameters, including low motility, worse sperm morphology, and increased sperm DNA fragmentation (Eskiocak et al., 2005; Janevic et al., 2014; Jurewicz et al., 2014). In addition, stress is thought to adversely impact the hypothalamus–pituitary–gonad axis and contribute to oxidative stress, thereby impacting sperm DNA fragmentation, motility, and morphology, as well as fertilization (Eskiocak et al., 2005; Nargund, 2015; Gualtieri et al., 2021).

Depression and anxiety have adverse effects on semen quality. Male partners in infertile couples with a Hospital Anxiety and Depression Scale (HADS) score ≥8 for anxiety (HADS-A) and depression (HADS-D), as an indicator of psychological stress, have lower levels of testosterone, higher levels of FSH and LH, and lower sperm counts and motility compared to those with a HADS score <8 (Bhongade et al., 2015). Similar relationships are seen with the severity of depression determined by the Beck Depression Inventory (BDI), as men with higher BDI scores have lower semen volume and motility (Ye et al., 2022).

Both male and female partners have increased rates of depression when undergoing IVF for the treatment of infertility. Studies have investigated the effects of female physiological and psychological stress on IVF outcomes and have found no association with clinical pregnancy rates (Cesta et al., 2018; Miller et al., 2019), but a possible association with miscarriage rates (Geisler et al., 2020); however, these data are lacking for male partners with physiological or psychological stress. Studies have shown that male anxiety causes low sperm motility, worse sperm morphology, and increased DNA fragmentation (Janevic et al., 2014; Jurewicz et al., 2014; Eskiocak et al., 2005), which are known to be influential factors on fertilization rates (Dcunha et al., 2022) and embryo quality (Cissen et al., 2016; Colaco and Sakkas, 2018). However, data are lacking on whether there is a direct correlation between male anxiety and/or depression on IVF outcomes. There is a need for further investigation into the potential relationship between the impact of male anxiety and depressive symptoms and IVF or IVF with ICSI (IVF/ICSI) outcomes. The objective of this study was to investigate the relationship between men with and without HADS ≥8 and their respective semen parameters and IVF outcomes. We hypothesized that men with an abnormal HADS-A or HADS-D score (≥8) would have lower semen parameters and live birth rates (LBRs) compared to their counterparts with normal HADS-A or HADS-D scores (<8).

Materials and methods

Study participants

A voluntary survey was given to couples who were undergoing IVF at a single, large hospital-affiliated fertility center from 6 September 2018 to 27 December 2022. The survey included a Hospital Anxiety and Depression Scale (HADS) questionnaire. Men who completed either the HADS anxiety (HADS-A) or HADS depression (HADS-D) portions of the questionnaire were included in the study. Exclusion criteria included not completing either the HADS-A or HADS-D questionnaire, use of frozen or donor sperm, prior chemotherapy or radiation exposure, prior vasectomy, or use of sperm obtained by testicular sperm extraction (TESE) or epididymal sperm aspiration.

There were 660 men who were approached, of whom 414 agreed and started the survey (63% response rate). There were 240 men who completed the HADS-A or HADS-D questionnaire (58% completion rate). Then 18 were excluded based on exclusion criteria leaving 222 respondents available for analysis. Groups were stratified based on male partner questionnaire responses into those with anxiety (HADS-A ≥ 8, N = 49) and without anxiety (HADS-A < 8, N = 169) as well as those with depression (HADS-D ≥ 8, N = 14) and without depression (HADS-D < 8, N = 201).

The primary outcome was LBRs between men with either elevated (≥8) HADS-A or HADS-D compared to those with normal (<8) HADS-A or HADS-D scores. Secondary outcomes included evaluation of semen parameters at the time of oocyte retrieval (semen volume, concentration, total motility, and total motile sperm count both pre- and post-processing), cycle outcomes, pregnancy outcomes, and prevalence of erectile dysfunction and low libido. This study was approved by the Mass General Brigham Institutional Review Board (Protocol # 2022P001043).

Survey design

The survey was split into several categories: general information, hormones and fertility history, medical history, medications, family history, and lifestyle. The HADS questionnaire was included within the lifestyle section of the survey. The HADS questionnaire is a validated tool used to assess anxiety and depression. It is divided into two parts, HADS-A and HADS-D, which contain seven intermingled questions. The HADS questionnaire was first developed in 1983 to identify the caseness (possible and probable) of anxiety disorders and depression in a non-psychiatric setting. Since that time, it has been under review for validation and has continually performed similarly or better when compared to other commonly used validation tools assessing anxiety and/or depression (Bjelland et al., 2002) (Supplementary Fig. S1).

Sperm handling and assessment

Semen specimens were collected by masturbation using a self-collection kit at home or within the hospital. Men were asked to adhere to a 48- to 72-h abstinence period prior to specimen collection. Specimens were evaluated within 1–2 h of production. Specimens were collected in a sterile cup and allowed to liquefy at 37°C for 15–30 min prior to processing.

Semen volume was measured using a calibrated pipette. Specimen viscosity and presence of debris were assessed and documented if abnormalities were noted. The raw specimen (approximately 6 µl) was placed on a Leja^® counting 2 chamber slide (IMV Technologies Group, France) and inserted into an automated semen analyzer (Hamilton Thorne, Danvers, MA, USA) to quantify sperm concentration and total motility. Sperm washing was done using density gradient centrifugation per a standard protocol (Zhao et al., 2016) with PureCeption™ (In Vitro Fertilization, Inc., USA), density gradient and HEPES-sperm wash (CooperSurgical, Inc., Denmark). Repeat evaluation of sperm volume, concentration, and total motility was calculated after sperm washing. Initial total motile sperm count and final total motile sperm count (fTMSC) (volume × concentration × percent total motility) were calculated for each specimen. We do not routinely assess sperm morphology at the time of oocyte retrieval; however, we did analyze semen morphology in the most recent semen analysis for each respondent and did not find a statistically significant difference between respondents with and without teratozoospermia (<4% based on Kruger criteria) between groups with high or normal HADS scores (P = 0.9621) (Table 1).

Table 1.

Baseline characteristics, n (%).

	HADS-A < 8	HADS-A ≥ 8	HADS-D < 8	HADS-D ≥ 8
	N = 169	N = 49	N = 201	N = 14
Male BMI
<18.5	2 (1.21)	0 (0)	2 (1.02)	0 (0)
18.5–24.9	42 (25.45)	11 (22.92)	51 (26.02)	0 (0)
25.0–29.9	83 (50.30)	21 (43.75)	98 (50.00)	5 (35.71)
30.0–39.9	33 (20.00)	12 (25.00)	38 (19.39)	7 (50.00)
40+	5 (3.03)	4 (8.33)	7 (3.57)	2 (14.29)
Missing	4	1	5	0
Race
White	136 (80.47)	43 (87.76)	164 (81.59)	13 (92.86)
Black or African American	3 (1.78)	0 (0)	3 (1.49)	0 (0)
Asian	23 (13.61)	6 (12.24)	26 (12.94)	1 (7.14)
Other	7 (4.14)	0 (0)	8 (3.98)	0 (0)
Ethnicity
Not Hispanic	158 (94.05)	47 (97.92)	191 (95.98)	12 (85.71)
Hispanic	10 (5.95)	1 (2.08)	8 (4.02)	2 (14.29)
Missing	1	1	2	0
Prior paternity
Yes	116 (68.64)	30 (61.22)	134 (66.67)	10 (71.43)
No	53 (31.36)	19 (38.78)	67 (33.33)	4 (28.57)
Smoking history
Yes	49 (29.17)	17 (35.42)	61 (30.65)	6 (42.86)
No	119 (70.83)	31 (64.58)	138 (69.35)	8 (57.14)
Missing	1	1	2	0
Alcohol history
Yes	148 (88.10)	45 (93.75)	178 (89.45)	12 (85.71)
No	20 (11.90)	3 (6.25)	21 (10.55)	2 (14.29)
Missing	1	1	2	0
Marijuana use
Yes	32 (19.05)	7 (14.29)	36 (18.00)	5 (35.71)
No	136 (80.95)	42 (85.71)	164 (82.00)	9 (64.29)
Missing	1	0	1	0
Physically active
Yes	79 (46.75)	20 (40.82)	89 (44.28)	8 (57.14)
No	90 (53.25)	29 (59.18)	112 (55.72)	6 (42.86)
Antidepressant use
Yes	7 (4.14)	6 (12.24)	10 (4.98)	3 (21.43)
No	162 (95.86)	43 (87.76)	191 (95.02)	11 (78.57)
Varicocele
Yes	9 (5.36)	2 (4.08)	11 (5.50)	0 (0)
No	159 (94.64)	47 (95.92)	189 (94.50)	14 (100)
Missing	1	0	1	0
History of PID/STI
Yes	23 (13.94)	2 (4.08)	24 (12.18)	1 (7.14)
No	142 (86.06)	47 (95.92)	173 (87.82)	13 (92.86)
Missing	4	0	4	0
Prior testosterone use
Yes	7 (4.17)	2 (4.08)	7 (3.52)	1 (7.14)
No	161 (95.83)	47 (95.92)	192 (96.48)	13 (92.86)
Missing	1	0	2	0
Baseline sperm morphology*
Abnormal	32 (22.86)	9 (22.50)	38 (22.62)	3 (25.00)
Normal	108 (77.14)	31 (77.50)	130 (77.38)	9 (75.00)
Missing	29	9	33	2
Female partner race
White	129 (78.66)	42 (87.50)	158 (81.03)	11 (78.57)
Black or African American	3 (1.83)	1 (2.08)	4 (2.05)	1 (7.14)
Asian	27 (16.46)	5 (10.42)	28 (14.36)	2 (14.29)
Other	5 (3.05)	0 (0)	5 (2.56)	0 (0)
Missing	5	1	6	0
Female partner BMI
18.5–24.9	92 (54.76)	28 (57.14)	112 (56)	7 (50)
25.0–29.9	47 (27.98)	10 (20.41)	53 (26.50)	2 (14.29)
30.0–39.9	22 (13.10)	6 (12.24)	26 (13)	3 (21.43)
40+	7 (4.17)	5 (10.20)	9 (4.50)	2 (14.29)
Missing	1	0	1	0
Infertility diagnosis
Diminished ovarian reserve	26 (15.38)	9 (18.37)	33 (16.42)	0 (0)
Uterine factor	1 (0.59)	0 (0)	1 (0.50)	0 (0)
Ovulatory dysfunction	8 (4.73)	3 (6.12)	11 (5.47)	0 (0)
Tubal factor	5 (2.96)	1 (2.04)	6 (2.99)	0 (0)
Endometriosis	5 (2.96)	3 (6.12)	7 (3.48)	1 (7.14)
Male factor	16 (9.47)	10 (20.41)	22 (10.95)	3 (21.43)
Unexplained	51 (30.18)	14 (28.57)	59 (29.35)	6 (42.86)
Other	27 (15.98)	5 (10.20)	30 (14.93)	2 (14.29)
Male factor + ≥1 more	17 (10.06)	2 (4.08)	17 (8.46)	2 (14.29)
Multiple female factors	13 (7.69)	2 (4.08)	15 (7.46)	0 (0)

*

Based on Kruger criteria: <4% = abnormal.

HADS-A: Hospital Anxiety and Depression Scale-Anxiety; HADS-D: Hospital Anxiety and Depression Scale-Depression; PID: pelvic inflammatory disease; STI: sexually transmitted infection.

Hormonal assessment

We retrospectively evaluated data on early morning serum FSH, luteinizing hormone, total testosterone, and free testosterone that were obtained within one year of respondents’ completion of the survey. We performed a sub-analysis to assess the possible correlation between mental health, hormonal status, and effects on libido and erectile dysfunction. Respondents with total testosterone <300 ng/dl on morning assay (per the AUA/ASRM testosterone deficiency guidelines (Schlegel et al., 2021)) and/or the use of clomiphene citrate/human chorionic gonadotropin/anastrozole as an on-going treatment for previously clinically diagnosed hypogonadism were excluded from the analysis for low libido and erectile dysfunction.

IVF stimulation

Female partners of male respondents underwent conventional ovarian stimulation per clinical routine. A combination of recombinant FSH [Gonal-f, EMD Serono (Rockland, MA) or Follistim, Merck (Kenilworth, NJ)] and urinary FSH [Menopur, Ferring (Parsippany, NJ)] was started on Day 2 of menses. Serial ultrasound monitoring and blood tests were done to monitor progress. Daily GnRH antagonist (250 µg Cetrorelix or Ganirelix acetate) was started when the estradiol level was >300 pg/ml or the mean diameter of the largest follicle was 14 mm. Stimulation continued until a median diameter of two follicles reached ≥18 mm, at which a ovulation trigger was administered (human chorionic gonadotropin, leuprolide acetate, or both) to induce final oocyte maturation. The oocyte retrieval was performed 36 h after the trigger.

Assisted reproductive technology

Intracytoplasmic sperm injection was performed by a trained embryologist using standard laboratory protocols (Van Steirteghem et al., 1993). Fertilization was checked 16–18 h after insemination. Embryos were checked on Days 3, 5, 6, and 7. Cleavage-stage embryos were graded based on number of cells, fragmentation, and symmetry (Racowsky et al., 2010). Blastocysts were graded on a modified Gardner grading scale (Bakkensen et al., 2019) (Supplementary Fig. S2). Top quality cleavage-stage embryos were defined as ≥8 cells, grade 0 or 1 fragmentation, and grade 1 or 2 symmetry. Top quality blastocysts were defined as stages 5 (expanding blastocyst) to 8 (hatching blastocyst) with inner cell mass and trophectoderm grades of A or B (e.g., 7Aa, 7Ba), except for 5Bb (equivalent to 2Bb with Gardner grading). The number of embryos transferred was based on internal institution guidelines and consistent with the ASRM guidelines for embryo transfer (ET) (Penzias et al., 2017).

Preimplantation genetic testing

The euploidy rate in couples who underwent preimplantation genetic testing for aneuploidy or monogenetic disorders (PGT-A or PGT-M) of supernumerary embryos was also assessed. Embryos destined for PGT underwent laser-assisted hatching on Day 3. They were evaluated and graded on Days 5–7 to determine if they were of top quality, as stated previously, for biopsy and vitrification. Embryos that did not meet the criteria for biopsy or vitrification were discarded.

Statistical analysis

Chi-squared tests were used for descriptive statistics. Physical activity was assessed based on CDC recommendations (‘How much physical activity do adults need?’, 2022). Respondents who did not meet these recommendations were categorized as not active (Table 1). Relative risk and 95% CI were produced with log-binomial regression for dichotomous outcomes, and Poisson regression was used for counts. Poisson regression with an offset was used for ratios. Sperm parameter data were log-transformed and linear regression models were performed. Models were adjusted for female age, female BMI, infertility diagnosis, male antidepressant use, male smoking, and day of transfer. Generalized estimating equations were used to account for patients contributing more than one record. An alpha of 0.05 was considered statistically significant. Chi-squared tests were calculated using an online calculator (Social Science Statistics, 2018). All other statistical analyses were performed with SAS^® version 9.4 (Cary, NC, USA).

Results

Study population

We analyzed 222 respondents who met the inclusion criteria. The mean age of all respondents was 37.38 ± 4.90 years. There were 22.5% respondents with a HADS-A ≥ 8 and 6.5% respondents with a HADS-D ≥ 8. Respondents with elevated HADS scores were compared to their normal counterparts for anxiety and depression separately. Baseline characteristics can be seen for HADS-A < 8 vs HADS-A ≥ 8 and HADS-D < 8 vs HADS-D ≥ 8 in Table 1. Antidepressant use was higher in respondents with a HADS-A or HADS-D ≥ 8 when compared to their normal counterparts (P < 0.05). Smoking use, marijuana use, male BMI, physical activity, and sperm morphology were similar between the groups for HADS-A < 8 and HADS-A ≥ 8 as well as for HADS-D < 8 and HADS-D ≥ 8 (P > 0.05).

Semen parameters

We evaluated the semen parameters seen at the time of oocyte retrieval between those with normal scores (<8) and those with elevated HADS-A or HADS-D scores (≥8). Between respondents with HADS-A < 8 and HADS-A ≥ 8, we found differences in the final concentration (adjusted ß (aß) = −0.41, 95% CI: −0.67, −0.15) and fTMSC (aß = −0.66, 95% CI: −1.22, −0.10) (Table 2). There were no other statistically significant differences seen in sperm parameters between those with and without anxiety. Given the small sample size for those with depression (HADS-D ≥ 8), we were unable to perform regression models for that cohort. However, the fTMSC appeared similar between men with and without depression (18.65 ± 25.3 HADS-D < 8 vs 16.02 ± 15.2 HADS-D ≥ 8) (Supplementary Table S1).

Table 2.

Association of Hospital Anxiety and Depression Scale-Anxiety (HADS-A) score and semen parameters, n ± SD.

	HADS-A < 8	HADS-A ≥ 8	Unadjusted ß estimate (95% CI)	Adjusted ß estimate* (95% CI)
	N = 169	N = 49
Initial total volume (ml)	2.78 ± 1.4	2.31 ± 1.2	−0.19 (−0.38, 0.00)	−0.13 (−0.32, 0.07)
Initial concentration (M/ml)	60.56 ± 44.7	59.60 ± 50.7	−0.06 (−0.30, 0.18)	−0.09 (−0.33, 0.16)
Initial motility (%)	50 ± 25	52 ± 26	−0.03 (−0.25, 0.19)	−0.09 (−0.32, 0.14)
Initial total motile sperm count (total × volume × motile)	91.30 ± 96.9	71.28 ± 74.5	−0.28 (−0.70, 0.13)	−0.31 (−0.74, 0.11)
Final total volume (ml)	0.89 ± 1.1	0.61 ± 0.6	−0.21 (−0.54, 0.13)	−0.22 (−0.56, 0.13)
Final concentration M/ml	22.10 ± 11.8	18.73 ± 12.1	−0.34 (−0.59, -0.08)	−0.41 (−0.67, -0.15)
Final motility (%)	87 ± 18	85 ± 19	−0.01 (−0.12, 0.10)	−0.02 (−0.15, 0.08)
Final total motile sperm count (total × volume × motile)	20.08 ± 26.5	11.81 ± 13.6	−0.55 (−1.10, -0.01)	−0.66 (−1.22, -0.10)

*

Adjusted for female age, female BMI, male BMI, infertility diagnosis, antidepressant use, physically active, marijuana use, and smoking.

Cycle outcomes

We evaluated the IVF cycle outcomes between those with and without anxiety or depression. For anxiety, female partners of male respondents with a HADS-A ≥ 8 had higher numbers of oocytes retrieved (adjusted relative risk (aRR) = 1.12, 95% CI: 1.02, 1.21), mature oocytes (MIIs) (aRR = 1.17, 95% CI: 1.06, 1.29), fertilized oocytes (2PNs) (aRR = 1.19, 95% CI: 1.06, 1.32), and biopsied blastocysts (aRR = 1.45, 95% CI: 1.04, 2.02) (Table 3). There were no differences in fertilization rates or blastulation rates. The findings for IVF cycle outcomes between respondents with and without depression can be seen in Supplementary Table S2.

Table 3.

Comparison of cycle outcomes based on Hospital Anxiety and Depression Scale-Anxiety (HADS-A) score, n ± SD.

	HADS-A < 8	HADS-A ≥ 8	Unadjusted Relative risk (95% CI)	Adjusted Relative risk* (95% CI)
	N = 169	N = 49
Number of oocytes	15.96 ± 10.8	16.80 ± 9.4	1.05 (0.97, 1.14)	1.12 (1.02, 1.21)
Number of MIIs	12.02 ± 8.8	12.98 ± 7.6	1.08 (0.99, 1.18)	1.17 (1.06, 1.29)
Fertilization rate (%)	79.37	79.43	1.00 (0.91, 1.11)	1.02 (0.95, 1.09)
Number of 2PNs	9.54 ± 7.6	10.31 ± 6.7	1.08 (0.98, 1.19)	1.19 (1.06, 1.32)
Blastulation rate (%)	42.14	39.57	0.94 (0.80, 1.10)	0.95 (0.80, 1.13)
Number of usable blastocysts	4.02 ± 4.6	4.08 ± 4.1	1.01 (0.87, 1.19)	1.11 (0.93, 1.32)
Number of embryos biopsied	5.45 ± 4.2	5.82 ± 4.6	1.07 (0.80, 1.42)	1.45 (1.04, 2.02)
Number of euploid embryos	2.42 ± 2.2	2.64 ± 2.4	1.09 (0.71, 1.66)	NA
Euploidy rate (%)	44.40	45.36	1.02 (0.67, 1.56)	NA

MII: metaphase II; 2PNs: 2 pronuclei; NA: not applicable.

*

Adjusted for female age, female BMI, male BMI, infertility diagnosis, antidepressant use, physically active, smoking, and marijuana use.

Live birth rates

There was a total of 316 ETs in the HADS-A group (n = 235 < 8 vs n = 81 ≥ 8). There were no statistically significant differences between the groups for transfer type (fresh vs frozen) or day of transfer (Day 3 vs Day 5/6) (data not shown). However, we controlled for the day of transfer when analyzing pregnancy outcomes given the known differences between Day 3 and Day 5/6 transfers, while also controlling for female age, female BMI, infertility diagnosis, and male antidepressant use. There were no differences in the LBR per ET between the HADS-A groups (45.11% <8 vs 43.21% ≥8; aRR= 0.83, 95% CI: 0.65, 1.06) (Table 4). We were unable to perform regression analyses on the HADS-D group due to the sample size, but the LBR per ET was 45.36% in the HADS-D < 8 vs 46.67% in the HADS-D ≥ 8 (Supplementary Table S3).

Table 4.

Pregnancy outcomes based on Hospital Anxiety and Depression Scale-Anxiety (HADS-A) score, n (%).

	HADS-A < 8	HADS-A ≥ 8	Unadjusted Relative risk (95% CI)	Adjusted Relative risk* (95% CI)
	N = 235	N = 81
Implantation rate (%)	41.35	48.08	1.16 (0.92, 1.47)	NA
Chemical pregnancy	25 (10.64)	7 (8.64)	0.81 (0.38, 1.75)	NA
Ectopic pregnancy	2 (0.85)	1 (1.23)	1.45 (0.14, 15.53)	NA
Miscarriage	23 (9.79)	10 (12.35)	1.26 (0.58, 2.73)	NA
Clinical pregnancy	109 (46.38)	36 (44.44)	0.96 (0.75, 1.23)	0.91 (0.73, 1.12)
Live birth (per ET)	106 (45.11)	35 (43.21)	0.96 (0.75, 1.23)	0.83 (0.65, 1.06)

ET: embryo transfer; NA: not applicable.

*

Adjusted for female age, female BMI, male BMI, infertility diagnosis, antidepressant use, and day of embryo transfer.

Erectile dysfunction and low libido

We evaluated the prevalence of erectile dysfunction and low libido between all respondents with a HADS <8 (n = 161) vs HADS ≥8 (n = 51) for either anxiety (HADS-A) or depression (HADS-D). Men with hypogonadism were removed from the analysis. We controlled for male BMI, antidepressant use, and smoking. We found no differences in erectile dysfunction (aRR = 1.12, 95% CI: 0.60, 2.06) or low libido (aRR = 1.70, 95% CI: 0.91, 3.15) between groups.

Discussion

Significant findings

There was an overall low prevalence of anxiety and depression within our cohort based on HADS score. Couples with men who had HADS-A scores ≥8 were more likely to have a low fTMSC but were also more likely to have a higher number of oocytes retrieved, number of MIIs, number of 2PNs, and number of blastocysts biopsied. There were no differences in LBR per ET between HADS-A < 8 and HADS-A ≥ 8 groups. The HADS-D < 8 and HADS-D ≥ 8 groups were clinically similar in all aspects.

Effects of anxiety on male fertility

Psychological stress is one of the potential causes for idiopathic infertility (Szkodziak et al., 2020). Prior studies have evaluated the effects of stress on semen quality (Giblin et al., 1988; Clarke et al., 1999; Auger et al., 2001; Pook et al., 2004; Bhongade et al., 2015), revealing lower semen volume, sperm concentration, motility, and morphology. However, these studies are difficult to compare due to variations in their assessment of stress. For example, a study by Vellani et al. evaluated semen quality using the State-Trait Anxiety Inventory Y questionnaire in 94 first-attempt IVF patients compared to age-matched controls (Vellani et al., 2013). They found that men with increased state and trait anxiety had lower semen volume, lower sperm concentration, lower motility, and increased sperm DNA fragmentation in IVF patients compared to men without anxiety. However, this study did not attempt to correlate their findings with IVF outcomes. Another study that utilized the 7-item Generalized Anxiety Disorder (GAD-7) scale found that men with an abnormal GAD-7 scale had a significantly lower sperm count, lower sperm concentration and lower sperm motility when compared to men with normal GAD-7 scores (Pan et al., 2022). Similar to the previous study, they did not attempt to correlate these findings with IVF outcomes. Our study adds to the limited research that associates semen quality affected by male anxiety and/or depression, while controlling for antidepressant medications, to IVF outcomes and is noteworthy in that regard.

Bhongade et al. utilized the HADS survey in their study and found lower levels of testosterone, higher levels of FSH and LH, and lower sperm counts and motility in those with scores >8 compared to those with a HADS score of <8. However, they did not separately assess individuals based on having either an elevated HADS-A or HADS-D score and they did not evaluate IVF outcomes. In addition, they had a relatively small sample size (n = 19) with elevated HADS scores. In comparison, our study had a larger sample size and analyzed HADS-A and HADS-D scores separately. We believe this to be a strength of our study as the HADS questionnaire was validated to interpret anxiety and depression scores separately. In addition, our study found similar initial concentrations, total motility, and total motile sperm in pre-wash samples for the elevated HADS-A and HADS-D cohorts when compared to their normal counterparts. However, post-wash samples did show a statistically significant difference in fTMSC for men with an elevated HADS-A. We were unable to statistically assess this variable in the HADS-D cohort due to the small sample size.

Effects of depression on male fertility

The prevalence of depression among infertile men varies widely from 5% to 38% (Volgsten et al., 2008; Gao et al., 2013). The etiology of infertility-related depression is likely multifactorial, and may in part be due to a male’s feelings of inadequacy or low self-esteem, as well as the number of ART cycles (Reis et al., 2013; Péloquin et al., 2018). A contemporary survey-based study of infertile Chinese men demonstrated that, in the setting of infertility, erectile dysfunction and presence of a varicocele are independent risk factors for depression, and the type of ART received was not a risk factor for depression (Yang et al., 2017). In our study, we were unable to make any comparisons based on the presence or absence of depression due to the small sample size of men with HADS-D ≥ 8. However, we found there was no difference in the rate of erectile dysfunction based on HADS-A or HADS-D score.

While the duration of ART and male partner coping strategies may impact levels of depression (Babore et al., 2017; Yang et al., 2017), and depression itself may impact the quality of sperm (Bhongade et al., 2015; Ye et al., 2022), it is unclear if and to what extent depression impacts the outcomes of ART. A cross-sectional study in China including infertile men did not demonstrate a significant impact of depression or anxiety scores (as measured by Zung Self-Rating Scales) throughout each stage of IVF on IVF outcomes. This study did not control for antidepressant use (Liu et al., 2021). However, a randomized controlled trial of men with major depression, defined by Patient Health Questionnaire score ≥10, suggested major depression among male partners may lower the likelihood of clinical pregnancy in couples using ovulation induction only or ovulation induction with intrauterine insemination (Evans-Hoeker et al., 2018). In contrast, we did not find a statistical difference in LBR per ET between men with and without anxiety. Similarly, we did not see a difference in LBR per ET in men with and without depression, but our sample size was too small to make any statistical comparisons.

Effects of antidepressant use on male fertility

Antidepressants are one of the most common types of prescribed medications in the US and are commonly used to treat anxiety, depression, or other psychological disorders in men of reproductive age. The effects of antidepressants on male fertility are still unclear. Studies have shown that the potential risks of antidepressant use on men wishing to conceive can cause sexual dysfunction, late-onset hypogonadism, and abnormal sperm DNA fragmentation (Beeder and Samplaski, 2020). However, these effects have been shown to be reversible with cessation of the medication. Additionally, studies have not shown the proposed side effects of antidepressants of male fertility effects patients who are utilizing IVF. Untreated depression or anxiety can have deleterious effects on a person’s well-being, social functioning, and ability to manage relationships. Family building in couples with infertility causes a strain on mental health and it would behoove physicians to mitigate this stress with medications if indicated after formalized diagnostic assessment (Cesta et al., 2018). Our study supports the safety of antidepressant use among men with anxiety or depression utilizing IVF for family building.

A recent literature review discussed how selective serotonin inhibitors (SSRIs) can decrease sperm concentration and morphology (Milosavljević et al., 2022). Tanrikut et al. conducted a prospective study on the effects of paroxetine on sperm quality in humans (Tanrikut et al., 2010). After 5 weeks of therapeutic paroxetine exposure and a 1-month washout period, there were no differences seen in semen parameters when compared to semen parameters prior to treatment. However, there was a significant increase in sperm DNA fragmentation, which suggests that SSRIs may still have an adverse impact on sperm quality in the presence of normal concentration, motility, and morphology.

Serotonin and norepinephrine reuptake inhibitors are another class of medication used for anxiety and/or depression. Very few studies have investigated their effects on male fertility. A randomized controlled trial investigated the impact of duloxetine on sperm parameters and DNA fragmentation (Punjani et al., 2021). Men were randomized to either 60 mg of duloxetine or placebo for 6 weeks (5-week full dose with 1-week taper). The authors found no significant difference in bulk semen parameters or DNA fragmentation. Therefore, we can presume that there may be a class of antidepressants that are safer for use by men whose partners are trying to conceive. However, until further evidence and research are conducted, there continues to be a debate about whether antidepressants should be prescribed among men while their partners are trying to conceive (Amano, 2020).

Our study helps to address this question. We controlled for antidepressant use when analyzing semen parameters and IVF outcomes. We found similar initial semen parameters and LBRs between men with and without anxiety, despite lower fTMSC in men with anxiety. We were unable to assess the same finding in men with depression due to small sample size. Also, we did not have data to assess DNA fragmentation. Nevertheless, we believe our data helps support the use of antidepressants, when clinically necessary, in men undergoing IVF or IVF/ICSI.

We believe it would be useful for fertility practices to include some opportunities to assess mental health for their patients. We found it useful to have patients complete the HADS survey at the time of their partner's oocyte retrieval since it is a period of time that requires both partners to be available. Another way to assess mental health would be through an intake questionnaire for new patients, which can be administered in advance of a clinic visit via a patient-access portal or by trained medical support staff during the patient’s visit with their male or female fertility specialist.

Strengths and limitations

To our knowledge, this is the first study to utilize the HADS questionnaire in male infertility patients and evaluate the association with IVF semen parameters, cycle outcomes, pregnancy outcomes, erectile dysfunction, and low libido. We were able to stratify the results based on either an elevated HADS-A or HADS-D, which has not been done previously for this evaluation. Additionally, we were able to control for confounding factors, such as antidepressant use and smoking, which has not been commonly done in prior studies. There are limitations to our study. The results are affected by the potential biases related to a survey-based study including recall bias; however, given that our study is based on the HADS survey which evaluates the participants’ feelings at the time of assessment on the day of IVF, recall bias may not be a key factor. Given that the patients were self-selecting to participate in our study increases the risk of selection bias. Also, our technique for sperm assessment is not a formal semen analysis but an assessment of sperm quality for use in IVF; and we did not assess for normal sperm morphology at the time of oocyte retrieval. Therefore, our evaluation of sperm quality differs from Bjorndahl et al. guidelines (Björndahl et al., 2016) on the assessment of semen analyses which could make our results less generalizable to other IVF practices that routinely perform a full semen analysis at the time of oocyte retrieval. However, we did not find a statistically significant difference in baseline presence of teratozoospermia between the groups. Furthermore, our study population was primarily Caucasian, so the results of our study may not be applicable to other races. Also, we did not have a large enough sample size to statistically evaluate respondents with depression (HADS-D ≥ 8) and therefore cannot make any conclusions on that cohort. Lastly, this was a single-site study within a mandated fertility coverage state. Participation in the study and results regarding anxiety or depression may not be generalizable to other non-mandated states or fertility clinics.

Another limitation to our study is the lack of universal hormonal testing for men who reported low libido or erectile dysfunction in our questionnaire set. We routinely assay morning serum hormonal levels (FSH, LH, testosterone profile) based on the AUA/ASRM 2021 guidelines (Schlegel et al., 2021). We controlled for male infertility diagnosis in an attempt control for this limitation, but we understand this may not completely address the limitation. We performed a sub-analysis of those with available hormonal data and found data on 22/222 (9.91%) patients. Out of those, 10/22 (45.45%) met the criteria for being hypogonadism and were subsequently removed from analysis in Table 5 due to potential confounding. We recommend that future studies should prospectively obtain hormonal evaluations of men who report low libido on erectile dysfunction on survey responses to provide a better understanding of the causation of effects.

Table 5.

Association of Hospital Anxiety and Depression Scale (HADS) score on erectile dysfunction and low libido, n (%).

	HADS <8	HADS ≥8	Unadjusted Relative risk (95% CI)	Adjusted Relative risk* (95% CI)
	N = 161	N = 51
Erectile dysfunction
Yes	25 (15.72)	12 (24.49)	1.56 (0.85, 2.86)	1.12 (0.60, 2.06)
No	134 (84.28)	37 (75.51)
Missing	2	2
Low libido
Yes	24 (14.91)	16 (31.37)	2.10 (1.22, 3.64)	1.70 (0.91, 3.15)
No	137 (85.09)	35 (68.63)

*

Adjusted for antidepressant use, male BMI, and smoking.

Excludes patients with hypogonadism.

Conclusion

Men with anxiety had lower fTMSCs during IVF compared to men without anxiety. However, regardless of antidepressant use, couples with men who reported anxiety and/or depression had similar LBRs when undergoing IVF. We conclude that the findings of this study underscore the importance of assessing both partners for mental health disorders during fertility treatment and supporting mental health treatments, including antidepressant medications, regardless of their fertility goals.

Supplementary data

Supplementary data are available at Human Reproduction online.

Data availability

The data that support the findings of this study are not openly available due to reasons of sensitivity. Data are available from the corresponding author upon reasonable request and with permission from the Mass General Brigham Institutional Review Board and Brigham and Women’s Hospital.

Authors’ roles

Z.W., A.L., S.S.S., and M.K. contributed to the study conception and design. Material preparation, data collection, and analysis were performed by Z.W. and A.L. The first draft of the manuscript was written by Z.W. and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding

There was no funding to report for this study.

Conflict of interest

Z.W. is a contributing author for UptoDate. S.S.S. is on the advisory board for Ferring Pharmaceuticals. E.G. was a medical consultant for Hall-Matson Esq, Teladoc, and CRICO and is a contributing author for UptoDate. The remaining authors have nothing to report.