Semen Quality Parameters Among U.S. Veterans of Operation Enduring Freedom, Operation Iraqi Freedom, and Operation New Dawn

ABSTRACT

Background: There is little research investigating exposures that occurred during war or conflict and the resulting influence on semen quality. We examined the associations between military service and semen quality among U.S. Veterans of Operation Enduring Freedom, Operation Iraqi Freedom, and Operational New Dawn. Methods: U.S. Veterans (n = 714), who sought fertility treatment and were participants of the Department of Veterans Affairs Women Veterans Cohort Study were included in the current study. Semen quality parameters (semen volume, total sperm count, sperm concentration, and percentage of total motile sperm) measured at fertility clinics in Veterans Affairs Healthcare System were analyzed. Findings: The median values of Veterans were 2.7 mL for semen volume, 55.3 × 10⁶ for total sperm count, 22.0 × 10⁶/mL for sperm concentration, and 55% for total motile sperm. No significant differences among Veterans were found between military-specific variables evaluated and sperm quality. In multivariate models, Veterans ≥46 years of age had lower semen volume (β = −0.10 ± 0.05; p < 0.05) and lower percent total motile sperm (β = −18.45 ± 7.0; p = 0.009) than Veterans 18 to 25 years of age. Black and Hispanic Veterans had lower sperm concentrations as compared to White Veterans (β = −0.17 ± 0.07; p = 0.01 and β = −0.20 ± 0.06; p < 0.001, respectively). Finally, Veterans diagnosed with post-traumatic stress disorder or major depression had lower total sperm motility (β = −0.10 ± 0.03; p = 0.004 and β = −0.09 ± 0.04; p < 0.05, respectively). Discussion: Veterans who were older, Hispanic or Black, or diagnosed with post-traumatic stress disorder or major depression were more likely to have lower sperm quality. Longitudinal studies are needed to better characterize the influence of military-specific exposures on semen quality parameters.

INTRODUCTION

Since October 2001, more than 2.7 million men and women of the U.S. Armed Forces have been deployed for Operation Enduring Freedom (OEF), Operation Iraqi Freedom (OIF), or Operation New Dawn (OND) in support of the Global War on Terrorism (GWOT).¹ Most research conducted to assess the prevalence of reproductive health-related treatment and diagnoses among OEF/OIF/OND Veterans has focused primarily on female Veterans, and limited data are available pertaining to males in this unique subset of the U.S. population.

A recent retrospective analysis of self-reported data from male OEF/OIF/OND Veterans who received medical care through the Veterans Affairs (VA) Healthcare System indicated that this cohort is seeking and receiving reproductive health treatment or evaluation at what appears to be a higher rate than military members who have not deployed in support of GWOT.² Furthermore, this rate appears to exceed that in the general U.S. civilian male population,^3,4 suggesting a possible link between military service-related exposure and an impact on male semen quality.

There is little research investigating exposures that occurred during war or conflict and the resulting influence on semen quality. U.S. military-specific studies have reported that Veterans who served in Vietnam had lower sperm quality than non-Vietnam Veterans.⁵ Veterans who were directly exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin, a chemical contaminant commonly associated with “Agent Orange” use during the Southeast Asia and Vietnam conflicts, displayed no differences in sperm quality as compared to Vietnam Veterans who were not directly exposed.⁶ Moreover, military occupational-based studies are limited to artillery and radar operators; however, their associations with sperm quality are conflicting.^7,8

Other studies, focused on non-U.S. military subjects and conducted outside the United States, report that exposure to increased stress or the chemical weapon bis (2-chloroethyl) sulfide (also known as sulfur mustard, or “mustard gas”) was associated with decreased sperm quality among military-aged males following the Lebanese Civil War and the Iraq–Iran war during the mid–to-late 1980s.^9,–11 Additionally, 1 study in France suggests a possible relationship between service on nuclear submarines and male infertility.¹² A recent military population-based study in China that evaluated the impact of physical environment on sperm quality found an association between altitude, climate, diet, and lifestyle with decreased sperm parameter values, namely volume, total count, and motility.¹³

To our knowledge, the potential impact of military occupation in support of OEF/OIF/OND on semen quality parameters has yet to be examined. Using clinical data from male Veterans in the OEF/OIF/OND Cohort, we evaluated semen quality parameters from men seeking fertility treatment to determine the relationship between these parameters and demographics, military-related characteristics, and select mental health diagnoses.

METHODS

Overview of Study Design and Data Sources

The Women Veterans Cohort Study (WVCS) includes OEF/OIF/OND Veterans enrolled for VA care with the aim to identify gender-associated disparities in health care utilization and outcomes. Data have been described previously elsewhere.^14,15 Briefly, data from the Department of Defense Manpower Data Center's (DMDC) Contingency Tracking System that provide information on sex, race, date of birth, and component and branch of service at the time of leaving active duty were linked to the VA's National Corporate Data Warehouse that contains much of the available clinical data from the electronic health record. A dataset that included information on military service and subsequent patient encounters with VA health care providers, as well as codified diagnostic and procedure data, laboratory, and pharmacy data were created. This dataset served as the basis for our study. This study was approved by the VA Connecticut Institutional Review Board.

Subjects

530,669 male OEF/OIF/OND Veterans had utilized the VA health services network between October 1, 2001, and December 30, 2010, for primary, emergency, or mental health care. Of those, 3,485 had received an infertility assessment, procedure, or select mental health diagnosis, quantified according to Current Procedural Terminology (CPT) codes. Within that group, 1,337 had semen analysis laboratory data (indicated by CPT codes 89300, 89310, and 89320) associated with their patient identification number. Semen analyses typically conducted for confirmation of success following a vasectomy (CPT code 89321) were excluded, because of the likelihood those tests did not quantify sperm parameters but were instead conducted to detect presence or motility of sperm.

We included individuals who had data for three sperm parameters of interest: semen volume, total sperm number, and sperm concentration. Of the 1,337 participants with semen parameter laboratory values in their records, a total of 44 (3.3%) subjects had all three parameters reported and 670 (50.1%) contained at least two of the three parameters. For those subjects with values for two of the three parameters, concentration was calculated by dividing count by volume, count by multiplying concentration with volume, and volume by dividing count by concentration, as appropriate, bringing the total number of subjects included to 714. In all cases, values recorded were manually converted to 10⁶ as appropriate for count and concentration. The remaining subjects (n = 623) were excluded from the cohort because of blank or missing values in the parameters of interest, zeros, or only a single parameter reported to restrict the analysis to those only with measurable parameters. Parameter values equal to zero, although a possible result from an analysis, were excluded to include only those subjects with measureable numbers of sperm. Additionally, 377 (52.8%) subjects from the final sample also had information on total motility (progressive [PR] + nonprogressive [NP]) included within their records.

To identify whether military occupation potentially impacts semen quality parameters, subjects from the final sample were classified into an occupational group (combat, maintenance, or support/all others) on the basis of the service-specific occupational code recorded in the DMDC roster on their discharge from military service. Although many service members perform duties outside the scope of their specific occupational identifier, only the primary duty was used for this grouping. For ease of comparison between the branches of service, and given the largest number of subjects served in the Army, the Army's “combat arms” branches served as a reference for the other services. Each subject's occupational code in the DMDC roster was individually cross-referenced between branches of service for applicability and grouped as appropriate. Supplemental Table II contains the breakdown of occupations and groupings for the final cohort.

Mental health diagnoses, including post-traumatic stress disorder (PTSD), depression, and bipolar disorder were evaluated to investigate associations with sperm quality parameters. Mental health diagnoses were quantified using International Classification of Diseases, 9th Revision, codes. These variables were coded as “ever/never” diagnosed; diagnoses may have taken place before or after the time the semen analysis samples were obtained. Additionally, we included service-connected disability rating scores, which are determined by the VA on discharge from military service and accounts for any disease or injuries that occurred or were aggravated as a direct result of military service, in our analysis to evaluate the association between disability rating and semen parameters.

Statistical Analysis

We started by describing the characteristics of our sample in relation to the semen quality parameters of interest. Because of the non-normal distribution of our semen parameters, we used Kruskal–Wallis tests to examine differences within demographic, service-related, and mental health categories by each semen parameter. Variables that were significantly different at the p < 0.05 level were included in a multivariate linear regression model, where each semen parameter was included as a dependent variable in a separate model. To account for the non-normality of the semen parameters, these measures were log transformed (base 10) before running the models. Statistical analyses were performed with SAS, version 9.3 (SAS Inc., Cary, North Carolina). All statistical tests were 2-tailed with an α level of 0.05.

RESULTS

The demographic and military-related data of our nested cohort are shown in Table I. Mean age at the time of semen analyses was 31.3 ± 6.9 years and 44.3% were married. Approximately 51% of Veterans in our cohort were White, whereas 23% and 16% were Hispanic and Black, respectively. As can be expected given the size and missions of each branch of service, 61.5% of Veterans served on active duty before discharge from service and 60.2% served in the U.S. Army. Additionally, 96% of the cohort served as enlisted members of the military, with only 3.9% serving in the officer ranks. Service-connected disability ratings were 36.4%, 27.9%, and 35.7% for the disability rating groups of 0%, 10 to 40%, and ≥50%, respectively. Furthermore, 44.3% of Veterans were diagnosed with PTSD and 31.8% were diagnosed with mild depression, either during service or after they were discharged.

TABLE I

Demographic Characteristics of the Study Population

Characteristic	N (% of 714)
Age Group (Mean Age at Time of Sample = 31.3; SD = 6.9)
18–25	175 (24.5)
26–35	385 (53.9)
36–45	117 (16.4)
≥46	37 (5.2)
Race
White	364 (51.0)
Black	114 (16.0)
Hispanic	164 (23.0)
Other	30 (4.2)
Unknown	42 (5.9)
Marital Statusa
Married	316 (44.3)
Not Married	23 (3.2)
Never Married	375 (52.5)
Educationb
<High School Diploma	8 (1.1)
High School Diploma or Equivalent	582 (82.0)
>High School	120 (16.9)
Rank
Enlisted	686 (96.1)
Officer	28 (3.9)
Component
Guard	154 (21.6)
Active	439 (61.5)
Reserve	121 (17.0)
Branch of Service
Army	430 (60.2)
Navy	108 (15.1)
Air Force	47 (6.6)
Marine Corps	129 (18.1)
Military Occupation
Combat	204 (28.5)
Maintenance	79 (11.1)
Support/Other	431 (60.4)
Service-Connected Disability
0%	260 (36.4)
10–40%	199 (27.9)
≥50%	255 (35.7)
PTSD
Ever Diagnosed	316 (44.3)
Mild Depression
Ever Diagnosed	227 (31.8)
Major Depression
Ever Diagnosed	123 (17.2)
Bipolar Disorder
Ever Diagnosed	53 (7.4)

^a

Marital status obtained at time of enrollment in the WVCS and does not necessarily reflect status at time of sample(s).

^b

Four subjects missing education data; % shown is of 710.

Table II shows the distribution of semen parameter values obtained from samples provided by Veterans in our cohort. As expected Veterans seeking infertility treatment had lower median sperm parameter data than reference values established by the World Health Organization (WHO) on the basis of time-to-pregnancy ≤12 months¹⁶: semen volume (2.7 mL vs. 3.7 mL), total sperm count per ejaculate (55.3 × 10⁶ vs. 255 × 10⁶), sperm concentration (22 × 10⁶/mL vs. 73 × 10⁶/mL), and total motility (55% vs. 61%). However, median sperm parameter data for Veterans and non-Veteran U.S. populations seeking infertility treatment¹⁷ are similar: semen volume (2.7 mL vs. 2.7 mL), total sperm count per ejaculate (55.3 × 10⁶ vs. 52 × 10⁶), sperm concentration (22 × 10⁶/mL vs. 20.6 × 10⁶/mL), and total motility (55% vs. 51%).

TABLE II

Distribution of Semen Quality Parameters

Parameter (Units)	N	Centiles
		2.5	5	10	25	50	75	90	95	97.5
Semen Volume (mL)	714	0.5	0.9	1.0	2.0	2.7	3.8	5.0	5.9	6.9
Total Sperm Number (106 Per Ejaculate)	714	1.7	2.9	6.0	20.7	55.3	117.0	247.0	387.0	531.2
Sperm Concentration (106 Per mL)	714	0.7	1.1	2.5	8.1	22.0	52.0	9.7	155.0	203.0
Total Motility (PR + NP, %)	377	0	7.0	15.0	33.0	55.0	72.0	80.7	86.0	90.0

Table III shows semen quality parameters stratified by demographic variables. The median semen volume of Veterans in this cohort differed by age groups (Table III; p = 0.02). Linear regression models adjusting for race showed that Veterans ≥46 years of age had lower semen volume than Veterans 18 to 25 years of age (β = −0.10 ± 0.05; p < 0.05). Similarly, median total sperm count per ejaculate differed by age groups (p < 0.05); however, statistical significance was not achieved in regression models adjusting for race. Although no group difference was observed between age and total motile sperm (p = 0.07), adjusting for race in linear regression models revealed that Veterans ≥46 years of age had a lower percentage of total motile sperm compared to those aged 18 to 25 years (β = −18.45 ± 7.0; p = 0.009).

TABLE III

Median Semen Quality Parameters by Demographic Variables

	N	Semen Volume (mL) Median (25–75)	Total Sperm Number (106 per Ejaculate) Median (25–75)	Sperm Concentration (106 per mL) Median (25–75)	Total Motility (PR + NP, %)
					Na	Median (25–75)
Age (Years)
18–25 (Ref)	175	3.00 (1.8–4.0)	58.60 (25.0–107.6)	22.00 (10.0–45.4)	68	56.50 (39.0–79.5)
26–35	385	2.89 (2.0–3.9)	62.30 (12.0–147.4)	23.33 (8.1–59.0)	191	55.00 (33.0–70.0)
36–45	117	2.50 (1.9–3.5)	36.50 (16.6–106.0)	22.40 (6.1–46.0)	62	55.00 (36.0–75.0)
≥46	37	2.00 (1.5–3.5)b	36.00 (18.9–90.0)	15.50 (6.7–40.4)	16	30.00 (11.0–62.5)b
p Value		0.02	0.047	0.373		0.07
Race
White (Ref)	364	2.84 (2.0–4.0)	69.05 (25.0–148.0)	25.35 (9.2–57.4)	181	60.00 (38.0–75.0)
Black	114	2.50 (1.5–3.9)	45.90 (13.0–110.0)c	16.57 (6.4–46.5)c	51	50.00 (24.0–70.0)
Hispanic	164	2.80 (1.6–3.5)	38.45 (16.25–89.0)c	16.28 (7.3–45.6)c	69	55.00 (35.0–72.0)
Other	30	3.00 (2.0–3.5)	55.00 (24.7–189.0)	22.25 (7.1–65.0)	15	55.00 (22.0–61.0)
Unknown/Missing	42	2.30 (2.0–3.4)	53.85 (19.33–118.0)	22.38 (6.4–56.7)	21	40.00 (28.0–59.0)
p Value		0.32	0.0009	0.02		0.12
Marital Statusd
Not Married (Ref)	23	2.20 (1.5–3.6)	65.00 (31.5–120.0)	22.42 (7.0–51.9)	12	66.50 (40.0–76.5)
Married	316	2.50 (1.75–3.6)	52.85 (17.95–127.6)	24.74 (10.5–74.0)	147	50.00 (25.0–70.0)e
Never Married	375	3.00 (2.0–4.0)	59.40 (21.0–111.4)	21.21 (8.1–51.0)	178	59.00 (39.0–75.0)
p Value		0.14	0.66	0.60		0.008
Education
<High School Diploma	8	2.19 (0.93–4.75)	19.48 (8.5–161.7)	9.08 (8.7–32.8)	3	38.00 (23.0–70.0)
High School Diploma or Equivalent (Ref)	582	2.79 (1.8–3.9)	58.50 (20.0–117.0)	22.29 (8.1–52.0)	277	57.89 (31.6–75.0)
>High School	120	2.55 (2.0–3.6)	49.70 (24.7–117.6)	22.00 (7.8–56.8)	55	50.00 (35.0–64.0)
Unknown/Missing	4	3.95 (2.95–4.85)	25.30 (21.55–218.4)	9.68 (5.5–42.7)	2	24.00 (8.0–40.0)
p Value		0.83	0.74	0.36		0.39

p values derived from Kruskal–Wallis nonparametric analysis of variance.

^a

Motility N = 377.

^b

p < 0.05 from PROC GLM (SAS Inc., Cary, North Carolina) for mean values within each group compared to reference (ref) group controlling for race.

^c

p < 0.05 from PROC GLM for mean values within each group compared to reference (ref) group controlling for age.

^d

Marital status obtained at time of WVCS enrollment and does not necessarily reflect status at time of sample(s).

^e

p < 0.05 from PROC GLM for mean values within each group compared to reference (ref) group controlling for age and race.

We also observed differences in median total sperm count and concentration by race (p = 0.0009 and p = 0.02, respectively). Subsequent linear regression models adjusting for age revealed that the significant differences in total sperm count by race were driven by lower total sperm counts among Black and Hispanic Veterans than Whites (β = −0.21 ± 0.07; p = 0.002 and β = −0.23 ± 0.06; p < 0.0001, respectively). Similarly, Black and Hispanic Veterans had lower sperm concentrations than White Veterans after adjusting for age (β = −0.17 ± 0.07; p = 0.01 and β = −0.20 ± 0.06; p < 0.001, respectively). Total motile sperm also differed by marital status (p = 0.008), which was largely driven by lower total motile sperm among married Veterans than those not married after adjusting for age and race (β = −0.19 ± 0.09; p = 0.04). No significant results were found between sperm parameters and education (Table III).

Table IV shows semen quality parameters stratified by military variables and occupational specialties held while serving. Median total sperm count for Veterans who were enlisted were lower than those who were officers; however, these differences did not reach statistical significance (p = 0.09) even after adjusting for age and race in linear regression analysis (β = −0.22 ± 0.12; p = 0.06). Likewise, median percentages of total sperm motility appeared to differ by military component and occupation but the differences only approached statistical significance (p = 0.07 and p = 0.10, respectively).

TABLE IV

Median Semen Quality Parameters by Military/Occupational-Specific Variables, Disability Rating, and Select Mental Health Conditions

	N	Semen Volume (mL) Median (25–75)	Total Sperm Number (106 per Ejaculate) Median (25–75)	Sperm Concentration (106 per mL) Median (25–75)	Total Motility (PR + NP, %)
					Na	Median (25–75)
Rank
Officer (Ref)	28	2.85 (2.0–3.9)	85.10 (35.5–158.3)	23.67 (9.0–76.3)	9	50.00 (35.0–60.0)
Enlisted	686	2.70 (1.9–3.8)	54.00 (20.0–116.0)	22.00 (8.1–51.1)	328	55.00 (32.0–73.5)
p Value		0.37	0.09	0.23		0.44
Component
Guard (Ref)	154	2.50 (1.65–3.5)	57.50 (20.0–135.0)	23.43 (8.0–62.0)	72	54.72 (36.5–70.0)
Active	439	2.80 (2.0–3.9)	54.00 (21.0–115.8)	21.14 (7.76–50.0)	211	51.00 (30.0–72.0)
Reserve	121	2.70 (1.8–4.0)	59.84 (20.7–108.0)	23.50 (8.4–50.7)	54	63.00 (40.0–80.0)
p Value		0.61	0.90	0.85		0.07
Branch of Service
Army (Ref)	430	2.79 (2.0–3.8)	55.50 (20.7–134.0)	22.00 (8.1–55.0)	205	59.00 (35.0–75.0)
Navy	108	3.00 (1.55–3.9)	53.95 (19.7–106.2)	20.70 (7.3–50.5)	54	49.50 (30.0–65.0)
Air Force	47	2.20 (1.6–3.0)	61.00 (24.5–115.8)	24.00 (24.0–10.8)	22	57.50 (29.0–64.0)
Marine Corps	129	2.80 (2.0–4.0)	55.20 (21.0–107.7)	24.00 (10.8–74.0)	56	52.00 (38.65–72.0)
p Value		0.28	0.92	0.70		0.40
Military Occupation
Support/Other (Ref)	431	2.60 (1.8–3.8)	55.40 (20.7–118.8)	23.33 (7.5–56.7)	198	50.50 (30.0–70.0)
Combat	204	2.85 (2.0–3.8)	52.40 (18.5–110.7)	20.76 (8.1–43.0)	100	58.50 (35.1–75.0)
Maintenance	79	2.90 (2.0–4.0)	64.00 (26.0–151.5)	22.00 (9.4–74.5)	39	64.00 (40.0–80.0)
p Value		0.39	0.34	0.31		0.10
Service-Connected Disability
0% (Ref)	260	2.60 (1.8–3.6)	60.00 (19.0–125.8)	23.15 (7.9–58.3)	119	61.00 (38.0–75.0)
10–40%	199	3.00 (1.6–4.0)	54.90 (20.0–124.0)	22.00 (7.7–58.0)	95	50.00 (32.0–70.0)
≥50%	255	2.70 (2.0–3.8)	53.00 (24.4–110.0)	21.90 (8.2–49.0)	123	53.00 (29.0–70.0)
p Value		0.68	0.94	0.83		0.23
PTSD
Never Diagnosed (Ref)	398	2.70 (1.8–3.9)	55.60 (20.0–115.8)	21.95 (7.0–51.0)	193	60.00 (37.0–75.0)
Ever Diagnosed	316	2.70 (2.0–3.8)	54.70 (23.0–119.0)	22.42 (9.0–54.3)	144	50.00 (30.0–70.0)**
p Value		0.78	0.42	0.39		0.04
Mild Depression
Never Diagnosed (Ref)	487	2.80 (1.8–4.0)	54.90 (20.7–118.8)	22.00 (7.7–50.0)	232	55.50 (34.0–73.5)
Ever Diagnosed	227	2.70 (2.0–3.5)	56.00 (20.7–117.0)	23.52 (8.5–57.6)	105	50.00 (30.0–70.0)
p Value		0.48	0.69	0.36		0.37
Major Depression
Never Diagnosed (Ref)	591	2.70 (2.0–3.9)	58.00 (20.5–118.0)	22.63 (8.1–51.9)	278	56.00 (35.0–75.0)
Ever Diagnosed	123	3.00 (2.0–3.7)	51.00 (22.0–117.0)	28.24 (7.5–54.0)	59	42.00 (24.0–70.0)*
p Value		0.97	0.56	0.76		0.09
Bipolar Disorder
Never Diagnosed (Ref)	661	2.75 (2.0–3.9)	57.00 (21.0–116.0)	22.00 (8.1–52.1)	314	55.00 (33.0–72.0)
Ever Diagnosed	53	2.50 (1.8–3.5)	39.00 (20.0–117.0)	20.00 (8.3–50.0)	23	50.00 (30.0–75.0)
p Value		0.22	0.74	0.91		0.95

p values derived from Kruskal–Wallis nonparametric analysis of variance.

*

p < 0.05 from PROC GLM for mean values within each group compared to reference (ref) group controlling for age and race.

**

p < 0.01 from PROC GLM for mean values within each group compared to reference (ref) group controlling for age and race.

^a

Motility N = 377.

Finally, we examined the influence of disability and mental health conditions on semen quality parameters (Table IV). We observed no significant results between sperm quality and overall percentages of service-connected disability. However, when examining specific mental health conditions, we observed that Veterans diagnosed with PTSD had lower total sperm motility (p = 0.04), which was more pronounced when adjusting for age and race in linear regression models (β= −0.10 ± 0.03; p = 0.004). Also, Veterans diagnosed with major depression had lower total motile sperm after adjusting for age and race (β = −0.09 ± 0.04; p < 0.05), whereas diagnosis of mild depression was borderline significant (β = −0.06 ± 0.04; p = 0.07). Bipolar disorder diagnosis among Veterans was not associated with significant changes in sperm quality parameters.

DISCUSSION

The objective of this study was to investigate the associations between military service and semen quality parameters among OEF/OIF/OND Veterans. We hypothesized that those Veterans who served in combat occupations would have lower sperm quality than their counterparts in maintenance or support jobs as a result of enduring high-stress scenarios on a more frequent basis or being subjected to austere conditions more regularly; however, no significant differences in sperm quality were observed.

Although this study did not detect any significant associations between decreased sperm quality and military occupation, we observed that a diagnosis of PTSD or major depression among Veterans was associated with a decrease in sperm motility after adjusting for age and race. It must be noted, however, that it is unknown whether these mental health diagnoses were made before or after semen samples were provided and whether Veterans were currently taking antidepressants at the time of semen collection. Furthermore, given our current study is a convenience sample with the OEF/OIF/OND Cohort, we are not positioned to investigate potential mechanisms for this association. However, a recent case study reported that use of the antidepressant citalopram was associated with abnormal sperm parameters including motility, which subsequently improved after discontinuing medication after 4 months.¹⁸ These results are further supported by a study in rats in which a high dose of bupropion (30 mg/kg for 30 days) resulted in higher levels of NP sperm.¹⁹ Thus, our results may be explained by pharmaceuticals prescribed to treat or manage PTSD and depression. Alternatively, our results between PTSD and sperm motility may be related to stress itself or its downstream effectors; however, a recent self-report study of psychological stress among Danish men found that stress was related with lower sperm concentration and count as well as semen volume but not with sperm motility.²⁰

The effect of age on sperm quality has remained a subject of much investigation in recent decades. The age-dependent decrease of semen volume, sperm motility, and to an extent total sperm number per ejaculate after 35 years of age observed in our study supports those observed by another infertility clinic cohort study. Stone et al²¹ observed in their study that all semen parameters remained the same up to 34 years of age, whereas, total sperm count and motility decreased thereafter, sperm concentration decreased at the age of 40, and semen volume at the age of 45. Supporting these findings, we observed a similar pattern in our study, though with less fidelity, possibly because of differences in age groupings.

Like that of age, the potential role of race on semen quality is a topic of great importance. The semen analysis, which remains the foundation in evaluating male fertility status²² despite its relatively “crude” methodology,²³ relies on reference values established by the WHO in diagnosing male factor infertility.²⁴ Although these reference values are averages on the basis of studies that encompassed thousands of semen samples from known fertile men across the globe,¹⁶ no measures are in place to control for variations in semen quality between different races or ethnicities; the end result is an average across all men. However, one study in the United States that sought to investigate the effect of race on semen quality found that Black males had lower sperm quality parameters than White males, whereas Hispanic males were similar to white males.²⁵ Our study of U.S. Veterans found similar results as Redmon et al²⁵ with respect to lower sperm quality among Blacks than Whites; however, Hispanic Veterans had lower total sperm count and concentration compared to White Veterans. Additionally, the findings of our study parallel those of an earlier VA study (1998–2003) that found Hispanic males received infertility treatment more frequently than Black males, followed by White males.²⁶ Although further studies will need to confirm our findings, the potential effect of race on semen quality measures should be considered when setting future reference guidelines.

Finally, it is evident that there are a growing number of U.S. military Veterans that are using VA health care infertility services. Within the literature, there exist an extremely limited number of prospective or longitudinal studies that seek to determine the prevalence of male factor infertility. Those that do are themselves based out of fertility clinics and are further limited by sample size or come from a distinct geographic region, thereby, limiting their applicability to the general population. Although our study is subject to similar selection biases such that Veterans who provided samples were users of VA health care and had deployed in support of OEF/OIF/OND, no generalizations can be made to the larger population. However, median semen parameters (Table II) of our cohort closely matched those of a larger 5-year study of 1,680 samples conducted at a single fertility clinic.¹⁷ Therefore, although sperm parameter data among U.S. Veterans are below normal reference guidelines for the general population put forth by the WHO,¹⁶ they are in agreement with other populations who are also seeking infertility treatment. Given these similarities in sperm parameters among subfertile men, another consideration that should be taken into account is the rate at which Veterans seek infertility treatment. Recent data from OEF/OIF has estimated the prevalence of lifetime history of infertility was 13.8% for male Veterans.² This exceeds prevalence estimates of the general population of 9.4%³ and 12%.⁴ Thus, among individuals seeking infertility, sperm quality parameters do not appear to be appreciably different between Veterans and the general population; however, the rate at which Veterans seek infertility treatment may be high.

Another limitation to this study is the multiple clinical laboratories that captured the data for the 714 veterans in this study. Although ostensibly trained to follow WHO guidelines, site-specific variations in semen quality analyses are possible; however, as mentioned earlier our sperm quality data are in line with other studies of subfertile males.¹⁷ We also have to acknowledge that more than 600 Veterans in our study had incomplete semen analysis data, and each clinic had their unique method for reporting the data, which limited our ability to capture all data across clinics. Moving forward, attention should be given to standardize reporting of semen quality data, and presumably other clinical data, across VA sites to facilitate its research capacity. Finally, the OEF/OIF/OND roster does not represent all OEF/OIF/OND Veterans; rather, it includes only those who have separated from service and enrolled or received VA health care services. It does not include those Veterans who remain on active duty, in the reserves, or as members of the National Guard; thus, the generalizability of our findings to the military population is also limited.

In conclusion, although we found no significant associations between service-specific variables and semen quality parameters, age, race, and mental health diagnoses of PTSD and depression were associated with select semen quality parameters. To better characterize potential effects of military-specific exposures on semen quality parameters, a carefully developed longitudinal study is clearly warranted. Continued deployments by U.S. military personnel in support of the GWOT present unique opportunities to accomplish such research objectives.

ACKNOWLEDGMENTS

We would like to give a special thank you to all Veterans for their service.

Footnotes

1

This article is based on work supported by the U.S. Department of Veteran Affairs, Office of Research and Development, Health Services Research and Development (DHI 07-065-2; C.A.B.).