Clinical Manifestations of Trichomoniasis in Men

Methods

Subjects

Subjects were selected from men attending the Seattle-King County Sexually Transmitted Disease (STD) Clinic at Harborview Medical Center from 1 May 1987 to 31 December 1990. Men were considered eligible for the study if they were heterosexual, 16 to 65 years old, had not used antimicrobial agents within the previous 14 days, spoke English, and were able to give informed consent. The protocols were approved by the University of Washington Human Subjects Committee. Subjects were classified into two categories, men who were sexual contacts of women with trichomoniasis and randomly selected men.

Contact men were defined as male sexual partners (within the previous 30 days) of women with trichomoniasis. The contact category included a total of 147 men. These included 52 men referred by women who were approached because we had identified T. vaginalis on their vaginal wet mount slides. The remaining 95 men were referred by women with trichomoniasis diagnosed at other facilities or who were said by the index men to have been treated for trichomoniasis at our clinic but who were not specifically identified by the index men.

Heterosexual men presenting with new problems were selected by matching the final two digits of their hospital numbers with a table of random numbers. Men were not eligible if they were returning for follow-up examinations or had had evaluations during the preceding 30 days. The 300 randomly selected men who enrolled in the study represented approximately 75% of heterosexual men presenting to the clinic for evaluation of new problems whose chart numbers matched the table of random numbers. Reasons for refusal included added time to complete the protocol (15%), unwillingness to enroll in a study (5%), and fear of discomfort from the urethral swabs (5%). Although we did not maintain demographic and clinical profile data on heterosexual men who refused enrollment in the database for this study, no obvious differences were noted from those who consented to enrollment.

Clinical Evaluation

An extensive sexual history and standardized physical examination were carried out for each participant. Current genitourinary tract complaints were recorded, including urethral discharge, urethral burning, urethral or perineal pain, urinary frequency, and dysuria. Demographic characteristics, medical, contraceptive, and sexual histories were recorded. Any history of therapy for previous urinary tract infections or sexually transmitted diseases was recorded. The physical examination emphasized evaluation of the genitalia for presence of inguinal adenopathy, genital lesions, urethral discharge, and intrascrotal abnormalities. The data were entered onto standardized forms.

Presence of urethral discharge was evaluated by inspection of the urethral meatus before and after stripping the urethra. The amount of discharge was classified as large if the discharge was profuse and present spontaneously at the meatus; moderate if discharge was not profuse but was present spontaneously or obvious after stripping the urethra; small if discharge was present in minimal amounts only after stripping the urethra; or none. The character of discharge was classified as purulent, if grossly yellow; intermediate, if cloudy but not grossly purulent; clear, if water-like in color and of slightly mucoid consistency; or none. Urethritis was defined as presence of at least two of the following: symptoms of urethral discharge or of urethral discharge on physical examination or presence of -w 5 polymorphonuclear leukocytes per oil immersion microscopic field (magnification, 1000 x) of the gram-stained urethral smear.

Microbiologic Evaluation

Urethral specimens were obtained using calcium alginate tipped swabs for gram stain and cultures. Trichomonas vaginalis was cultured using modified Diamond's medium [10] incubated at 37 C in an atmosphere of 5% CO₂ and room air. Neisseria gonorrhoeae was cultured using modified Thayer-Martin medium [11]. Oxidase-positive gram-negative diplococci were evaluated with biochemical tests, which included measurement of their ability to produce acid from glucose, maltose, lactose, sucrose, and fructose [12]. Chlamydia trachomatis was cultured using cycloheximide-treated McCoy cells in microtiter plates [13]. Fluorescein-conjugated monoclonal antibodies were used to identify chlamydial inclusions in infected monolayers of McCoy cells [14]. The genital mycoplasmas were cultured using broth and agar media containing urea and sulfite [15]. Additional specimens were obtained to determine if T. vaginalis was present; the specimens were from the external genitalia (subpreputial space of uncircumcised men or coronal sulcus of circumcised men) and the centrifuged sediment of 10 mL of first-voided urine (obtained after the urethral specimens). Trichomonas vaginalis was identified based on the distinctive motility and morphology exhibited by the protozoa in wet mounts prepared from the culture tubes.

Statistical Analysis

The chi-square test or Fisher exact test was used to compare categorical data, and the Mann-Whitney U test [16] was used to compare ordinal data. Mantel-Haenszel procedures [17] were used to test associations between infection with T. vaginalis and selected clinical findings after stratifying for exposure to a sexual partner with trichomoniasis. Logistic regression analyses [18] were used to adjust the association between T. vaginalis infection and urethritis for variables found by univariate analysis to be associated with either trichomoniasis or nonchlamydial nongonococcal urethritis. Variables satisfying these criteria were included simultaneously in the model.

Results

Descriptive Characteristics

The 447 men in this study ranged from 14 to 64 years old. Among the contact males, no statistical differences existed in mean age, income, or racial distribution between men named by women with confirmed trichomoniasis and those referred by partners diagnosed at other facilities or those who did not identify the partner. Therefore, all contact men were combined for subsequent analyses. No statistical difference (P > 0.2) existed in annual income between the 75% random sample and the contact males. The contact males were older (mean 33 9 years [standard deviation] versus 29 8 years, P < 0.001). The racial distribution was 51% black, 43% white, and 6% other for the 75% random sample; versus 75% black, 22% white, and 3% other for the contact men (P < 0.001).

Prevalence of Urogenital Trichomoniasis

Trichomonas vaginalis was isolated from 50 (11%) of the 447 subjects in this study, including 17 (6%) of 300 among the 75% random sample of men undergoing evaluation for new problems and 33 (22%) among the 147 contacts of women with trichomoniasis (P < 0.01, Table 1). Only one man who had a positive culture from the external genitalia also had negative urine and urethral cultures for T. vaginalis. Among contacts of women with trichomoniasis, the prevalence of T. vaginalis was 9 [17%] of 52 for men whose partners had confirmed trichomoniasis and 24 (25%) of 95 for the remaining contacts (P > 0.2).

Symptoms, Physical Findings, and Urethral Smear Findings

The symptom of urethral discharge was the most common chief complaint (see Table 1). Urethral discharge was reported by 125 [42%] of 300 men in the 75% random sample and by 49 (33%) of 147 contacts of women with trichomoniasis. In each group, isolation of T. vaginalis was associated with symptoms of urethral discharge (P < 0.05 for men in the 75% random sample, P < 0.04 for contact men) but not with signs of urethral discharge. Of the 50 men with trichomoniasis in our study, 15 had received treatment for trichomoniasis in the past. Of the 15 cases who had received previous treatment, 10 (67%) had current urethral symptoms, although 17 (49%) of the 35 cases with no history of previous treatment had urethral symptoms (Fisher exact test, P > 0.2). Thus, no clear difference existed in the prevalence of current urethral symptoms between men with and without a history of previous treatment for trichomoniasis, but the numbers in each group were small. No other symptom or physical finding was associated with an increased rate of T. vaginalis isolation. However, the median number of polymorphonuclear leukocytes per 1000 x oil immersion microscopic field of the gram-stained urethral smear was higher among men with trichomoniasis than among those without trichomoniasis. Among contact men, the median number of polymorphonuclear leukocytes per oil immersion field (1000 x) was 13 for T. vaginalis-positive men compared with a median of 1 for T. vaginalis-negative men (P < 0.03). The contact men from whom T. vaginalis was isolated were almost as likely to be completely asymptomatic (42%) as they were to acknowledge symptoms of urethral discharge (48%).

Of the 50 men with trichomoniasis, 27 had urethral symptoms (see Table 1). The mean duration of symptoms reported by the 16 symptomatic contact men was 0.4 days [median, 0 days; SD 0.4 days] compared with a mean duration of 0.9 days reported by the 11 symptomatic men in the 75% random sample (median, 1 day; SD 6 days, P = 0.13 by Mann-Whitney U test). If analysis is limited to the 40 men with trichomoniasis who did not have coinfections with either N. gonorrhoeae or C. trachomatis, then 21 men had urethral symptoms. The mean duration of symptoms reported by the 13 symptomatic contact men was 0.5 days (median, 0 days; SD 0.7 days) compared with a mean duration of 0.8 days reported by the 8 symptomatic men in the 75% random sample (median, 0.5 days; SD 0.8 days, P > 0.2 by Mann-Whitney U test). Thus, the duration of urethral symptoms was similar in symptomatic men, in the contact men, and in the 75% random sample, but the numbers of men in each group were small.

Excluding men with dual infections, urethral discharge associated with trichomoniasis was usually small to moderate in amount (Figure 1) and clear to intermediate in character (Figure 2). In contrast, urethral discharge with gonorrhea was usually purulent and large or moderate in amount. Only in occasional men was the discharge associated with trichomoniasis profuse (two patients) or purulent (three patients). Overall, urethral discharge associated with trichomoniasis resembled the scant, clear discharge associated with C. trachoma-tis, more than it resembled the profuse, purulent discharge characteristic of gonorrhea.

Microbiologic Findings

Neisseria gonorrhoeae was isolated from 55 (12%) of the 447 men, including 41 (14%) of 300 among men in the 75% random sample (2 [12%]) of 17 randomly selected men with trichomoniasis and 39 [14%] of 283 without trichomoniasis) and 14 (10%) of 147 among contacts of women with trichomoniasis (4 [12%] of 33 contact men with trichomoniasis and 10 [9%] of 114 without trichomoniasis). Chlamydia trachomatis was isolated from 40 (9%), including 29 (10%) of 300 men in the 75% random sample (3 [18%] of 17 randomly selected men with trichomoniasis and 26 [9%] of 283 without trichomoniasis) and 11 (7%) of 147 contacts of women with trichomoniasis (1 [3%] of 33 contact men with trichomoniasis and 10 [9%] of 114 without trichomoniasis). Genital herpes or other sexually transmitted diseases were identified in 23 (5%) men. None of these conditions was associated with a statistically increased (P > 0.5) rate of isolation of T. vaginalis among either the 75% random sample or the contact men. We also found no statistical difference (P > 0.5) in the prevalence of the genital mycoplasmas, Mycoplasma hominis and Ureaplasma urealyticum, among the various patient populations (data not presented).

To clarify the role of T. vaginalis as a cause of urethral pathologic results, we re-evaluated our findings excluding the 10 men with dual infections by T. vaginalis with N. gonorrhoeae or C. trachomatis (Table 2). After stratifying patients by status (men in the 75% random sample or contact males), T. vaginalis was associated with symptoms (odds ratio 3.8, 95% CI, 1.7 to 8.3; P < 0.01) and signs of urethral discharge (odds ratio 2.3; CI, 1.1 to 5.1; P < 0.03), as well as with inflammation on the urethral smear (odds ratio 3.5; CI, 1.6 to 7.9; P < 0.01). Clinical diagnosis of urethritis was made in 50% of the men with T. vaginalis as the sole urethral pathogen compared with 32% of the men with none of the three pathogens (odds ratio 2.8; CI, 1.3 to 6.2, P < 0.01). Of 121 men with nongonococcal nonchlamydial urethritis in this study, 20 (17%) were associated with T. vaginalis. This includes 7 (8%) of 85 men in the 75% random sample group and 13 (36%) of 36 cases in the contacts of women with trichomoniasis (P < 0.001). Thus, T. vaginalis was associated with nongonococcal nonchlamydial urethritis in this population of sexually active men, particularly among sexual partners of women with trichomoniasis. Of the etiologic forms of urethritis summarized in Table 2, T. vaginalis was the mildest in all respects.

To identify potential confounders, we assessed the relation of trichomoniasis and nongonococcal nonchlamydial urethritis to other variables. By univariate analyses, either trichomoniasis or nonchlamydial nongonococcal urethritis showed statistically significant (P < 0.05) associations with the following variables: black race, previous trichomoniasis, method of recruitment (contact of woman with trichomoniasis or 75% random sample), previous nongonococcal urethritis, previous gonorrhea, age, and number of sex partners in the previous 6 months. Multiple logistic regression techniques [18], adjusting for these variables, were then used to evaluate the apparent association of T. vaginalis with nonchlamydial nongonococcal urethritis. Trichomonas vaginalis was the variable most strongly associated with nonchlamydial nongonococcal urethritis in this population of sexually active men (adjusted odds ratio 3.8; CI, 1.1 to 11.2; P = 0.02) followed by black race (adjusted odds ratio 3.4; CI, 1.5 to 7.7; P = 0.005). Other variables of potential clinical importance as risk factors for nonchlamydial nongonococcal urethritis included previous trichomoniasis (adjusted odds ratio 2.2; CI, 0.7 to 6.7; P = 0.19), membership in the contact group (adjusted odds ratio 2.1; CI, 0.9 to 4.8; P = 0.09), and previous treatment for nongonococcal urethritis (adjusted odds ratio 1.7; CI, 0.9 to 3.5; P = 0.15).

Discussion

Although seldom identified in men in routine clinical practice, T. vaginalis is common in those at risk for sexually transmitted diseases. We identified 50 patients among 447 heterosexual men, including 17 of 300 in a 75% random sample of men attending our sexually transmitted disease clinic and 33 of 147 sexual contacts of women with trichomoniasis. Because they attended the clinic for evaluation and treatment of new problems, men with trichomoniasis in the 75% random sample were more likely to have symptoms than men who had contact with women who had trichomoniasis (see Table 1). These contact men often attended the clinic for postexposure treatment. The 6% prevalence of T. vaginalis infection among men in the 75% random sample compares with rates of 14% for gonorrhea and 9% for C. trachomatis infection.

As anticipated, the prevalence of trichomoniasis was higher among sexual contacts of women with trichomoniasis than among men in the 75% random sample of heterosexual men attending the same clinic. The prevalence of trichomoniasis in this study was within the wide range of 14% to 60% found in previous studies [19-27]. Most of these studies were completed before 1965; they were done to support the idea that sexual contact was the primary mechanism for transmission. Clinical characterization of men in these studies was often limited, and no evaluation was done for associated genitourinary pathogens other than N. gonorrhoeae. Our study also includes a random sample of men attending our sexually transmitted disease clinic who were characterized clinically and microbiologically for several genitourinary pathogens.

We found that men with trichomoniasis were usually symptomatic. More than half complained of urethral discharge, and they had more urethral inflammation than did men without trichomoniasis. Forty (80%) of the 50 men with trichomoniasis were not infected with either N. gonorrhoeae or C. trachomatis. In the subset of men who were not infected with either of these urogenital pathogens, those with trichomoniasis were more likely to complain of urethral discharge and to have objective manifestations of urethritis. The association of T. vaginalis with nonchlamydial nongonococcal urethritis remained significant (P = 0.2) after adjustment for other variables potentially associated with an increased risk for other sexually transmitted diseases.

Urethral discharge associated with trichomoniasis tended to be only mild or moderately severe, and men with symptomatic discharge plus positive gram-stain results for urethritis often lacked evidence of urethral discharge on physical examination. These findings suggest that trichomoniasis is an important consideration in sexually active men with urethral symptoms or inflammation who have little or no discharge on physical examination. As with women [2], some men with trichomoniasis lacked any features of inflammation produced by the infection, and half did not meet our criteria for urethritis. Comparison of T. vaginalis isolates from symptomatic and asymptomatic men as well as analysis of host factors associated with the presence or absence of symptoms might help define the determinants of the inflammatory response to trichomoniasis.

Nongonococcal urethritis has been associated with trichomoniasis in other studies [23, 28, 29]. The reported prevalence of trichomoniasis has ranged from 1% to 18% among men with gonococcal or nongonococcal urethritis [19, 21, 22, 30, 31]. Other clinical syndromes attributed to trichomoniasis in previous reports include nonbacterial prostatitis [6], urethral stricture disease [23], epididymitis [32], other inflammations of the genitalia [33], and infertility [34]. None of these conditions was evident in our study, which was not specifically designed to include these syndromes.

The prevalences of N. gonorrhoeae and C. trachomatis were similar among men with and without trichomoniasis in this study. We have previously observed that infections with N. gonorrhoeae, but not with C. trachomatis, were associated with trichomoniasis in women [2]. The association of gonorrhea with trichomoniasis in women could be attributed to lower rates of early treatment for both infections in women compared with men, differences in the natural history of either infection in men and women, differences in anatomy such as exposure to periodic urine flow in the male urethra, host-defense factors such as antibacterial activity of prostatic secretions [35], or to undefined patient selection bias.

Simultaneous treatment is recommended for male sexual partners of women with trichomoniasis [36, 37]. This practice results in a small but important increase in the cure rate for women receiving single-dose therapy, although no statistical improvement in the cure rate has been shown with concomitant treatment of male partners of women treated for 7 days [38-40]. The major rationale for treating men has been to decrease subsequent transmission of infections to women. Our data show that more than half of infected men had morbidity, although other infected men remained asymptomatic. Thus, routine partner notification and treatment are warranted in most clinical settings, not only to reduce transmission to susceptible women, but also to reduce morbidity in infected men. Our observation that symptoms of urethritis in infected men consistently disappear after metronidazole therapy (including negative test-of-cure cultures; Unpublished observation) supports the recommendation for routine partner notification and treatment of both partners. Efforts at identification and treatment in the men should focus on eliciting a history of sexual exposure to trichomoniasis as well as on testing for T. vaginalis among men with the syndrome of mild urethritis.