Differences in testosterone levels in perianal diseases: A comparative study of abscesses and fistulas

Abstract

BACKGROUND

Anorectal abscess and cryptoglandular anal fistula are common conditions with high recurrence rates, reflecting an incomplete understanding of their pathophysiology. A classic, yet unproven hypothesis proposes that high androgen levels increase infection risk by stimulating anal gland secretion, potentially explaining the male predominance. However, direct evidence regarding sex hormone levels in patients remains scarce. We hypothesized that active perianal infection is associated with significantly altered testosterone (T) levels.

AIM

To investigate the association between perianal diseases and sex hormone levels and to identify factors associated with T levels.

METHODS

This cross-sectional study included 249 men: 64 with hemorrhoids, 64 with anorectal abscess, and 121 with anal fistula. Preoperative serum sex hormone levels were measured. Multivariate linear regression identified factors associated with T levels.

RESULTS

There were no significant differences in age among the three groups at baseline. T levels differed significantly among the groups (P < 0.001). Multivariate regression analysis revealed that in the overall study population, anorectal abscess (B = -2.999, P < 0.001) and body mass index (BMI; B = -0.368, P < 0.001) negatively correlated with T levels. In the abscess group, white blood cell count (B = -0.565, P < 0.001) and BMI (B = -0.298, P = 0.011) negatively correlated with T levels. In the fistula group, BMI (B = -3.866 P < 0.001) and internal opening above the dentate line (B = -4.479, P = 0.048) were associated with lower T levels.

CONCLUSION

Patients with active perianal infections exhibited a significantly lower T level compared to those with hemorrhoids or fistulas, which is independently associated with systemic inflammation severity and obesity.

Key Words: Perianal diseases; Testosterone; Sex hormones; Anorectal abscess; Anal fistula; Inflammation; Body mass index; Cross-sectional study

Core Tip: This study challenges the classic high-androgen hypothesis in perianal infections by demonstrating that patients with active anorectal abscess exhibit a significantly lower testosterone (T) level compared to those with hemorrhoids or fistulas, which is independently associated with systemic inflammation and obesity. We propose two novel hypotheses, namely a dynamic shift from high to low T, and a pathogenic role of inherent low T. The new endocrinological perspective suggests that restoring T could be a future adjunctive therapeutic strategy.

INTRODUCTION

Anorectal abscess and its chronic manifestation, anal fistula, are among the most common infectious conditions in colorectal surgery. The pathogenesis of cryptoglandular perianal fistula is generally attributed to the obstruction and subsequent infection of anal glands[1]. Although surgery is currently the gold standard of treatment[2] and numerous surgical techniques are available, recurrence rates remain unsatisfactory. This indicates persistent gaps in our understanding of the underlying factors influencing the disease. Thus, a deeper comprehension of the etiology could pave the way for innovative treatment strategies[3].

In addition to anatomical, iatrogenic, and human factors, sex hormone levels have long been recognized as important regulatory factors in the development and progression of various diseases. In the field of proctology, a classic, yet simplistic hypothesis suggests that elevated androgen levels may lead to increased secretion by the anal glands, thereby raising the risk of glandular obstruction and infection. This mechanism has been proposed as an explanation for the significantly higher incidence of these conditions in males relative to females[4]. However, this hypothesis is largely derived from physiological speculation and lacks direct evidence from patient populations. At present, there is a lack of systematic research on the actual state of sex hormone levels in patients with perianal infectious diseases.

Therefore, we designed this cross-sectional study to compare sex hormone levels among patients with anorectal abscess, anal fistula, and hemorrhoids (serving as a non-infectious control group). The study specifically aimed to: (1) Examine differences in testosterone (T) levels among these three patient populations; and (2) Identify factors independently associated with T levels in patients with anorectal abscess or anal fistula. The findings are expected to provide new endocrinological insights into the systemic pathophysiology of perianal infectious diseases and to refine current understanding of their pathogenic mechanisms.

MATERIALS AND METHODS

Settings and patients

A prospective observational study was implemented at Longhua Hospital, Shanghai University of Traditional Chinese Medicine, with patients enrolled from October 2021 to March 2023. The study was approved by the ethics review board of the hospital. For exploratory purposes, only male patients were included in the study to avoid potential interference with variations in sex hormone levels caused by menstrual cycles and menopausal status in female patients.

Study design

Anal fistulas or anorectal abscesses were diagnosed by board-certified anorectal surgeons and confirmed using ultrasonography or magnetic resonance imaging. The control group included age-matched patients who received diagnosis of hemorrhoids during the same study period. Given the high comorbidity of hemorrhoids in individuals with anorectal abscesses or anal fistulas, all participants (both cases and controls) were required to have hemorrhoids to minimize potential confounding factors at baseline. The hemorrhoids group was selected from patients with hemorrhoids who visited our hospital during the same period, and was randomly selected at a 1:1 ratio with the abscess group, with age matching (in 10-year age groups).

Inclusion criteria: (1) Male patients diagnosed with anorectal abscess, anal fistula, and hemorrhoids; (2) First hospitalization for the diagnosed disease; and (3) Not infected with coronavirus disease 2019, with a negative viral nucleic acid test.

Exclusion criteria: (1) Simultaneous diagnosis of necrotizing fasciitis; (2) Confirmed inflammatory bowel disease (IBD); (3) Presence of infectious diseases, such as tuberculosis, acquired immune deficiency syndrome, or syphilis; (4) Postoperative pathology suggesting cysts, teratomas, or malignant tumors; (5) Presence of hepatitis, cirrhosis, renal failure, or other causes of abnormal liver and kidney function; (6) Use of steroidal anti-inflammatory drugs, contraceptives, or other drugs with sex hormone effects; (7) Malignant tumors; and (8) Suspected or definite trauma-caused disease.

Procedures

To exclude IBD, patients underwent colonoscopy if clinically indicated. Those with findings conclusive for Crohn’s disease or ulcerative colitis were excluded. Cases with only minor, nonspecific inflammatory changes were included and annotated.

Basic patient information, such as duration of disease, age, and body mass index (BMI), was collected for the study. At 6:00 AM on the day of the operation, each patient’s peripheral blood was drawn in a fasting state, and the following parameters were tested: T, estradiol (E₂), follicle-stimulating hormone (FSH), and luteinizing hormone (LH).

The patients with abscesses were further grouped based on the following characteristics: White blood cell (WBC) count, overweight, concomitant diabetes, concomitant hypertension, type of abscess, fistula formation, abscesses involving more than two perianal quadrants, and T levels. The patients with anal fistula were further grouped based on the following characteristics: Overweight, diabetes mellitus, hypertension, type of anal fistula, location of the internal opening, number of fistulas, and colonoscopy report suggesting concomitant intestinal inflammatory manifestations. Then, we observed whether there were differences in T levels among different groups.

Statistical analysis

The measurement data were described by the mean ± SD, and severely skewed data were expressed as the median (P25, P75). Categorical data were presented as numbers and percentages. The t-test or corrected t-test was used for two groups of measures that followed a normal distribution, whereas the Mann-Whitney U test was used for those that did not follow a normal distribution. For comparisons of categorical variables between groups, the χ² test was applied. In this study, ANOVA was used for statistical analysis of multiple sets of measurement data; if the data showed a severely skewed distribution, the Kruskal-Wallis test was applied instead. In addition, the Bonferroni correction was employed for multiple-comparison analysis.

To examine the association between disease characteristics and T levels, we performed separate univariate analyses for the abscess and fistula groups. Variables demonstrating a marginal association (P < 0.1) in the univariate testing were included as candidate predictors for multivariate linear regression. The final model was constructed through stepwise variable selection. For correlation tests, the Pearson’s test was used in the case of a normal distribution, while the Spearman’s rank-sum test was applied in the case of non-normal distribution. Model significance was assessed using ANOVA, with multicollinearity evaluated via variance inflation factor (VIF; threshold < 5). Residual independence was verified using the Durbi-Watson test to exclude autocorrelation.

All statistical calculations and graphs relied on standard software (SPSS version 30.0; SPSS Inc.), with significance set at P < 0.05.

Priori sample size calculation for the anorectal abscess group, based on an estimated standard deviation of 3.67 nmol/L, a margin of error of 1 nmol/L, and a two-sided alpha of 0.05, indicated a minimum requirement of 52 patients. This target was exceeded in all the groups. A larger sample size, particularly in the anal fistula group, was intentionally enrolled to facilitate robust age matching and enable exploratory analyses across anatomical subtypes. Post-hoc power analysis confirmed that the final cohort (n = 249) provided a high statistical power of 0.94 for detecting the observed differences in T levels among the three groups.

RESULTS

Clinical characteristics and hormone levels

Ultimately, a total of 249 male patients were included, encompassing 64 patients with hemorrhoids, 64 patients with anorectal abscess, and 121 patients with anal fistula.

The clinical, demographic, and endocrine characteristics of the study population are summarized in Table 1.

Table 1 Baseline characteristics and comparison of sex hormone levels among the three groups.

Group	Hemorrhoids (n = 64)	Anorectal abscess (n = 64)	Anal fistula (n = 121)	Test statistic	P value
Age (years)	37.50 (31.00, 49.00)	35.00 (29.25, 49.75)	36.00 (29.00, 49.00)	0.8091	0.667
Symptom duration (months)	48.00 (24.00, 120.00)a	0.23 (0.18, 1.00)2	10.00 (3.00, 24.00)a,2	119.8911	< 0.001
BMI (kg/m2)	24.25 ± 5.49	25.00 ± 4.06	25.60 ± 3.48	2.1722	0.116
E2 (pg/mL)	21.78 ± 7.12	21.87 ± 7.90	23.57 ± 7.88	1.6062	0.203
FSH (IU/L)	5.42 ± 4.13	5.36 ± 3.92	5.09 ± 4.20	0.1842	0.173
LH (IU/L)	4.31 (2.73, 6.55)	4.51 (3.14, 7.56)	4.08 (2.76, 5.50)	3.8211	0.148
T (nmol/L)	13.94 ± 5.53a	10.68 ± 4.84	13.08 ± 5.25a,b	6.9322	< 0.001
Hypertension	9 (14.06)	9 (14.06)	16 (13.22)	0.0373	0.982
Diabetes mellitus	14 (21.88)	14 (21.88)	18 (14.88)	2.0233	0.364

^aP < 0.05 vs the abscess group.

^bP < 0.05 vs the hemorrhoids group.

¹Kruskal-Wallis test.

²One-way ANOVA.

³χ² test.

Data were showed as mean ± SD, n (%) or median (P25, P75). BMI: Body mass index; E₂: Estradiol; FSH: Follicle-stimulating hormone; LH: Luteinizing hormone; T: Testosterone.

There were no significant differences in age, BMI, or demographic characteristics among the three groups, indicating successful matching on these parameters. As expected, a significant difference in symptom duration was observed among the groups (η² = 0.48), where the hemorrhoids group showed the longest median disease duration. Analysis of preoperative sex hormone levels revealed no significant inter-group differences in E₂, FSH, or LH concentrations (all P > 0.05). Post-hoc analysis confirmed that the anorectal abscess group had significantly lower mean T levels compared with both the hemorrhoids group [mean difference: 3.27 nmol/L; 95% confidence interval (95%CI): 1.45-5.08; P < 0.001] and the anal fistula group (mean difference: 2.40 nmol/L; 95%CI: 0.81-3.99; P = 0.030). There was no significant difference in T levels between the hemorrhoids and fistula groups.

To assess the independent association between anorectal abscess and T levels, multiple linear regression analysis was performed adjusting for age, BMI, hypertension, diabetes mellitus, and disease group (Table 2).

Table 2 Independent association between anorectal abscess and lower testosterone levels in the overall study population.

Predictor	B	SE	β	t value	P value	B diff (95%CI)	VIF
Constant	23.946	2.161	-	11.081	< 0.001	19.69-28.20	-
Age	-0.032	0.027	-0.074	-1.156	0.249	-0.09 to 0.02	1.17
BMI	-0.368	0.076	-0.292	-4.850	< 0.001	-0.52 to -0.22	1.04
Hypertension	-0.377	0.999	-0.024	-0.378	0.706	-2.35 to 1.59	1.19
Diabetes mellitus	1.056	0.815	0.077	1.296	0.196	-0.55 to 2.66	1.01
Anorectal abscess1	-2.999	0.880	-0.246	-3.408	< 0.001	-4.73 to -1.27	1.49
Anal fistula1	-0.346	0.778	-0.032	-0.444	0.657	-1.88 to 1.19	1.53

¹The reference group for disease category is the hemorrhoids group.

Model summary: Adjusted R² = 0.136, F (6, 242) = 7.485, P < 0.001. Multicollinearity: All variance inflation factor values were < 1.6, indicating no substantial multicollinearity. BMI: Body mass index; 95%CI: 95% confidence interval; VIF: Variance inflation factor.

After adjusting for covariates, a diagnosis of anorectal abscess (B = -2.999, P < 0.001) and higher BMI (B = -0.368, P < 0.001) were identified as significant, independent negative predictors of T levels. No other variables, including anal fistula, showed significant associations in the model (Adjusted R² = 0.136; all VIFs < 1.6).

Determinants of T levels

To explore the sources of differences in T levels, we carried out univariate analysis of disease characteristics of patients with anorectal abscess (shown in Tables 3 and 4).

Table 3 Univariate analysis between disease characteristics and testosterone levels in the patients with anorectal abscess.

Grouping category	n (%)	T (nmol/L; mean ± SD)	Mean/median diff (95%CI)	Test statistic	P value
Hypertension
No	55 (85.9)	11.07 ± 4.95	Reference	2.142	0.050
Yes	9 (14.1)	8.27 ± 3.37	2.80 (0.00-5.60)
Diabetes mellitus
No	50 (78.1)	10.08 ± 4.18	Reference	-1.921	0.059
Yes	14 (21.9)	12.83 ± 6.41	-2.75 (-5.61 to 0.11)
BMI (kg/m2)
≤ 24	28	11.75 ± 5.29	Reference	1.983	0.056
> 24	36	9.84 ± 4.34	1.91 (-0.50 to 4.31)
Symptom duration (days)
< 7	21 (32.8)	8.51 ± 3.62	Reference	-2.627	0.011
≥ 7	43 (67.2)	11.74 ± 5.03	-3.23 (-5.69 to -0.77)
Type of abscess
Perianal subcutaneous	48 (75.0)	10.61 ± 4.99	Reference	0.224	0.879
Ischioanal abscess	10 (15.6)	10.2 8 ± 5.47	0.33 (-3.09 to 3.76)
Intersphincteric abscess	5 (7.8)	12.33 ± 2.15	-1.72 (-6.35 to 2.92)
Submucosal abscess	1 (1.6)	9.721	0.891
Fistula formation
No	18 (28.1)	8.82 ± 5.56	Reference	-2.226	0.032
Yes	46 (71.2)	11.35 ± 5.09	-2.53 (-4.83 to -0.23)
Abscesses involving more than two perianal quadrants
No	40 (62.5)	11.62 ± 4.17	Reference	2.057	0.044
Yes	24 (37.5)	9.11 ± 5.52	2.50 (0.07-0.49)

¹There was only one sample in the group, so no pairwise comparisons were conducted.

T: Testosterone; 95%CI: 95% confidence interval; BMI: Body mass index.

Table 4 Correlation analysis of disease characteristics and testosterone levels in the patients with anorectal abscess (n = 64).

Variable	r value1	95%CI	P value
Age	-0.018	-0.27 to 0.24	0.886
WBC (109/L)	-0.467	-0.64 to -0.24	< 0.001

¹Pearson correlation coefficient.

95%CI: 95% confidence interval; WBC: White blood cell.

Given that this was an exploratory study, variables with P < 0.1 (hypertension, diabetes mellitus, BMI, symptom duration, fistula formation, abscesses involving more than two perianal quadrants, and WBC count) were used as the independent variables, while T level was selected as the dependent variable in multiple linear regression analysis. The results are shown in Table 5.

Table 5 Multivariate linear regression analysis of testosterone levels in the patients with perianal abscess (n = 64).

Predictor	B	SE	β	t value	P value	B diff (95%CI)	VIF
Constant	14.995	2.138		7.014	< 0.001	10.71-19.28
WBC (109/L)	-0.565	0.160	-0.441	-3.540	< 0.001	-0.885 to -0.245	1.165
BMI	-2.879	1.088	-0.298	-2.647	0.011	-5.058 to -0.700	1.138

Model summary: R² = 0.377, adjusted R² = 0.300, F (7, 56) = 4.851, P < 0.001. Multicollinearity: All variance inflation factor values were < 1.2, indicating no substantial multicollinearity. 95%CI: 95% confidence interval; VIF: Variance inflation factor WBC: White blood cell; BMI: Body mass index.

The multivariate linear regression model significantly predicted T levels in patients with anorectal abscess [F (7, 56) = 4.851, P < 0.001], explaining 30.0% of the variance (adjusted R² = 0.300). Higher WBC count (B = -0.565, β = -0.441, P < 0.001) and higher BMI (B = -0.289, β = -0.298, P = 0.011) were both significant independent negative predictors. The model showed no multicollinearity (all VIFs < 1.2) and residual independence (Durbin-Watson = 1.901).

We next performed a similar analysis to identify the factors associated with T levels in patients with anal fistula, a condition characterized by chronic inflammation. The univariate associations are presented in Tables 6 and 7.

Table 6 Univariate analysis between disease characteristics and testosterone levels in the patients with anal fistula (n = 121).

Grouping category	n (%)	T (nmol/L), mean ± SD or median (P25, P75)	Mean/median diff (95%CI)	Test statistic	P value
Diabetes mellitus
No	103 (85.1)	12.30 (9.25, 16.60)	Reference	0.328	0.743
Yes	18 (14.9)	12.95 (9.93, 16.15)	-0.41 (-3.05 to 2.10)
Hypertension
No	105 (86.8)	13.37 ± 5.33	Reference	1.631	0.105
Yes	16 (13.2)	11.02 ± 4.17	2.34 (-0.50 to 5.19)
BMI (kg/m2)
≤ 24	44 (36.4)	15.84 ± 4.88	Reference	4.766	0.001
> 24	77 (63.6)	11.45 ± 4.80	4.35 (2.54-6.16)
Type of anal fistula
Trans sphincteric	84 (69.4)	13.25 ± 5.23	Reference	8.897	0.064
Inter sphincteric	17 (14)	13.09 ± 6.07	0.16 (-3.78 to 4.10)
Supra sphincteric	8 (6.6)	14.62 ± 4.65	-1.37 (-6.85 to 4.11)
Extra sphincteric	10 (8.3)	9.43 ± 3.23	3.82 (-1.13 to 8.77)
Submucosal	2 (1.7)	17.75 ± 1.63	-4.50 (-15.10 to 6.10)
Internal opening above the dentate line
No	112 (92.6)	13.42 ± 5.27	Reference	4.460	0.001
Yes	9 (7.4)	8.81 ± 2.72	4.61 (2.38-6.84)
Number of fistula tracts
< 2	97 (80.2)	12.97 ± 5.17	Reference	-0.462	0.645
≥ 2	24 (19.8)	13.52 ± 5.65	-0.55 (-2.93 to 1.82)
Intestinal inflammatory manifestations (n = 73)
No	55 (75.3)	11.7 (9.25, 15.30)	Reference	-0.557	0.578
Yes	18 (24.7)	13.10 (8.13, 16.00)	-0.07 (-3.40 to 2.10)

T: Testosterone; 95%CI: 95% confidence interval; BMI: Body mass index.

Table 7 Correlation analysis of disease characteristics and testosterone levels in the patients with anal fistula (n = 121).

Category	Correlation coefficient (ρ)1	95%CI	P value
Age	-0.201	-0.37 to -0.02	0.027
Symptom duration (months)	-0.029	-0.21 to 0.16	0.775

¹Spearman's rank correlation coefficient.

95%CI: 95% confidence interval.

Given that this was an exploratory study, variables with P < 0.1 (BMI, type of anal fistula, internal opening above the dentate line, and age) were used as the independent variables, while T level was selected as the dependent variable in multiple linear regression analysis. The results are shown in Table 8.

Table 8 Correlation analysis of disease characteristics and testosterone levels in the patients with anal fistula.

Predictor	B	SE	β	t value	P value	B diff (95%CI)	VIF
Constant	18.030	1.492		12.083	< 0.001	(15.07-20.99)
BMI	-3.866	0.917	-0.356	-4.214	< 0.001	(-5.68 to -2.05)	1.054
Internal opening above the dentate line	-4.479	2.238	-0.225	-2.001	0.048	(-8.91 to -0.05)	1.866

Model summary: R² = 0.215, adjusted R² = 0.188, F (4, 116) = 7.941, P < 0.001. Durbin-Watson ratio = 1.992. 95%CI: 95% confidence interval; VIF: Variance inflation factor; BMI: Body mass index.

In patients with anal fistula, higher BMI (B = -3.866, P < 0.001) and an internal opening located above the dentate line (B = -4.479, P = 0.048) were significant independent predictors of lower T levels in the multivariate linear regression model (adjusted R² = 0.188, P < 0.001). The model showed no multicollinearity (all VIFs < 1.9) or autocorrelation.

DISCUSSION

In this study, we compared sex hormone levels among patients with anorectal abscess, anal fistula, and hemorrhoids, revealing a previously underrecognized endocrine characteristic in perianal infectious diseases. Our study revealed that patients with active anorectal infectious diseases exhibited a significantly lower T level compared to those with fistulas or hemorrhoids, which was independently associated with systemic inflammation severity and obesity.

It should be noted that due to the limitations of the research methods, the conclusions of multiple models in our study should be regarded as exploratory rather than definitive. In this study, we selected patients with hemorrhoids as the control group given their high comorbidity with anorectal abscesses or fistulas. Their inclusion in all of the groups helped minimize baseline confounding and allowed for a more specific comparison of endocrine differences attributable to infectious status. Patients with hemorrhoids may have long-term and persistent chronic inflammation. Considering that T levels are negatively correlated with inflammation levels, the T levels of patients with hemorrhoids may be similar to or lower than those of healthy people. The T level of patients with anorectal abscesses is lower than that of patients with hemorrhoids, which means that the T level of patients with anorectal abscesses may be even lower than that of healthy people, but this needs further confirmation. Therefore, it must be acknowledged as a limitation of this study that chronic hemorrhoidal disease itself may not represent a perfectly healthy endocrine state. The subsequent discussion will be based on the limitations of this study.

A strong inverse correlation between systemic inflammation and T levels is well-established. This is evidenced by studies linking endotoxemia to suppressed T[5], and clinical observations demonstrating that elevated C-reactive protein (CRP) is a robust, independent predictor of low bioavailable T and biochemical hypogonadism[6,7], an association also noted in diabetic populations[8]. The underlying mechanism likely involves proinflammatory cytokines (e.g., IL-1, IL-6, TNF-α) directly inhibiting Leydig cell steroidogenesis[9] and/or suppressing the hypothalamic-pituitary-gonadal axis[10]. While inflammation is recognized as an independent risk factor for hormonal and metabolic disturbances[11], its specific role in anorectal abscesses has been unexplored, a gap that our study addresses. Patients with IBD may also present perianal symptoms, which is a pathophysiological mechanism that also occurs in this group. Disease severity correlates with hypogonadism and impaired semen quality[12,13], and genetic evidence suggests a potential bidirectional causality with sex hormones[14].

Obesity is also closely associated with low T. Through multiple mechanisms—including triggering the release of inflammatory factors, suppressing hypothalamic-pituitary function, reducing sex hormone-binding globulin, and increasing the aromatization of T to E₂—obesity leads to a functional hypogonadal state[11]. Furthermore, obesity itself is a well-established independent risk factor for anorectal abscess[15-17]. Włodarczyk et al[18] demonstrated that complex cryptoglandular anal fistulas are associated with reduced levels of anti-inflammatory adipokines, such as adiponectin. This suggests that obesity may foster a synergistic proinflammatory environment characterized by both low T and dysregulated adipokine signaling, which collectively drive disease progression.

The classic theory regarding the pathogenesis of cryptoglandular anal fistula posits that elevated androgen levels lead to increased secretion by the anal glands, subsequently causing their obstruction and infection. This appears to contradict our finding that active infection is associated with low T levels. We propose two possible hypotheses for this apparent discrepancy. The first hypothesis attempts to reconcile our findings with the classical theory: We postulate a dynamic shift in T levels during the disease process, from an initial high to a subsequent low state. In the preliminary stage, a higher baseline T level may induce excessive anal gland secretion, thereby increasing the risk of glandular obstruction and secondary infection. Following the establishment of infection, sharply elevated inflammatory factors suppress the gonadal axis, potentially leading to secondary, reactive hypogonadism observed in our study. Limited by the cross-sectional design, this proposed causal sequence requires future validation through measurement of pre-disease baseline T levels.

The second hypothesis, grounded in existing research, proposes that an inherently low T level may itself act as a predisposing factor for perianal inflammatory diseases. T has been demonstrated to exert anti-inflammatory effects[19] and negatively correlates with proinflammatory parameters[20]. T replacement therapy has been shown to reduce proinflammatory cytokine levels[21]. The mechanisms may involve the suppression of key proinflammatory factors such as TNF-α and IL-6, and the promotion of macrophage polarization toward the reparative M2 phenotype[22]. Therefore, we speculate that an inherently low T state may compromise local immune defense and tissue repair capacity, thereby creating a condition conducive to the progression of minor infections into abscesses or fistulas. A similar speculation has been proposed by Judge et al[23] in their study on IBD patients.

The hypothesis that low T may promote perianal inflammatory diseases offers a novel therapeutic perspective. Beyond traditional surgical approaches, restoring physiological T levels may potentially improve outcomes in refractory anal fistula and abscess. This could be explored through anti-inflammatory strategies, weight loss interventions[24], or T supplementation under strict medical supervision. Although Nasser et al[25] and Mason et al[26] have explored hormone replacement therapy in Crohn’s disease and adolescents with IBD, observing some benefit, current clinical guidelines recommend T treatment only for symptomatic men with androgen deficiency due to underlying hypothalamic, pituitary, or testicular disorders[27]. Therefore, any future interventional studies should prioritize thorough safety evaluations, particularly considering the complex and potentially dual role of androgens in immune regulation.

This study has several limitations. Its cross-sectional design precludes causal inference, and the lack of a healthy control group means that the findings indicate relative differences between disease groups. The all-male cohort limits generalizability to females, and the absence of pre-disease hormone levels prevents direct validation of etiological hypotheses. Furthermore, unmeasured confounders (e.g., smoking and alcohol) and the modest explanatory power of our models suggest that additional, unidentified factors influence T levels. Finally, the results of the multivariate models should be interpreted as exploratory. Subsequent longitudinal studies should be carried out and the condition of inflammatory factors such as IL-6, CRP, and TNF-α should be paid attention to.

CONCLUSION

In summary, patients with anorectal abscesses exhibit significantly lower T levels compared with those with hemorrhoids or anal fistulas, and the difference is independently associated with systemic inflammation severity and obesity. Our findings challenge the conventional view of the endocrine environment in perianal diseases. We propose two hypotheses, namely a transition from initially high to subsequently low T levels, and the potential to initiate and promote disease development by an inherently low T level. This endocrinological perspective offers a new avenue for understanding recurrence and developing adjunctive therapeutic strategies.

ACKNOWLEDGEMENTS

Thanks to the support of the Department of Anorectal, Longhua Hospital, Shanghai University of Traditional Chinses Medicine.