Targeting obesity and lipid metabolism profiles to prevent perianal abscesses: A case-control study and Mendelian randomization analysis

Abstract

BACKGROUND

Perianal abscesses (PAs) are associated with significant complications, such as recurrent infections, pain, anal fistulas, rectovaginal fistulas, rectourethral fistulas, and rectovesical fistulas. However, established primary and secondary prevention strategies for PAs are lacking.

AIM

To explore the relationships between obesity and lipid metabolites, including perianal abscess onset.

METHODS

We conducted two independent studies under a unified research question. Case-control analysis was conducted at a single hospital between May 2023 and November 2023. Inpatients diagnosed with a perianal abscess and matched healthy controls were included. Body dimensions and serum metabolites were measured. Genome-wide association study data regarding genetic variants of PAs, obesity, and serum metabolites were obtained for the Mendelian randomization (MR) analysis. The study outcomes were perianal abscess onset and the number and location of PAs.

RESULTS

In the case-control study, higher body mass index (BMI), waist-to-hip ratio (WHR), waist-to-height ratio (WHtR), blood glucose levels, uric acid (UA) levels, total cholesterol levels, triglyceride levels, and low-density lipoprotein (LDL) levels were associated with increased risk of PAs. Higher high-density lipoprotein levels were associated with reduced risk of PAs. The BMI, WHR, WHtR, UA level, triglyceride level, and LDL level were associated with the number and severity of PAs. In MR analysis, the BMI, WHR, body fat percentage, whole body fat mass, limb fat percentage, limb fat mass, and various lipid profiles were significantly associated with the risk of PAs.

CONCLUSION

A hospital-based case-control study and an independent MR analysis consistently support obesity and lipid metabolism profiles are associated with an increased risk of perianal abscess. These findings provide a basis for developing primary and secondary prevention strategies for perianal abscess.

Key Words: Perianal abscess; Prevention; Obesity; Lipid metabolism profile; Case-control study; Mendelian randomization

Core Tip: Our case-control study revealed that the risk of perianal abscesses (PAs) was significantly correlated with various obesity/metabolic markers, including higher body mass index (BMI), waist-to-hip ratio (WHR), waist-to-height ratio, glucose levels, uric acid levels, cholesterol levels, triglyceride levels, and low-density lipoprotein (LDL) cholesterol levels, as well as lower levels of high-density lipoprotein cholesterol. The mendelian randomization analysis further confirmed relationships between the PA risk and factors such as BMI, WHR, large and medium LDLs, and lipid ratios. These findings suggest that lipid levels, body fat distribution, and metabolic markers are crucial to the understanding of PA risk.

INTRODUCTION

Perianal abscesses (PAs) are painful lumps around the anus caused by infectious anorectal diseases; they are often accompanied by redness, swelling, and a burning sensation[1]. In the United States, approximately 68000-96000 individuals are affected by PAs annually[2]; in contrast, the United Kingdom reports approximately 18000 cases each year[3,4]. Because of the anatomical structure of the anus and hormonal influences, PAs are more prevalent in male patients[5] (male-to-female ratio of approximately 8:1 and 9:1), predominantly affecting individuals aged 20-40 years[6]. PAs usually present acutely; however, inadequate or delayed treatment may result in severe complications such as necrotizing fasciitis or sepsis, thus posing a significant threat to life. Chronic PAs can cause recurrent infections, pain, anal fistulas, rectovaginal fistulas, rectourethral fistulas, and rectovesical fistulas, potentially leading to cancer and severely impacting the quality of life[7-10].

The precise etiology of PAs is not well understood. Studies indicate that factors such as anal gland infections[11], central gap infections[12], immune functions[13], inflammatory bowel disease[14], and sex hormones significantly contribute to their development[15]. Lifestyle factors, including an unhealthy diet, diabetes, sedentary habits, lack of exercise, and irregular bowel movements are associated with higher incidence of PAs[16]. Zhang et al[17] identified mean platelet volume as an inflammatory marker indicative of the severity of PAs. Additionally, obesity and various abnormalities may influence the occurrence and severity of PAs. Adamo et al[4] demonstrated that obesity elevates the risk of PAs. Although previous studies have suggested that overweight and obesity are potential risk factors for PAs[18], a direct relationship involving obesity, metabolic markers, and PA has not yet been established.

Most prior epidemiology on PA has relied on body mass index (BMI) alone, few studies have evaluated distributional adiposity metrics [e.g., waist-to-hip ratio (WHR) or waist-to-height ratio (WHtR)] alongside a granular lipid panel [triglycerides (TG), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), uric acid (UA)]. Additionally, prior studies linking adiposity and perianal abscess are largely correlational, leaving reverse causation and residual confounding unresolved. We therefore complemented our case-control analysis with two-sample Mendelian randomization (MR) to strengthen interpretation. MR uses germline variants associated with adiposity/Lipid traits as instruments, which are allocated at conception and thus less prone to confounding and reverse causation[19].

We designed a case-control study of patients with PAs to identify potential exposure factors. Subsequently, we performed a two-sample MR analysis based on a population to further validate the relationships between obesity, metabolic markers, and PAs.

MATERIALS AND METHODS

Research design

This study integrates two independent lines of evidence addressing the same etiologic question. First, we performed a single-center case-control study to estimate associations between adiposity/Lipid traits and PAs. Second, we conducted a two-sample MR analysis using genome-wide association summary statistics to infer causality under standard MR assumptions (Figure 1).

Figure 1 Schematic overview of the study design. BMI: Body mass index; CHO: Cholesterol; GA: Glycated albumin; GLU: Glucose; GWAS: Genome-wide association study; HDL-C: High-density lipoprotein cholesterol; LDL-C: Low-density lipoprotein cholesterol; MR: Mendelian randomization; PA: Perianal abscess; SNP: Single nucleotide polymorphism; TG: Triglycerides; UA: Uric acid; WHR: Waist-to-hip ratio; WHtR: Waist-to-height ratio.

Study variables and covariates

We now present a directed acyclic graph (DAG) specifying our target estimand (Supplementary material). From this DAG, the minimally sufficient adjustment set is: (1) Age; and (2) Sex. To investigate the associations between obesity, metabolic markers, and PA outcomes, we used PA onset as the primary outcome and the number and location of PAs as secondary outcomes. The main exposure factors included BMI, WHR, WHtR, and metabolic markers such as glucose, glycated albumin (GA), UA, cholesterol, TG, HDL-C, and LDL-C. Covariates included sex and age.

A case-control study

Our study included 1071 participants from the Anorectal Center at China-Japan Friendship Hospital in Beijing, China. This cohort comprised patients seeking treatment and healthy individuals undergoing medical examinations. We collected statistical data, anthropometric measurements (height, weight, waist, and hip circumference), metabolic indices, and information regarding PAs (presence or absence, number, and location). The case group comprised 476 patients with PA, and the control group included 595 healthy individuals. Details regarding the diagnostic criteria, inclusion and exclusion criteria, and sample size calculation methods are provided in Supplementary Tables 1 and 2. The sample size calculation is provided in Supplementary material. The study was conducted in accordance with the Declaration of Helsinki. It was approved by the Medical Ethics Committee of China-Japan Friendship Hospital, Beijing (No. 2024-KY-101). All participants provided informed consent.

MR analysis

Data regarding obesity, metabolic markers, and PAs were obtained from the Integrative Epidemiology Unit Open GWAS database, an open-access resource dedicated to facilitating the research and application of genome-wide association data. Details of the GWAS datasets used for this study are detailed in Supplementary Table 3. We performed a two-sample MR analysis to evaluate the relationships between obesity, metabolic markers, and PA outcomes. Single nucleotide polymorphisms (SNPs) were the instrumental variables and were selected on the basis of the following criteria: (1) The SNP must be significantly associated with the marker (P < 5 × 10^-8); (2) The SNP must be independent of any confounding factors; and (3) The SNP must influence the outcome solely through its association with the marker.

To mitigate linkage disequilibrium effects, we excluded SNPs with R² < 0.001 and a distance < 10000 kb. Additionally, we excluded weak instruments with an F < 10 to reduce potential bias.

Statistical analysis

For the case-control study, continuous variables that were not normally distributed are shown as median [interquartile range (IQR)]; categorical variables are shown as n (%). We used Spearman correlation to examine relationships among obesity and metabolic markers, and the Mantel test to further assess the correlation between these factors and PAs, including the number and location of PAs.

To estimate associations between obesity/metabolic markers and PAs, we fitted unadjusted and adjusted regression models, odds ratio (OR) (or β coefficients where appropriate) with 95%CI were reported. Adjustments were made for age and sex. Sensitivity analyses handled missing data with Markov chain Monte Carlo-based multiple imputation, repeated the models in complete cases, and applied inverse-probability weighting.

For MR, the inverse-variance weighted method was the primary analysis. We assessed robustness with MR-Egger, weighted median, simple mode, and weighted mode methods. Horizontal pleiotropy was evaluated using the MR-Egger intercept, and we performed leave-one-out analyses to assess the influence of individual variants.

All analyses were conducted using R software (version 4.2.1) and the two-sample MR package (version 0.5.6). P < 0.05 was considered significant for both the case-control study and MR analysis.

RESULTS

Baseline characteristics

This study initially included 1160 Chinese participants, with 514 in the case group (PAs). A total of 38 participants were excluded (18 diagnosed with secondary PA and 20 on long-term lipid/glucose/UA-regulating agents). The control group comprised 646 participants; 51 were excluded (37 without blood metabolic index examinations and 14 on long-term lipid/glucose/UA-regulating agents). Finally, 1071 participants were included in the analysis, and there were 476 new cases of PAs (Figure 1). Table 1 presents the baseline characteristics and serum metabolic data of the participants. The median age was 36.0 years (IQR: 31.0-44.0 years). In comparison with the control group (n = 595), the PA group (n = 476) exhibited significant differences in several characteristics, including age, sex, BMI, WHR, WHtR, glucose, GA, UA, cholesterol, TG, HDL-C, and LDL-C (P < 0.05).

Table 1 Baseline characteristics of all participants, n (%).

Variable	Status	Total (n = 1071)	Normal (n = 595)	Case (n = 476)	P value	W value
Age (years) [median (IQR)]		36.0 (31.0-44.0)	38.0 (32.0-47.0)	35.0 (30.0-42.0)	< 0.001	166281.5
Sex	Female	398 (37.2)	321 (53.9)	77 (16.2)	< 0.001
	Male	673 (62.8)	274 (46.1)	399 (83.8)
Body mass index (kg/m2) [median (IQR)]		24.2 (21.8-27.2)	23.1 (20.8-25.6)	26.1 (23.3-29.4)	< 0.001	83028.5
waist-to-hip ratio [median (IQR)]		0.8 (0.8-0.9)	0.8 (0.8-0.8)	0.9 (0.9-1.0)	< 0.001	40010.5
waist-to-height ratio [median (IQR)]		0.5 (0.4-0.5)	0.4 (0.4-0.5)	0.5 (0.5-0.6)	< 0.001	50096
glucose (mmol/L) [median (IQR)]		5.0 (4.7-5.5)	5.0 (4.6-5.3)	5.2 (4.8-5.6)	< 0.001	109062.5
glycated albumin (%) [median (IQR)]		13.2 (12.4-14.2)	13.4 (12.6-14.3)	13.0 (12.1-14.0)	< 0.001	167984
Urea (mmol/L) [median (IQR)]		4.5 (3.9-5.4)	4.5 (3.8-5.4)	4.6 (3.9-5.5)	0.048	131657.5
Uric acid (μmol/L) [median (IQR)]		358.0 (290.0-420.0)	323.0 (268.0-384.0)	403.0 (343.0-473.5)	< 0.001	78655
Cholesterol (mmol/L) [median (IQR)]		4.6 (4.0-5.1)	4.4 (3.8-4.9)	4.8 (4.3-5.4)	< 0.001	96592.5
Triglycerides (mmol/L) [median (IQR)]		1.1 (0.8-1.6)	0.9 (0.7-1.4)	1.4 (1.0-2.1)	< 0.001	84128.5
High-density lipoprotein cholesterol (mmol/L) [median (IQR)]		1.2 (1.0-1.5)	1.3 (1.1-1.5)	1.1 (1.0-1.3)	< 0.001	199218.5
Low-density lipoprotein cholesterol (mmol/L) [median (IQR)]		2.5 (2.1-2.9)	2.3 (2.0-2.7)	2.7 (2.3-3.2)	< 0.001	94032.5

W: Mann-Whitney test’s W statistics. P < 0.05 was considered statistically significant. IQR: Interquartile range.

Risk factors for PAs

A correlation analysis using Spearman coefficients and Mantel tests revealed significant associations between obesity/metabolic markers and PA outcomes. Specifically, the PA incidence was significantly correlated with BMI (R = 0.08), WHR (R = 0.30), WHtR (R = 0.25), glucose (R = 0.02), GA (R = 0.01), UA (R = 0.11), cholesterol (R = 0.03), TG (R = 0.05), HDL-C (R = 0.06), and LDL-C (R = 0.07) (P < 0.001). Additionally, BMI, WHR, WHtR, UA, TG, and LDL-C were significantly correlated with the lesion number. BMI, WHR, WHtR, GA, UA, cholesterol, TG, HDL-C, and LDL-C were also significantly correlated (P < 0.05) with the PA location (Figure 2A and Supplementary Table 4).

Figure 2 Mantel test and logistic regression analysis of exposure factors and perianal abscess outcomes. A: Mantel test. Mantel P value: P value; Mantel r: Mantel correlation coefficient; Spearman r: Spearman correlation coefficient. Not significant: P > 0.05. Outcome diagnosis, diagnosed with perianal abscess; outcome number, number of abscesses; outcome location, location of abscesses. ^aP < 0.05; ^bP < 0.01; ^cP < 0.001; B: Logistic regression analysis. Adjusted OR: Odds ratio value of the logistic regression model with sex and age; Adjusted P value: P value of the logistic regression model with sex and age; BMI: Body mass index; CHO: Cholesterol; Crude OR: Odds ratio value of the logistic regression model without sex and age; Crude P value: P value of the logistic regression model without sex and age; GA: Glycated albumin; GLU: Glucose; HDL-C: High-density lipoprotein cholesterol; LDL-C: Low-density lipoprotein cholesterol; OR: Odds ratio; TG: Triglycerides; UA: Uric acid; WHR: Waist-to-hip ratio; WHtR: Waist-to-height ratio.

A binary logistic regression analysis was conducted to further explore PA risk factors. After adjusting for confounding variables, the analysis showed positive correlations between the PA risk and BMI (OR = 1.16; 95%CI: 1.12-1.20), WHR (OR = 8.43; 95%CI: 6.36-11.35), WHtR (OR = 10.10; 95%CI: 7.32-14.18), glucose (OR = 1.44; 95%CI: 1.23-1.69), UA (OR = 1.00; 95%CI: 1.00-1.01), cholesterol (OR = 1.81; 95%CI: 1.56-2.13), TG (OR = 1.82; 95%CI: 1.52-2.20), and LDL-C (OR = 2.20; 95%CI: 1.78-2.75). Conversely, HDL-C was negatively associated with the PA risk (OR = 0.17; 95%CI: 0.10-0.28) (P < 0.05; Figure 2B).

A further evaluation using multiclass logical regression and linear regression was performed to assess the relationships between obesity/metabolic markers and lesion characteristics, including the location and number of PAs (Supplementary Tables 5 and 6). The results indicated that BMI, WHR, WHtR, UA, TG, and LDL-C were particularly associated with the increased number and severity of PA lesions.

Subgroup and sensitivity analyses

We conducted a subgroup analysis based on the median age and sex. For male participants, glucose (OR = 1.74; 95%CI: 1.40-2.21) and UA (OR = 1.01; 95%CI: 1.00-1.01) were significant risk factors for PAs, whereas HDL-C (OR = 0.10; 95%CI: 0.05-0.20) was a protective factor for PAs. For female participants, these associations were not significant (P > 0.05). Additionally, WHR, WHtR, cholesterol, TG, and LDL-C were more strongly associated with PAs in male participants than with PAs in female participants (Figure 3A).

Figure 3 Subgroup analysis of exposure factors and perianal abscess outcomes according to age and sex. A: Subgroup analysis according to age; B: Subgroup analysis according to sex. P < 0.05 was considered statistically significant. Adjusted OR: Odds ratio value of the logistic regression model with sex and age; Adjusted P value: P value of the logistic regression model with sex and age; BMI: Body mass index; CHO: Cholesterol; Crude OR: Odds ratio value of the logistic regression model without sex and age; Crude P value: P value of the logistic regression model without sex and age; GA: Glycated albumin; GLU: Glucose; HDL: High-density lipoprotein; LDL-C: Low-density lipoprotein cholesterol; TG: Triglycerides; UA: Uric acid; WHR: Waist-to-hip ratio; WHtR: Waist-to-height ratio.

For participants younger than 36 years, GA (OR = 0.83; 95%CI: 0.72-0.96) was a protective factor for PAs. However, this protective effect was not statistically significant in participants older than 36 years (P > 0.05). The WHR, WHtR, cholesterol, and LDL-C were risk factors for PAs in both older and younger participants; however, their impact was more pronounced in older participants (Figure 3B). We also analyzed the influences of age and sex on the number and location of PAs (Supplementary Tables 7-10). The sensitivity analysis results were consistent with the main findings (Supplementary Tables 11-16).

MR analysis

Exposure factors associated with PAs: To further investigate the relationships between PA and various exposure factors, we performed a two-sample MR analysis of obesity and metabolism data from the GWAS database. The MR analysis results corroborated our case-control study findings. BMI and WHR were significant risk factors for PAs (P < 0.05). Additionally, several markers related to body fat were identified as risk factors for PA. These included the body fat percentage, whole body fat mass, arm fat (right) mass, arm fat (left) percentage (left), arm fat (right) percentage, leg fat (left) mass, leg fat (right) mass, leg fat (left) percentage, and leg fat (right) percentage (P < 0.05).

We also identified lipid-related metabolic markers that are risk factors for PA, including the free cholesterol-to-total lipids ratio in large LDL, medium LDL, and very large very LDL (P < 0.05). However, no significant correlations (P > 0.05) were found between PAs and glucose, GA, urea, UA, cholesterol, TG, HDL-C, and LDL-C (Figure 4, Supplementary Figure 1, and Supplementary Table 17).

Figure 4 Association between various influencing factors and the perianal abscess risk based on the results of the two-sample mendelian randomization. Data are expressed as the odds ratio (OR) with the 95%CI (error line). OR > 1.00 indicates that the exposure factor increases the perianal abscess risk. P < 0.05 was considered statistically significant. BMI: Body mass index; CHO: Cholesterol; GA: Glycated albumin; GLU: Glucose; HDL-C: High-density lipoprotein cholesterol; LDL-C: Low-density lipoprotein cholesterol; nSNP: Number of single nucleotide polymorphisms; OR: Odds ratio; TG: Triglycerides; UA: Uric acid; WHR: Waist-to-hip ratio.

Sensitivity analysis

To validate the robustness of our findings, we conducted several sensitivity analyses, indicating no significant evidence of pleiotropy. The weighted median method corroborated the influence of body fat-related and lipid metabolic markers, aligning with the findings of the main analysis. The weighted mode method showed no significant heterogeneity or bias. The leave-one-out method also showed robustness in the results. The results of these analyses collectively supported the credibility of our results (Supplementary Table 18).

DISCUSSION

This study utilized a case-control approach to identify risk factors associated with Pas, emphasizing on their number and location. Our analysis revealed significant correlations between the increased risk of PAs and higher BMI, WHR, WHtR, glucose levels, UA levels, cholesterol levels, TG levels, and LDL-C levels; additionally, lower HDL-C levels were also associated with an increased risk of PAs. The MR analysis further confirmed relationships between the PA risk and factors such as BMI, WHR, large and medium LDLs, and lipid ratios. These findings suggested that lipid levels, body fat distribution, and metabolic markers are crucial to the understanding of PA risk.

Obesity is a recognized risk factor for multiple chronic conditions, including cardiovascular disease, diabetes, and certain cancers[20-22]. It is also characterized by chronic low-grade inflammation[23]. In line with this, two studies have linked higher BMI to more severe disease phenotypes[24,25]. Regarding perianal disease, our previous work identified BMI as a potential risk factor for PA[18], and a cohort study similarly reported an increased risk of PA among individuals with obesity[4]. Obesity has also been associated with poorer outcomes after anal fistula surgery, including higher failure rates in complex cases[26]. Mechanistically, heightened inflammatory activity and a greater burden of metabolic dysfunction in obesity may plausibly increase susceptibility to PA[21,25,27,28]. Together, these observations support and contextualize our findings.

Our study identified novel associations and significant correlations between elevated WHRs and WHtRs and the incidence, number, and location of PAs. Allisha et al[29] identified that imbalances in the WHR and WHtR associated with morbid obesity are likely to cause complex PAs, such as horseshoe abscesses, thus complicating treatment. Similarly, Obst et al[30] reported that abnormal WHRs and WHtRs create difficulty in the healing of PA wounds, thus complicating early wound management. Previous studies primarily focused on BMI and overlooked central obesity markers such as the WHR and WHtR. Unlike BMI, the WHR and WHtR better reflect visceral fat accumulation[31] and may be more closely associated with PA development.

We found that the body fat percentage, body fat mass, and fat distribution (including limb fat content) were significantly correlated with the risk of PAs. Although no studies have directly linked these factors to PAs and other abscesses, research has shown associations with hypertension[32], cardiovascular disease[33], and erectile dysfunction[34]. These markers may be more valuable than the BMI, particularly for assessing risks related to glucose and lipid metabolism[35]. Therefore, maintaining a healthy body fat percentage with a normal BMI could enhance the effects of primary prevention strategies for PA by reducing central obesity and limb fat.

Lipid metabolism is crucial in the human body, and its disorder is linked to many diseases[36-38]. We found that HDL levels were negatively correlated with the risk of PAs but that LDL, cholesterol, and TG levels were positively correlated. Additionally, TG and LDL-C levels were associated with the increased number and severity of PAs. Previous studies have linked decreased HDL levels with more severe anal fistula disease, abscesses, and fistula areas[39]. Hypertriglyceridemia may indirectly affect the occurrence of PAs by impacting the immune function[40], and blood lipid markers such as TGs and LDL-C may be indirectly associated with inflammatory markers and metabolic diseases, thereby affecting the number and severity of PAs[21]. These results further highlight the role of lipid metabolism in PA pathogenesis and suggest that regulating blood lipid levels could help prevent PAs.

Our findings align with those of previous studies that found that hyperuricemia and hypertriglyceridemia are linked to an increased risk of inflammatory diseases[41-44]. UA, a product of purine metabolism, is associated with various diseases[45,46]. During this study, the positive correlation between the UA level and PA risk suggested its significant role in PA pathogenesis. This finding provides new insights into the role of UA in PA development.

Studies have found that patients with diabetes are at higher risk for PA infections, especially those caused by pathogens such as Klebsiella pneumonia; such infections are associated with compromised cell-mediated immunity, phagocyte function, and defective neutrophil activity caused by poor blood glucose control[13]. Increased blood glucose levels, which are reflected by higher glycated hemoglobin levels, increase the PA risk and complicate treatment[47], thus frequently leading to refractory wounds after anal fistula surgery[48]. Our study did not establish a direct link between increased blood glucose and UA levels and PA onset in the general population. However, observational studies have suggested that these metabolites are significant risk factors for PAs, especially in male patients. These findings further emphasize the critical role of metabolic disorders in PA onset and inform clinical secondary prevention strategies, particularly those for managing obesity, hyperlipidemia, hyperuricemia, and diabetes. Despite observational correlations, MR showed no significant causal effects of glucose or UA on PA. This discrepancy likely reflects confounding/reverse causation, weaker and less portable instruments, phenotype misalignment (lifelong genetic liability vs acute measurements), and possible non-linear or time-varying effects. Overall, glucose/UA appear more consistent with biomarkers/mediators than independent causes, whereas adiposity/Lipid traits show more coherent evidence under MR assumptions.

This study had some limitations. Triangulating a hospital-based case-control study with two-sample MR increases causal plausibility under MR assumptions but does not establish causation; observational and genetic results should be viewed as converging, not definitive. Our genetic instruments were largely European-ancestry whereas the clinical cohort is Chinese; cross-ancestry differences in linkage disequilibrium and allele frequencies may weaken instruments or introduce pleiotropy, and cross-ancestry MR (used when East Asian outcome GWAS were unavailable) should be considered exploratory. Additionally, the retrospective design and reliance on self-reported lifestyle variables limit the scope of conclusions. The potential for residual confounding, measurement error in metabolic variables, and misclassification of abscess diagnoses should be addressed. Finally, using hospital-based controls may introduce selection bias, as health-exam attendees can differ from the general population in care-seeking and socioeconomic/comorbidity profiles.

CONCLUSION

In this case-control and two-sample MR study, obesity indices (BMI, WHR, body-fat percentage/content, limb-fat distribution) and lipid traits (free cholesterol-to-total lipids ratios in large/medium LDL and very-large VLDL) were associated with higher odds of perianal abscess. The convergence of observational and genetic analyses increases causal plausibility under MR assumptions, but does not establish causation. These findings should be interpreted as consistent, hypothesis-generating evidence that can inform mechanistic studies and prospective, multi-center evaluations, and may help guide research on risk stratification and prevention.