Prevalence of uveitis in syphilis patients in Taiwan

ABSTRACT

Few population-based studies have looked at the risk of uveitis among syphilis patients. Our study addresses the knowledge gap by reporting on uveitis risk in syphilis patients through a retrospective cohort study. The Taiwan National Health Insurance database was used for this study, covering the period from January 1st, 2009, to December 31st, 2020. We created a 1:4 propensity score matched cohort between the syphilis patients and controls, which accounted for gender, age, and comorbidities. The primary endpoint was the incidence of newly recorded uveitis. The assessment of uveitis risk in syphilis patients included the use of the Kaplan-Meier method and multivariate Cox proportional hazard model. A total of 31,597 syphilis patients and 126,379 matched comparisons were recruited. The uveitis incidence rate from our syphilis patients was 1.25 per 1000 person-years. The uveitis incidence rate from our non-syphilis group was 0.8 per 1000 person-years. After matching, the syphilis group was found to have a higher risk of developing uveitis (adjusted hazard ratio (aHR) [95% CI]: 1.57 [1.36–1.81], P < .001). Among males and individuals aged 20–34 years, subgroup analysis showed an increased risk of uveitis in the presence of syphilis infection. The Kaplan-Meier survival curve showed a significant difference in uveitis incidence between syphilis and non-syphilis groups (log-rank test P < .001). In summary, our study revealed that Taiwanese syphilis patients were at a higher risk of developing uveitis. These results highlight the need for regular ocular monitoring and screening in individuals with syphilis.

KEYWORDS:

• Uveitis
• syphilis
• intraocular involvement
• prevalence
• Taiwan

Introduction

Syphilis is a systemic infection caused by the spirochete Treponema Pallidum. The incidence of syphilis has been on the rise in various countries, including the United Kingdom [1] and the United States [2]. In Taiwan, the syphilis incidence rate reached 60.6 per 10,000 person-years in 2016 [3]. However, uveitis due to syphilis infection is relatively uncommon and possibly contributes to the lack of studies to date on such associations. One estimate reported that the annual incidence of ocular syphilis is only 0.3 cases per one million adults in the UK [4]. Gaining a better understanding of the association between uveitis and syphilis is essential for preventing poor visual outcomes.

To address these knowledge gaps, we conducted a retrospective, population-based study to assess the risk of developing uveitis among syphilis patients in Taiwan.

Method

Data source

The NHI Research Database (NHIRD) is managed by the Taiwanese Food and Drug Administration (FDA) and includes comprehensive information on more than 99% of Taiwanese nationals. It encompasses data related to age, gender, diagnoses, medications, and medical procedures. A subset of the NHIRD is the Longitudinal Health Insurance Database (LHID), which contains data from one million patients randomly selected since 2000. Data from the LHID has been deidentified in compliance with the Taiwanese “Personal Information Protection Act”. Because of this anonymized database design, the requirement for informed consent has been waived. The present study was conducted in accordance with the Declaration of Helsinki.

Study design and data collection

This retrospective population-based cohort study utilized the Taiwanese LHID electronic health-registry database. Diagnosis in the registry was coded according to the International Classification of Diseases, Ninth and Tenth Revision, Clinical Modification (ICD-9-CM and ICD-10-CM). The study recruited both syphilis and non-syphilis cohorts in order to investigate the incidence of uveitis.

Regarding the inclusion criteria, patients who were 20 years of age or older with a new diagnosis of syphilis (ICD-9: 090–097 or ICD-10: A50-A53) were included in the syphilis study cohort (see Supplementary Table 1). To ensure diagnostic accuracy, participants from the syphilis study cohort had to have at least one outpatient or one inpatient encounter related to syphilis infection between January 1st, 2009, and December 31st, 2020. The index date was defined as the date when the first diagnosis of syphilis was given. Individuals diagnosed with syphilis before the study period were excluded.

The non-syphilis cohort comprised of participants who had never been diagnosed with syphilis within the study period. The index date for the non-syphilis comparison cohort was randomly generated. The follow-up period ended upon the onset of the new diagnosis of uveitis (ICD-9-CM: 360.00, 360.11, 360.12, 362.18, 363.00, 363.01, 363.03, 363.05, 363.08, 363.1x, 363.20, 363.21, 363.4x, 364.00, 364.02, 364.04, 364.1x, 364.3x or ICD-10-CM: H44.00x, H44.13x, H44.11x, H35.06x, H30.00x, H30.01x, H30.02x, H30.03x, H30.1x, H30.89x, H30.9x, H30.2x, H31.10x, H31.11x, H31.12x, H35.33, H20.00, H20.02x, H20.04x, H20.1x, H20.9), with the end date of the study period being December 31st, 2020.

To reduce confounding, we excluded patients that were: (1) below 20 years old or those with missing demographic information; (2) had a history of uveitis diagnosed before our study period; and (3) had underlying disease of viral hepatitis (ICD-9-CM code 070), human immunodeficiency virus infection (ICD-9-CM code 042–044, 795.8, V08), TB (ICD-9-CM 010-018), or autoimmune disease (ICD-9-CM code 135, 279.49, 283, 443, 571.42, 696, 710, 714, 715) (see Supplementary Table 1).

Propensity score matching was also applied to reduce bias. Through propensity score matching, patients from the syphilis cohort were matched to non-syphilis patients in a 1:4 ratio. The propensity score was determined using logistic regression modelling, which factors in probabilities from potential confounders, including age (within +/- 5 years), year of index date, and gender.

Comorbidities of interest for this study included: hypertension (ICD-9-CM 401–405; ICD-10-CM I10–I13, I15, N26.2), dyslipidemia (ICD-9-CM 272; ICD-10-CM E71.30, E75.21, E75.22, E75.24, E75.3, E75.5, E75.6, E77, E78.0, E78.1, E78.2, E78.3, E78.4, E78.5, E78.6, E78.70, E78.79, E78.8, E78.9), current cigarette smoking status (ICD-9-CM 305.1, V15.82; ICD-10-CM F17.2, Z87.891) and diabetes (ICD-9-CM code 250).

Outcome measures

The primary outcome of this study was the incidence of newly recorded uveitis. This was identified by the corresponding ICD-10 codes in the database. We only included patients who underwent at least two medical visits related to uveitis, with each visit separated by at least seven days. We followed both cohorts from their respective index date until a pre-defined endpoint. Such endpoint included new-onset uveitis, withdrawal from the programme for any reason, patient’s death, or the end of the study period on December 31st, 2020.

Statistical analysis

The baseline distributions of demographic characteristics and comorbidities between the syphilis and non-syphilis cohorts were compared using a two-sample t-test for continuous variables and the chi-squared test for categorical variables. 1:1 propensity-score matching was applied to minimize selection bias. This involved estimating the closest propensity scores between the syphilis and non-syphilis groups. We used the nearest-neighbor algorithm to construct matched pairs, with a p-value of 0.05 signifying a significant difference between the case and control groups. Our study also employed crude and multivariable-adjusted Cox proportional hazards models in order to determine the risk of uveitis from the covariates of genders, ages, and comorbidities of interest. Hazard ratios and their corresponding 95% confidence intervals (95% CIs) were presented for patients with syphilis compared to those without syphilis. The adjusted Kaplan–Meier method was utilized to estimate the cumulative risk of uveitis in patients with and without syphilis. The log-rank test was applied to examine for any differences. All data were analyzed using the SAS version 9.5 (SAS Institute, Cary, NC, USA) platform. A two-tailed p-value of 0.05 was considered significant.

Result

Baseline data

After propensity score matching, a total of 31,597 syphilis patients (77% male, mean age at index 43.2 years) and 126,379 patients without syphilis (77.8% male, mean age at index 43.3 years) were recruited from 2009 to 2020. The mean ages and genders in both groups were comparable (Table 1). The two groups were well-matched regarding the distribution of baseline demographics and comorbidities of interest. The mean follow-up time was 6.62 ± 3.32 (average years ± SD) for the syphilis group and 6.64 ± 3.3 for the non-syphilis group (see Table 1).

Table 1. Characteristics of individuals with and without Syphilis.

	Non-syphilis		Syphilis
	N = 126379		N = 31597
Variables	n	%	n	%	p-value
Sex					0.99
Female	28121	22.30	7032	22.30
Male	98258	77.80	24565	77.70
Age (year)					0.99
20–34	56663	44.80	14166	44.80
35–49	30452	24.10	7613	24.10
50–64	16966	13.40	4243	13.40
>=65	22298	17.60	5575	17.60
mean, (SD)	43.3	19.4	43.2	19.40	0.8
Comorbidity
Current smoker	7679	6.08	1922	6.08	0.96
Lipid metabolism disorders	33692	26.70	8423	26.70	0.99
Hypertension	38860	30.80	9715	30.80	0.99
DM	20347	16.10	5089	16.10	0.98
Mean follow-up for Uveitis, (SD)	6.64	3.30	6.62	3.32	0.35

Time-to-event analysis

During the study period, the incidence rate of uveitis among the syphilis cohort was 1.25 per 1000 person-years. The incidence rate of uveitis among the non-syphilis cohort was 0.8 per 1000 person-years. Using Cox regression analysis, the syphilis group was found to have a significantly increased risk of developing uveitis (adjusted hazard ratio (aHR) [95% CI]: 1.57 [1.36–1.81], P < .001) when compared to our control cohort (see Table 2).

Table 2. Incidences and hazard ratios of Uveitis for individuals with and without Syphilis.

	Non-Syphilis			Syphilis			Crude			Adjusted
Variables	Events	PY	IR*	Events	PY	IR*	cHR	(95% CI)	p-value	aHR ^†	(95% CI)	p-value
Overall	672	839225	0.8	262	209203	1.25	1.57	(1.36, 1.81)	<0.001	1.57	(1.36, 1.81)	<0.001
Sex
Woman	193	190989	1.01	57	47692	1.20	1.19	(0.88, 1.59)	0.260	1.18	(0.88, 1.59)	0.260
Man	479	648236	0.74	205	161511	1.27	1.72	(1.46, 2.02)	<0.001	1.28	(1.47, 2.04)	<0.001
Age
20–34	168	375509	0.45	110	93540	1.18	2.63	(2.07, 3.34)	<0.001	2.64	(2.07, 3.36)	<0.001
35–49	170	222175	0.77	56	55993	1.00	1.31	(0.97, 1.77)	0.080	1.31	(0.97, 1.78)	0.080
50–64	138	123493	1.12	48	31396	1.53	1.37	(0.99, 1.90)	0.060	1.37	(0.98, 1.90)	0.060
>=65	196	118048	1.66	48	28274	1.70	1.02	(0.75, 1.40)	0.890	1.01	(0.73, 1.38)	0.970
Comorbidity
Current smoker	39	55747	0.70	7	13922	0.50	0.72	(0.32, 1.61)	0.420	0.73	(0.32, 1.62)	0.430
Lipid metabolism disorders	266	253412	1.05	69	63574	1.09	1.04	(0.80, 1.35)	0.790	1.04	(0.79, 1.35)	0.800
Hypertension	308	252312	1.22	76	64178	1.18	0.97	(0.76, 1.25)	0.820	0.97	(0.75, 1.24)	0.790
DM	177	128908	1.37	50	32994	1.52	1.11	(0.81, 1.51)	0.530	1.10	(0.80, 1.50)	0.570

Abbreviation: DM, diabetes; IR, Incidence rate; PY, Person Years; cHR, crude hazard ratio; aHR, adjusted hazard ratio.

* Incidence rate per 1000 person-years.

† adjusted hazard ratio, adjusted for age, gender, smoker status, lipid metabolism disorder, hypertension, diabetes.

Kaplan–Meier survival statistics revealed the cumulative incidence of uveitis increased in the syphilis cohort compared to the non-syphilis cohort (log-rank test P < .001) (see Figure 1).

Figure 1. Kaplan-Meier curve for cumulative incidence of uveitis.

Other possible clinical factors related to uveitis

When comparing the syphilis and non-syphilis cohort and stratifying by gender and age, our results revealed increased risks for developing uveitis in male patients with syphilis (aHR:1.28 [95% CI: 1.47 - 2.04]). Conversely, no increased risk was found among the female syphilis patients (aHR: 1.18 [CI: 0.88- 1.59]) (P < 0.001) (Table 2). With regards to age, a significant increase in uveitis was found in the 20–34 age group within the syphilis cohort (aHR: 2.64 [CI: 2.07 - 3.36]) compared to the non-syphilis cohort.

When stratified by the three medical comorbidities of interest, it was observed that syphilis patients did not demonstrate a higher risk of developing uveitis compared to matched controls, irrespective of whether they had any of these comorbidities (adjusted hazard ratio for hyperlipidemia: 1.04 [0.79–135], p = 0.8; for hypertension: 0.97 [0.75–1.24], p = 0.79; and for diabetes: 1.1 [0.8–1.5], p = 0.57) (see Table 2).

Smoking status was also found to be non-contributory to developing uveitis risk among syphilis patients (aHR: 0.73 [0.32–1.62], p = 0.43).

In addition to syphilis, our cox-regression analysis revealed other clinical factors that might influence the occurrence of uveitis (see Table 3 and Supplementary Figure 1).

Table 3. Cox model-measured hazard ratios and 95% confidence intervals of Uveitis associated with Syphilis and covariates or not.

Characteristics	Event			Crude			Adjusted
	n	PY	IR	HR	(95%CI)	p-value	HR	(95%CI)	p-value
Syphilis
No	672	839225	0.80	1	(reference)	-	1	(reference)	-
Yes	262	209203	1.25	1.57	(1.36, 1.81)	<0.001	1.57	(1.36, 1.81)	<0.001
Sex
Woman	250	238681	1.05	1	(reference)	-	1	(reference)	-
Man	684	809747	0.84	0.81	(0.70, 0.93)	0.004	1.04	(0.89, 1.21)	0.64
Age, years
20–34	278	469048	0.59	1	(reference)	-	1	(reference)	-
35–49	226	278168	0.81	1.37	(1.15.1.64)	<0.001	1.38	(1.15, 1.65)	<0.001
50–64	186	154889	1.20	2.03	(1.68, 2.44)	<0.001	1.99	(1.61, 2.46)	<0.001
>=65	244	146323	1.67	2.81	(2.36, 3.33)	<0.001	2.72	(2.19, 3.38)	<0.001
Comorbidity
Current smoke
No	888	978760	0.91	1	(reference)	-	1	(reference)	-
Yes	46	69669	0.66	0.73	(0.54, 0.98)	0.04	0.76	(0.56, 1.03)	0.07
Lipid metabolism disorders
No	599	731442	0.82	1	(reference)	-	1	(reference)	-
Yes	335	316986	1.06	1.30	(1.13, 1.48)	<0.001	0.95	(0.81, 1.11)	0.5
Hypertension
No	550	731939	0.75	1	(reference)	-	1	(reference)	-
Yes	384	316489	1.21	1.62	(1.42, 1.84)	<0.001	0.95	(0.80, 1.13)	0.57
DM
No	707	886526	0.80	1	(reference)	-	1	(reference)	-
Yes	227	161902	1.40	1.76	(1.51, 2.04)	<0.001	1.27	(1.06, 1.51)	0.009

Abbreviations: HR, hazard ratio; CI, confidence interval.

Adjusted HR: Adjusted for age, sex, and comorbidities in the Cox proportional hazards model.

In terms of gender, no increased risk for uveitis was found among male patients when compared to female patients. However, among patients aged 35 years old or older, a higher risk of uveitis was seen when compared to their younger counterparts (adjusted HR for age 35–49: 1.38 [1.15–1.65];50–64: 1.99 [1.61–2.46]; > = 65: 2.72 [2.19–3.38]). Additionally, participants with diabetes were at higher risk of developing uveitis than those without diabetes (adjusted HR for diabetes: 1.27 [1.06–1.51]). The comorbidities of smokers (aHR: 0.76 [CI: 0.56–1.03])(p = 0.07), lipid metabolism disorders (aHR: 0.95 [CI: 0.81–1.11])(p = 0.5) and hypertension (aHR: 0.95 [CI: 0.8–1.13]) (p = 0.57) were found to have no significant impact on uveitis risk when compared to their counterparts (Table 3 and Supplementary Figure 1).

In terms of human immunodeficiency virus (HIV) infection status, an increased risk was seen among syphilis patients with concomitant HIV-positive status (HR 1.77 [1.22–2.56], p = 0.003) compared to syphilis patients that were HIV-negative (see Supplementary Table 2).

Discussion

To the best of our knowledge, this study represents the largest investigation to date that has aimed to assess uveitis risk among patients with syphilis.

Novel findings

Our study showed that syphilis patients have a higher risk of developing uveitis. This increased risk for uveitis persisted even after adjusting for age, gender, and specific baseline comorbidities. Among male patients and those aged 20–34 years, there was an increased risk of uveitis occurrence. Furthermore, comorbidities such as diabetes, hypertension, hyperlipidemia, and smoker status were found to not contribute significantly to the occurrence of uveitis among syphilis patients. Our subgroup analysis on HIV status further revealed that syphilis patients with concomitant HIV infection exhibited a higher risk of uveitis compared to syphilis patients without HIV infection.

Clinical implications

Few studies have looked at the risk of uveitis among syphilis patients, despite previous public health concerns raised by the American Center of Disease Control (CDC) [5]. Consequently, our findings hold significance in addressing these knowledge gaps. Furthermore, syphilis uveitis has been known to be easily misdiagnosed. Such misdiagnosis has been previously shown to lead to poor visual outcomes [6,7]. Our data aims to raise awareness about the risk of uveitis among syphilis patients and to support clinicians in achieving the best clinical outcome for these individuals.

Comparison to literature

We also found some relevant studies in the literature. Albini et al. reported that the cumulative incidence of syphilis patients diagnosed with uveitis each year from 2002 to 2009 was around 1.29% [8]. They also found that the odds ratio of uveitis occurrence among syphilis patients was lower for women (Odds Ratio [OR]: 0.4, CI 0.28–0.57). Although we used different statistical analyses, we found notable similarities with Albini et al. Firstly, our incidence rate for uveitis among syphilis was 1.25 per 1000 person-years. Furthermore, our results also indicated that males are at increased risk for uveitis among syphilis patients [9]. This predilection towards the male gender was further supported by another population-based study as well [10]. Differences in chromosomes and sex hormone profiles between genders may offer explanations for the increased risk of uveitis among syphilitic patients who are male [11].

In another study by Oliver et al., they recruited 4,232 syphilis patients and showed that 63 (equivalent to 1.5% of all their syphilis cases) were diagnosed with ocular syphilis. They also found that ocular syphilis was more prevalent among males (prevalence ratio (PR): 2.82; 95% CI: 1.03–7.73), individuals aged 40 years of age or older (PR: 3.27; 95% CI: 1.98–5.39), and those of white ethnicity (PR: 2.63; 95% CI: 1.60–4.32). Our results showed some similarities with these findings, but there were also some minor differences. Specifically, we found those aged 20–34 were at higher risk of uveitis among syphilis patients. We propose the following reasons for these divergent findings. Firstly, we hypothesize that HIV status is a confounding factor in the study conducted by Oliver et al. Notably, 56% of the syphilis patients from Oliver et al. had confirmed HIV-positive status. This has implications for the results from Oliver et al., as older HIV patients are known to be at an increased risk of uveitis [12,13]. Secondly, the difference in healthcare systems between Taiwan and the United States could also be a contributing factor. The existence of universal healthcare in Taiwan ensures that Taiwanese citizens of all ages have equal access to healthcare. In contrast, the United States has a combination of public and private healthcare. This greater reliance on private health care in the United States can affect the results obtained. Studies have indicated that individuals under 40 in the United States are at higher risk of being uninsured [14]. Therefore, this could lead to the underdiagnosis of uveitis in this younger age group. This partially explains the differences between the results from Oliver et al. and ours. However, our results also highlighted the influence of age on uveitis risk when the factor of syphilis infection was removed. In Table 3, we demonstrated a significantly increased risk of uveitis among the age groups of 35–49 (aHR: 1.38 [CI: 1.15–1.65]), 50–64 (aHR: 1.99 [CI: 1.61–2.46]) and those aged 65 or older (aHR:2.72 [CI: 2.19–3.38]) compared to individuals in the age group younger than 34 years (p < 0.001). These findings align with existing literature, which has indicated the elderly population’s greater susceptibility to developing ocular inflammatory conditions [15].

Another point of discussion pertains to HIV status. While specific case reports have documented posterior and pan-uveitis presentation among syphilis patients with concurrent HIV-positive status [16,17], such an association remains relatively unexplored in studies with a larger sample size. To address this knowledge gap, we conducted a separate subgroup analysis that investigated the impact that a positive HIV status would have on the uveitis risk among syphilis patients. Our analysis revealed an elevated risk of uveitis among syphilis patients with HIV-positive status in comparison to those without HIV infection. Such results align with the case reports mentioned earlier. However, caution is warranted when interpreting our findings, given the following reasons. Firstly, it is important to note that antiviral drugs commonly prescribed for HIV patients have been linked to drug-induced uveitis. Examples of such antiviral drugs associated with uveitis include Rifabutin and Cidofovir [18]. Another essential aspect to emphasize was that our ICD-code-based registry does not facilitate the examination of laboratory results, such as CD4+ T cell count. HIV infection has been known to target and destroy CD4+ T-lymphocyte cells. The resulting reduction in CD4 count from HIV infection plays a pivotal role in the clinical staging of HIV [19]. Existing studies have indicated that a lower CD4 + count was associated with an increased risk of ocular manifestations in patients with HIV [20]. Unfortunately, we lack information on how the CD4⁺ T cell count from our enrolled HIV-positive status patients might have influenced the uveitis risk obtained. One final point worth noting from our subgroup analysis was the observation of a lower number of syphilis patients with a positive HIV status compared to those without HIV. The observed lower number of HIV-affected syphilis patients from our study was interesting, especially considering the shared mode of sexual transmission and common risk factors known to exist between HIV and syphilis infection [21]. We propose that the relatively low number of HIV-positive syphilis patients from our study might reflect cultural stigma. The negative stigma linked to HIV status, particularly in Asian countries like Taiwan [22], may have contributed to the underestimation of HIV prevalence among such populations. Furthermore, a study has shown that individuals who experience self-stigmatization from one health condition may be more prone to perceiving stigma from a second medical condition [23]. In other words, the additional self-stigmatization stemming from patients with syphilis and HIV coinfection might exacerbate the already perceived stigma due to syphilis. The perceived stigma may influence individuals’ decisions regarding seeking care for HIV or syphilis-related issues, potentially contributing to the underestimation of incidences of syphilis patients with concomitant HIV infection.

Our study also revealed no association between smoking status, various medical comorbidities (such as diabetes, hyperlipidemia, and hypertension) and the risk of uveitis among syphilis patients. These factors were analyzed because they represent some of adult patients’ most frequently encountered comorbidities. Furthermore, it has been theorized that these factors might predispose such patients to developing intraocular inflammation such as uveitis, particularly after syphilis infection. While there have been some reported associations between some of these variables investigated in our study and the risk of uveitis [24,25], our study may be one of the first to report a lack of associations between these specific comorbidities and the risk of uveitis among syphilis patients.

Interestingly, other studies have shown an increased risk of uveitis with the comorbidities of hypertension and hyperlipidemia [26]. This heightened risk is thought to be due to inflammation. Conditions like hypertension and atherosclerosis upregulate systemic inflammatory markers like interleukin [27]−23. Diabetes has also been linked to uveitis [26], primarily due to the blood-retinal barrier disruption by glycated haemoglobin A1c (HbA1c). In our study, we did observe an increased risk of uveitis among diabetic patients when removing syphilis infection as a factor (adjusted HR for diabetes: 1.27 [1.06–1.51]) compared to their non-diabetic counterparts. However, it remains unclear how syphilis affects the relationship between diabetes and uveitis. Smoking is another known risk factor for uveitis, and it exerts this effect by exacerbating inflammation [24]. However, our study found no increased risk for uveitis among smokers with syphilis infection. We suggest several reasons for these findings. One primary factor could be confounding variables related to medication history. It is known that medications like metformin and statins can reduce the occurrence of uveitis due to their anti-inflammatory properties. Another factor could be the clinical severity of the recruited comorbidities. It is possible that the comorbidities that we recruited were of milder clinical grade. For example, patients with well-managed diabetes with tighter glycemic control have been shown to have reduced risk for uveitis.

Strengths and limitations

Our study has several strengths. It is one of the first large-scale retrospective studies that examined uveitis risk in syphilis patients. Additionally, our study benefits from using a population-based electronic health registry, employing propensity score matching, and conducting long follow-ups. These strengths enhance the reliability of our findings.

Our study has some limitations that warrant discussion. Firstly, we acknowledge that not all potential confounding factors were accounted for within our analysis. An example is medication history and the lack of analysis of laboratory results. Another limitation is our absence of information regarding the clinical stages of syphilis in the patients we studied and the specific subtypes of uveitis involved. The natural history of syphilis can be broadly classified into primary, secondary, and tertiary stages (see supplementary Figure 2) based on their respective classical presentations. Additionally, there is a subset of syphilis infection called neurosyphilis. Neurosyphilis occurs when syphilis infection has reached the central nervous system. Ocular involvement can manifest at any stage of syphilis, but previous research indicated a propensity for ocular syphilis to occur during secondary and tertiary syphilis [28]. Furthermore, some studies have shown a close association between ocular syphilis and neurosyphilis [29]. This is another limitation as it is unknown which type of syphilis we may have recruited more of.

Conclusion

In summary, our study assessed uveitis risk among syphilis patients in Taiwan. Our findings indicate an elevated uveitis risk among syphilis patients, with an increased risk observed among males and those aged 20–34. These results emphasize the importance of raising awareness about syphilis-associated uveitis among healthcare professionals and individuals at-risk.

Ethics statement

The study adhered to the tenets of the Declaration of Helsinki. As the Taiwan National Health Insurance database is comprised of de-identified data, the requirement of written consent from study participants was deemed unnecessary and waived by the Ethics Research Committee. Institutional Review Board approval was also not required because de-identified data were used for this retrospective analysis.

Author contributions

CYC, HTK, and AYH wrote and edited the paper's main text.

CYC, HTK, AYH, CHL, and NYH participated in the study design and drafting of the paper.

CYC, HTK, AYH, CHL, and NYH designed, discussed, edited, and guided the overall process of the paper.

CYC, HTK, AYH, CJL, NYH, PTT, CTL, HSC, and YYT contributed to the drafting of the paper.

CYC, HTK, AYH, CJL, NYH, PTT, CTL, HSC, and YYT contributed to the study's design.

CYC, HTK, AYH, CJL, NYH, PTT, CTL, HSC, and YYT provided valuable insight and contributed to the drafting of the paper.

All authors (CYC, HTK, AYH, CJL, NYH, PTT, CTL, HSC, YYT) contributed equally to the manuscript revisions.

All authors (CYC, HTK, AYH, CJL, NYH, PTT, CTL, HSC, YYT) approved the final version of the manuscript and agreed to be held accountable for the content therein.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data generated during and analyzed in this article are available from the corresponding author under reasonable request.