Incidence of Chronic Hepatitis B Virus Infection Among People With HIV in Yunnan Province, China: A Retrospective Observational Cohort Study (2004–2024)

Abstract

Background

Hepatitis B virus (HBV) exacerbates mortality and morbidity in individuals with human immunodeficiency virus (HIV), but limited research has been conducted on incidence of HBV in people with HIV (PWH).

Methods

We herein conducted a retrospective cohort study of PWH in Yunnan province, China. PWH who underwent antiretroviral therapy (ART) and whose status was initially negative for hepatitis B surface antigen (HBsAg) were enrolled and followed-up with from 2004 to 2024 to analyze chronic HBV infection incidence and associated factors.

Results

A total of 40 086 HBsAg-negative PWH underwent at least 1 follow-up test, contributing to a cumulative observation time of 221 049.8 person-years (PYs). Of the included individuals, 894 chronic HBV infections were identified, corresponding to a HBsAg-emia incidence of 2.23%, resulting in an overall incidence rate of 0.40 per 100 PYs (95% CI: 0.38–0.43). We ascertained that individuals who acquired their infection as an injectable drug user (adjusted hazard ratio [aHR, 1.50]) and who exhibited low ART adherence (aHR, 1.46) were primarily associated with a higher incidence of chronic HBV infection. Those individuals who were enrolled post-2014 (aHR, 0.81) were linked to a lower incidence of HBV, indicating a 19% relative reduction in chronic HBV infection compared to those enrolled pre-2014.

Conclusions

The reduction may have been linked to factors such as introduction of hepatitis B vaccination, a decrease in proportion of HIV infections transmitted through injectable drugs, and improved effectiveness of ART. We posit that our efforts should focus in the future on enhancing immunization, treatment, and behavior intervention in this population.

Hepatitis B virus (HBV) infection is a major public health concern worldwide [1]. According to the Hepatitis Report 2024 of the World Health Organization (WHO), an estimated 254 million people were living with HBV in 2022, with 1.1 million deaths attributed to the infection [2]. With highly effective preventive measures and treatments now available, the WHO established an ambitious goal in 2016 to eliminate HBV as a public health issue by 2030 [3, 4]. Specifically, this goal aims to achieve, by 2030, a 95% reduction in new infections and a 65% reduction in mortality compared to the 2015 baseline. Additionally, it aims to reduce the prevalence of hepatitis B surface antigen (HBsAg) in children under 5 years of age to below 0.1% by 2030.

Hepatitis B is primarily transmitted through sexual contact, blood contact, and from mother to child [1, 5]. Vertical (mother-to-child) transmission and horizontal transmission during early childhood are the primary routes of HBV transmission, leading to the majority of chronic infections [6]. Chronic infection occurs in 90% of infected neonates and infants but in less than 5% of individuals who contract the infection in adulthood [1, 6]. Vaccination with the hepatitis B vaccine (HepB) is the most effective way to prevent HBV infection. In China, HepB has been included in the national immunization program since 2002, and the 3-dose series for newborns is provided free of charge. The schedule recommends the first dose within 24 hours of birth, followed by the second and third doses at 1 month and 6 months of age, respectively [7]. As a result, the prevalence of HBsAg in children aged 1–4 fell from 9.67% in 1992% to 0.3% in 2020 [8], which was very close to meeting the WHO's indicator requirements for the elimination of hepatitis B.

People with HIV (PWH) represent a key population for the targeted control of HBV. They are more likely to progress to chronic infection than the general population upon exposure to HBV. Those with established coinfection experience accelerated liver disease progression and higher rates of liver-related complications. They also exhibit reduced immunogenicity and protective efficacy following hepatitis B vaccination [1, 9, 10]. Even with effective virologic suppression via antiretroviral therapy (ART), PWH may experience persistent immune dysregulation and chronic inflammation, contributing to non-AIDS-related morbidity [9]. Furthermore, due to shared transmission routes and often overlapping high-risk behaviors, the prevalence of chronic HBV coinfection among PWH is substantial, estimated at approximately 10% [11].

Given their heightened risks and complex care needs, PWH constitute a key population for focused efforts toward HBV elimination. Over 1.355 million people were reported to be living with HIV/AIDS by 2024 in China [12]. Consequently, understanding the incidence of new HBV infections and related factors within this group is crucial for formulating targeted interventional strategies to reduce new infections and contribute to the broader goal of hepatitis B elimination. In China, initial ART regimens often included lamivudine (3TC)-based combinations. Since 2015, tenofovir disoproxil fumarate (TDF) + 3TC has become the preferred first-line regimen, providing dual activity against both HIV and HBV. Since the introduction of free ART in China in 2004, treatment coverage has increased substantially, reaching 92.8% among diagnosed PWH by the end of 2022 [13]. In this study, we used the HIV antiviral-treatment follow-up cohort data from Yunnan Province, one of China's regions most profoundly impacted by HIV/AIDS, and analyzed the incidence of chronic HBV infection and associated factors among PWH. Whether ART regimens conferred a preventive benefit against incident HBV infection would also be explored. We posit that our research findings will provide a basis for reducing new infections in the population and that they will ultimately contribute to a diminution in the overall number of new infections.

METHODS

All individuals in China who test positive for HIV are reported to China's HIV/AIDS Comprehensive Response Information Management System that integrates the epidemiological data on HIV/AIDS as well as data on ART (detailed descriptions of the system's data information and operational mechanisms can be found elsewhere [14, 15]). Free ART was launched in China in 2004, and the China Free Antiretroviral Therapy Manual suggests that all PWH undergo a series of clinical and laboratory evaluations (including HBsAg testing) before initiating ART [16]. People with HIV on ART were recommended to undergo regular follow-ups every 3–6 months. Liver function tests were routinely performed during follow-up. For those who were HBsAg-negative at baseline, additional HBsAg testing was performed if there was clinical suspicion of hepatitis, elevated liver enzymes, or reported high-risk behavior during the follow-up. The testing was performed using either chemiluminescence immunoassay or enzyme-linked immunosorbent assay.

We herein conducted a retrospective analysis of data from Yunnan province up to 2024 and downloaded it on 1 January 2025. Because HBV core antibody (anti-HBc) testing was not routinely performed in this programmatic setting, we utilized the HBsAg serological marker to assess the incidence of chronic HBV infection among PWH. Therefore, our analysis identifies new onset of HBsAg positivity, which is a marker of chronic infection, but cannot identify resolved infections or acute infections that subsequently clear HBsAg.

The exclusion criteria comprised individuals who had not undergone HBsAg testing and who were HBsAg-positive at baseline; individuals who were HBsAg-negative at baseline were also included in the cohort. The first HBsAg test served as the baseline test, while subsequent tests were regarded as follow-up tests. Statistical comparison of demographic characteristics between the follow-up and nonfollow-up groups was performed using the χ² test. Initial HBsAg positivity was considered indicative of an incident infection during the follow-up period, with the incident infection time defined as the period between the time of the first positive test and the time of the last negative test. We calculated follow-up time from the baseline test until either the onset of an incident infection or the last negative HBsAg test; HBsAg-negative patients at the last visit were right-censored.

The incidence of chronic HBV infection was stratified based on various baseline characteristics of the study participants—including sex, age, educational status, marital status, occupation, enrollment year, route of HIV infection, baseline CD4 cell count, type of ART clinic attended, ART regime used, ART adherence, and HIV viral load. Enrollment year was dichotomized at 2014, coinciding with the regional rollout of HBV vaccination for PWH and growing proportion of PWH receiving ART containing TDF. ART regimens were classified into 3 categories: 3TC-containing, 3TC and TDF-containing, and others. Alternative regimens could be introduced in cases of drug toxicity or resistance. Low adherence to antiviral therapy was defined as missed appointments or discontinuation of ART during treatment for any ART associated with higher incidence. We subsequently evaluated HIV viral load 3–6 months after initiating ART. We employed Kaplan–Meier analysis to compute survival probabilities for the remaining individuals who were HBsAg-negative, categorized by their year of enrollment. Furthermore, we exploited univariate and multivariate Cox proportional hazards models to compare the incident rates of chronic HBV infections among different subgroups and to estimate the adjusted hazard ratios (aHRs).

Patient Consent Statement

Our study involved a retrospective observational cohort analysis aimed at assessing the incidence of chronic HBV infection among individuals with HIV in Yunnan, China. This research received approval from the Scientific Research Ethics Committee of the Yunnan Center for Disease Control and Prevention (Ethics Approval: 2024–31). The committee granted a waiver for informed consent for this study.

RESULTS

As of 31 December 2024, the HIV/AIDS Comprehensive Response Information Management System recorded a total of 134 888 HIV-positive individuals who had undergone at least 1 HBsAg test, resulting in a cumulative total of 436 415 test records. Of these individuals, 9639 tested positive for HBsAg at baseline, resulting in a positivity rate of 7.1%; and 125 249 individuals who subsequently tested negative at baseline were enrolled in the HBV observation cohort. Of these, 40 086 individuals had at least 1 follow-up HBsAg test, and 85 163 experienced no follow-up tests for a proportion of 68.0% (Figure 1).

Figure 1.

Study profile.

The individuals who underwent follow-up HBV tests ranged in age from 0 to 96 years, with a median age of 38 years (interquartile range, 30–47 years). Among the 40 086 individuals with HIV, 61.1% were male; 74.4% were aged between 23 and 49 years; 82.2% had completed secondary education or less; 52.6% were married; 63.3% worked as farmers; 50.9% were enrolled in care after 2014; and 78.6% reported transmission through heterosexual intercourse. At baseline, 34.8% had CD4 cell counts between 200 and 349 cells per microliter; and the majority of participants (78.7%) underwent treatment at county or township ART clinics, with 54.3% using 3TC-containing regimen and 44.6% using a combination of 3TC and TDF-containing regimen. Additionally, 77.9% of individuals achieved an HIV viral load of less than 200 copies/mL within 3–6 months after initiating ART; however, 27.0% of individuals experienced low ART adherence. Table 1 presents the demographic and HIV-related profiles of the patients, categorized by follow-up status. The data show differences in all characteristics between the 2 groups.

Table 1.

Baseline Characteristics of the Participants

Characteristics	Population	With HBV Follow-up Test	Without HBV Follow-up Test	P Value
Total	125 249	40 086	85 163
Sex	…	…	…	<.001
Male	79 463	24 495(61.1)	54 968(64.5)
Female	45 786	15 591(38.9)	30 195(35.5)
Age (y)	…	…	…	<.001
0–22	6548	2357(5.9)	4191(4.9)
23–49	86 003	29 825(74.4)	56 178(66.0)
≥50	32 698	7904(19.7)	24 794(29.1)
Education	…	…	…	<.001
Secondary school and below	104 425	32 964(82.2)	71 461(83.9)
Postsecondary school	20 268	6889(17.2)	13 379(15.7)
Missing data	556	233(0.6)	323(0.4)
Marital status	…	…	…	<.001
Unmarried	33 542	11 173(27.9)	22 369(26.3)
Married	64 533	21 084(52.6)	43 449(51.0)
Divorce/Widow	26 835	7698(19.2)	19 137(22.5)
Missing data	339	131(0.3)	208(0.2)
Occupation	…	…	…	<.001
Farmer	81 578	25 370(63.3)	56 208(66.0)
Unemployed	18 788	6047(15.1)	12 741(15.0)
Others/unknown	24 883	8669(21.6)	16 214(19.0)
Route of HIV infection	…	…	…	<.001
Heterosexual intercourse	101 388	31 495(78.6)	69 893(82.1)
MSM	6408	2065(5.2)	4343(5.1)
Injecting drug use	14 725	5408(13.5)	9317(10.9)
Others/unknown	2728	1118(2.8)	1610(1.9)
Baseline CD4 cell count (per μL)	…	…	…	<.001
<200	42 924	12 770(31.9)	30 154(35.4)
200–349	41 559	13 954(34.8)	27 605(32.4)
350–499	22 373	7365(18.4)	15 008(17.6)
≥500	18 163	5997(15.0)	12 166(14.3)
Unknown	230	0(0)	230(.3)
Enrollment y	…	…	…	<.001
2004–2013	49 922	19 669(49.1)	30 253(35.5)
2014–2024	75 327	20 417(50.9)	54 910(64.5)
Type of ART clinic	…	…	…	<.001
County/Township	97 020	31 534(78.7)	65 486(76.9)
Provincial/municipal	27 454	8239(20.6)	19 215(22.6)
Unknown	775	313(0.8)	462(0.5)
ART regimen	…	…	…	<.001
3TC-containing	59 603	21 757(54.3)	37 846(44.4)
3TC and TDF-containing	63 436	17 897(44.6)	45 539(53.5)
Others	2210	432(1.1)	1778(2.1)
HIV viral load (copies/mL) a	…	…	…	<.001
<200	90 490	31 214(77.9)	59 276(69.6)
≥200	27 277	8517(21.2)	18 760(22.0)
Unknown	7482	355(0.9)	7127(8.4)
ART adherence	…	…	…	<.001
High	95 664	29 262(73.0)	66 402(78.0)
Low	29 585	10 824(27.0)	18 761(22.0)

^a: The HIV viral load was evaluated 3–6 m after initiating ART.

During the study period, a total of 221 049.8 person-years (PYs) were observed for these 40 086 individuals who had undergone HBsAg follow-up tests, with a median of 4.9 years (interquartile range, 2.3–8.5 years). Of these individuals, 894 chronic HBV infections were identified, with a HBsAg-emia incidence of 2.23%, resulting in an overall incidence rate of 0.40 per 100 PYs (95% confidence interval [CI], .38–.43).

Univariate analysis revealed higher incidence rates for PWH who were infected through injectable drug use, those with an HIV viral load ≥200 copies/mL or without HIV viral load data, and those with low ART adherence compared to their respective reference subgroups. Conversely, females, individuals with postsecondary school education, those enrolled post-2014, those engaged in men who have sex with men, and those receiving treatment at provincial and municipal-level ART clinics showed a relatively lower incidence (Table 2). The cumulative probability of HBV infection demonstrated a significant decline in the post-2014 period compared to the pre-2014 period (Figure 2). No significant difference in incidence rate was found across different first-line ART regimens containing HBV-active drugs, such as lamivudine and tenofovir.

Table 2.

Risk Factors for Incident HBV Infection Among People With HIV

Characteristics	Total	Person-Y	New HBsAg (%)	Incidence Rate Per 100 Person-y (95% CI)	Univariate Analyses		Multivariate Analyses
					Hazard ratio (95%CI)	P value	Adjusted hazard ratio (95% CI)	P value
Total	40 086	221 049.8	894(2.23)	.40(.38–.43)	-	-	-	-
Sex							…
Male	24 495	129 895.9	607(2.48)	.47(.43–.50)	1		1
Female	15 591	91 153.9	287(1.84)	.31(.28–.35)	.71(.62–.82)	<.001	.74(.63–.86)	<.001
Age (y)							…
0–22	2357	14 150.7	44(1.87)	.31(.22–.40)	1		1
23–49	29 825	171 781.3	713(2.39)	.42(.38–.45)	1.31(.97–1.78)	.081	.92(.65–1.31)	.648
≥50	7904	35 117.8	137(1.73)	.39(.32–.46)	1.06(.76–1.49)	.728	.81(.55–1.19)	.285
Education status							…
Secondary and below	32 964	179 004.2	761(2.31)	.43(.39–.46)	1		1
Postsecondary	6889	40 589.3	127(1.84)	.31(.26–.37)	.76(.63–.92)	.004	.98(.79–1.21)	.863
Missing data	233	1456.3	6(2.58)	.41(.08–.74)	1.11(.50–2.48)	.796	.93(.34–2.50)	.88
Marital status							…
Unmarried	11 173	60 631.4	223(2.00)	.37(.32–.42)	1		1
Married	21 084	119 701.1	487(2.31)	.41(.37–.44)	1.15(.98–1.35)	.083	1.31(1.10–1.57)	.003
Divorce/Widow	7698	39 919.6	180(2.34)	.45(.39–.52)	1.21(.99–1.47)	.059	1.39(1.13–1.72)	.002
Missing data	131	797.7	4(3.05)	.50(.01–.99)	1.53(.57–4.11)	.399	1.21(.37–3.93)	.755
Occupation							…
Farmer	25 370	133 770.7	572(2.25)	.43(.39–.46)	1		1
Unemployed	6047	35 521.8	145(2.40)	.41(.34–.47)	1.01(.84–1.21)	.928	1.01(.83–1.23)	.935
Others/unknown	8669	51 757.3	177(2.04)	.34(.29–.39)	.86(.73–1.02)	.077	1.02(.84–1.23)	.869
Heterosexual intercourse	31 495	172 256.1	640(2.03)	.37(.34–.40)	1		1
MSM	2065	10 400.6	21(1.02)	.20(.12–.29)	.50(.32–.77)	.002	.64(.40–1.02)	.063
Injecting drug use	5408	30 937.6	210(3.88)	.68(.59–.77)	1.92(1.64–2.24)	<.001	1.50(1.25–1.81)	<.001
Others/unknown	1118	7455.5	23(2.06)	.31(.18–.43)	.97(.64–1.47)	.871	.99(.61–1.61)	.968
Baseline CD4 (cells/μL)							…
<200	12 770	70 850.6	279(2.18)	.39(.35–.44)	1		1
200–349	13 954	80 650.7	341(2.44)	.42(.38–.47)	1.09(.93–1.28)	.288	1.08(.92–1.26)	.363
350–499	7365	39 211.7	155(2.10)	.40(.33–.46)	.94(.77–1.14)	.531	.99(.81–1.21)	.901
≥500	5997	30 336.8	119(1.98)	.39(.32–.46)	.90(.73–1.12)	.347	.96(.77–1.19)	.69
Enrollment y							…
2004–2013	19 669	134 033.7	543(2.76)	.41(.37–.44)	1		1
2014–2024	20 417	87 016.0	351(1.72)	.40(.36–.45)	.70(.61–.80)	<.001	.81(.69–.95)	.008
Type of ART clinic							…
County/Township	31 534	167 378.2	752(2.38)	.45(.42–.48)	1		1
Provincial/municipal	8239	52 637.9	135(1.64)	.26(.21–.30)	.62(.52–.75)	<.001	.70(.57–.85)	<.001
Unknown	313	1033.7	7(2.24)	.68(.18–1.18)	1.19(.56–2.50)	.651	1.26(.60–2.66)	.547
ART regime							…
3TC-containing	21 757	138 336.6	513(2.36)	.37(.34–.40)	1		1
3TC and TDF-containing	17 897	80 794.5	372(2.08)	.46(.41–.51)	.97(.85–1.11)	.659	1.14(.98–1.32)	.092
Others	432	1918.7	9(2.08)	.47(.16–.77)	1.06(.55–2.05)	.861	1.30(.67–2.52)	.435
HIV viral load (copies/mL) a							…
<200	31 214	174 270.7	668(2.14)	.38(.35–.41)	1		1
≥200	8517	45 963.0	217(2.55)	.47(.41–.53)	1.27(1.09–1.48)	.003	1.10(.94–1.29)	.238
Unknown	355	816.1	9(2.54)	1.10(.39–1.82)	2.06(1.07–3.98)	.031	1.44(.74–2.79)	.287
ART adherence							…
High	29 262	159 793.1	554(1.89)	.35(.32–.38)	1		…
Low	10 824	61 256.7	340(3.14)	.56(.50–.61)	1.64(1.43–1.87)	<.001	1.46(1.26–1.68)	<.001

^a: The HIV viral load was evaluated 3–6 m after initiating ART.

Figure 2.

Time to HBV infection by enrollment year.

Multivariate analysis showed that being married (adjusted hazard ratio [aHR], 1.31; 95% CI, 1.10–1.57), divorced or widowed (aHR 1.39; 95% CI, 1.13–1.72), having acquired infection through injectable drug use (aHR 1.50, 95% CI, 1.25–1.81), and having low ART adherence (aHR 1.46; 95% CI, 1.26–1.68) were associated with a higher incidence of chronic HBV infection. In contrast, being female (aHR 0.74; 95% CI, .63–.86), enrollment post-2014 (aHR, 0.81; 95% CI, .69–.95), and receiving treatment at provincial and municipal-level ART clinics (aHR, 0.70; 95% CI, .57–.85) were associated with a lower incidence of chronic HBV infection. The aHR of 0.81 for individuals enrolled post-2014 compared to that of those enrolled pre-2014 indicated a 19% relative reduction in chronic HBV infection.

DISCUSSION

We herein conducted a retrospective cohort analysis of chronic HBV infection among PWH in Yunnan Province and found an incidence rate of 0.4 per 100 PYs; and we also observed that higher rates of chronic HBV infections were primarily associated with injectable drug users and PWH showing poor adherence to ART. Nevertheless, the incidence of chronic HBV infection showed a declining trend, attributable to the use of HepB vaccine and positive behavioral changes.

The incidence of chronic HBV infection in our cohort among PWH with ART was 0.40 per 100 PYs. It is important to note methodological differences when comparing this rate to other studies. For instance, a study from Rakai, Uganda, which utilized HBV core antibody testing to capture all new infections (including resolving ones), reported an incidence of 0.49 per 100 PYs [17]. HBsAg seroclearance rates among HIV/HBV coinfected patients on ART ranged from 7.2% to 10.9% across studies in China and the United States [18–20]. Given the higher rate of progression to chronicity in PWH, one might expect the chronic infection incidence in a comparable setting to be a substantial proportion of the total exposure incidence. In contrast, the reported incidence from Madrid, Spain (0.02 per 100 PYs), which also reflects total infection, was considerably lower [21], highlighting geographical disparities in HBV burden. According to the Global Hepatitis Report 2024, the majority of global HBV morbidity and mortality disproportionately afflicts African and western Pacific regions [2]. China, with 79.7 million individuals affected by HBV, is one of the countries showing the greatest burden of hepatitis B [2], and the incidence of PWH in Yunnan and Uganda was therefore much higher than that in Spain. However, the estimated number of new HBV infections in China was only 20 914 in 2022, which was equivalent to 0.0015 per 100 PYs in the general population [2]. The high incidence of chronic HBV infection in PWH in our analysis indicated that individuals with HIV constituted a high-risk population for HBV infection. The persistent incidence of chronic HBV infection suggests ongoing risk exposures in this population. While dedicated behavioral data were not collected prospectively in this cohort, a survey in Yunnan reported that high-risk sexual behavior, though declining after HIV diagnosis, remained measurable [22].

Some studies have suggested a stronger protective effect of HBV-active ART medications, particularly for TDF-based regimens [17, 23–25]. We did not observe a statistically significant difference in HBV incidence between patients on a 3TC + TDF-containing regimen and those on a 3TC-containing regimen without TDF (aHR 1.14, 95% CI 0.98–1.32, P = .092). In this cohort, most participants received regimens containing 3TC (54.3%) or both 3TC and TDF (44.6%). The lack of a significant difference in our study may be explained by several factors. First, the inherent antiviral activity of 3TC against HBV, present in both comparison groups, may provide a baseline level of protection, making an additional effect of TDF harder to detect. Second, the critical factor of treatment adherence appears to be a major modulator of this potential benefit. As our results show, low ART adherence was independently associated with a 46% higher risk of HBV infection (aHR 1.46), suggesting that the prophylactic potential of any ART regimen is contingent upon consistent intake. This underscores that while the choice of regimen is important, ensuring high adherence is paramount for achieving both optimal HIV suppression and maximal potential protection against incident HBV. Based above analysis, ART is not a reliable substitute for vaccination due to the realities of imperfect adherence and persistent risk exposure. HBV vaccination must be viewed as a foundational, nonsubstitutable component of preventive care for PWH, complementary to—and not replaced by—effective ART.

The aHR of 0.81 for individuals enrolled post-2014 compared to that of those enrolled pre-2014 indicated a 19% relative reduction in HBV infection. We attribute this to the use of HepB vaccine and changes in the behavior of PWH. Since 2002, China has implemented a free universal HBV vaccination program for newborns and a catch-up program for children under 15, including those with HIV. Moreover, free HepB vaccine was offered for PWH and with all 5 HepB infection biomarkers negative upon their voluntary participation in high HIV-prevalence areas in Yunnan Province in 2014, with the support of a national major science and technology project [26]. Follow-up data revealed that 5 years postvaccination, the HepB surface antibody positivity rate was 21.2% among vaccinated individuals with HIV, which was significantly higher than that among the unvaccinated PWH (4.2%) [26]. In line with broader patterns, a meta-analysis indicates that anti-HBs positivity in adult PWH falls below 50% after 2 years and to just 17% after 5 years [10]. While our study did not assess individual antibody levels, the expansion of vaccination coverage at the population level likely played a role in reducing chronic infections after 2014. Additionally, shifts in the epidemiological profile of HIV and associated risk behaviors may have contributed. In Yunnan's HIV epidemic, the proportion of infections attributed to injecting drug use has decreased over time, while sexual transmission has become predominant [27, 28]. As injecting drug use is a more efficient route for HBV transmission, this shift may represent 1 contextual factor associated with the observed decline in incidence post-2014. A study conducted in Yunnan also indicated that sustained behavioral interventions had reduced high-risk behaviors over time [22]. This may help to clarify the declining trend in incident HBV chronic infections observed in our research. The chief target groups for HepB vaccination are newborns currently, followed by infants, children under 15 who are not immune, and high-risk adults such as those with HIV [29, 30]. In Yunnan, only some regions administered free HBV vaccination to individuals with HIV in 2014 with their informed consent [26]. Our study confirmed that the incidence of chronic HBV infections among individuals with HIV remained relatively high, however, highlighting the need to strengthen HBV vaccination efforts.

In our study, we found that being married or divorced/widowed and having acquired the infection through injectable drug use were associated with a higher incidence of chronic HBV infection; while being female, enrolled post-2014, and having undergone treatment at provincial and municipal-level ART clinics were associated with a lower incidence of chronic HBV infection. The disparities in marriage status and sex in our results were generally consistent with the findings of a meta-analysis conducted by Dong in Yuxi City, Yunnan Province, which revealed that within 1 year of HIV diagnosis, the risky behavior among married individuals was 2.4 times higher than among unmarried individuals [22]. This result could have contributed to the higher incidence of chronic HBV infections among married individuals with HIV in Yunnan Province. Another study suggested that the greater prevalence of chronic HBV among men may be linked to the role of sex hormones in regulating the immune response and responses to external pathogens. Some authors observed that immune clearance of HBsAg occurred in a higher proportion of women than men and that women tended to gain better protection from HBV vaccination than men [31]. These data might explain why men had a relatively higher HBV incidence, but it does not rule out the possibility that men engaged in higher-risk behaviors than women with respect to HBV infection.

We acknowledge several limitations in this analysis. First, our study relied solely on HBsAg testing without concurrent testing for anti-HBc. We could not identify individuals who experienced an acute HBV infection that subsequently resolved (with HBsAg clearance and anti-HBc development). HBsAg levels can fluctuate (eg, during the window period or in occult infection), and detection depends on the performance and threshold of the immunoassay used. Therefore, our reported chronic HBV incidence based on HBsAg should be considered a conservative estimate from public health surveillance and may slightly underestimate the true infection rate. Second, due to the criminalization of drug use and social stigma faced by men who have sex with men, transmission routes may have been misclassified. Third, the analysis was restricted to individuals with at least 2 HBsAg tests, potentially introducing selection bias. Irregular follow-up intervals may also have influenced the findings. Fourth, the characteristics of PWH in Yunnan may not be generalizable to other populations in China or globally. Further research across broader and more diverse populations is warranted.

In conclusion, by conducting this cohort study, we observed a decline in chronic HBV infection incidence and identified several risk factors that were associated with incident chronic HBV infections among PWH. However, the incidence of chronic HBV infections remained high, highlighting the need to strengthen HBV screening, vaccination, and behavioral interventions in the population. This study also provides valuable insights for other similar geographical regions that attempt to control hepatitis B, thus minimizing the impacts of HBV on PWH.