The prevalence of anxiety disorders among people living with HIV in India: A systematic review and meta-analysis

ABSTRACT

Background:

Globally, 301 million face anxiety disorders, notably common among people living with human immunodeficiency virus (PLHIV), causing major challenges and reduced quality of life.

Aim:

This review assesses India’s prevalence, offering evidence to guide targeted interventions.

Methods:

A systematic literature search was conducted across PubMed, Excerpta Medica database (EMBASE), Scopus, and Web of Science up to November 2024. Observational studies using validated tools to assess anxiety prevalence in PLHIV were included. The pooled prevalence of anxiety disorders was estimated using a random-effects meta-analysis. Subgroup and sensitivity analyses were conducted to explore sources of heterogeneity. Study quality was assessed using the Joanna Briggs Institute Critical Appraisal Tool. Between-study heterogeneity was evaluated using Cochran’s Q and I² statistics.

Results:

Fifteen studies (5,336 PLHIV) were included. The pooled prevalence of anxiety was 36% (95% confidence interval: 24%–47%), with higher prevalence in hospital-based studies (39%) compared to community-based studies (23%). Females had a higher prevalence (46%) than males (36%), though this difference was not statistically significant. Diagnostic tools influenced prevalence estimates, with self-report measures (e.g., Hospital Anxiety and Depression Scale [HADS] and Depression Anxiety and Stress Scale [DASS]) reporting higher prevalence (48%–49%) than structured interviews (ICD-10: 7%, DSM-IV: 1%). Subgroup analysis revealed significant heterogeneity across study settings and diagnostic tools.

Conclusion:

This meta-analysis demonstrates a high prevalence of anxiety disorders among PLHIV in India, underscoring the need for integrated mental health services within HIV care frameworks. Routine anxiety screening, stigma reduction initiatives, and gender-sensitive interventions are essential to address this mental health burden effectively. Future research should focus on longitudinal outcomes and community-based approaches to improve mental health care for PLHIV.

INTRODUCTION

Human immunodeficiency virus (HIV) remains a significant global public health challenge, having caused an estimated 42.3 million deaths to date. By the end of 2023, approximately 39.9 million people were living with HIV, with 0.6% of adults aged 15–49 globally affected.[1] While the burden of HIV varies significantly across regions, India, with an estimated 2.4 million people living with the condition, is among the countries most affected.[2] Although there is no cure for HIV, advancements in prevention, diagnosis, and treatment, particularly through antiretroviral therapy (ART), have transformed the disease into a manageable chronic condition.[3]

Despite the increased life expectancy following the introduction of ART, living with HIV continues to pose significant challenges.[4] Living with HIV can significantly impact mental health, increasing the risk of mood, anxiety, and cognitive disorders, with depression being particularly common.[5] Anxiety disorders, as defined by the DSM-5, are characterized by excessive fear and anxiety that result in significant impairment in daily functioning. Unlike normal developmental or stress-related anxiety, these disorders persist and often lead to serious disruptions in quality of life.[6,7] In India, societal stigma surrounding HIV, coupled with cultural taboos about mental health, creates a dual burden for people living with HIV (PLHIV).[8] Limited access to mental health services, particularly in rural areas where healthcare resources are scarce, further compounds these challenges, as mental health often remains deprioritized in HIV care programs.[9]

The connection between HIV and anxiety disorders is both reciprocal and dynamic. HIV-related stressors, such as stigma and treatment challenges, can exacerbate anxiety, while heightened anxiety can impair adherence to ART and worsen health outcomes. Quantifying the burden of anxiety disorders in PLHIV is essential to inform policy decisions and design targeted interventions. Reliable prevalence estimates can help integrate mental health into HIV care frameworks, allocate resources effectively, and develop culturally sensitive programs addressing the unique needs of PLHIV in India. Despite extensive research in Western contexts, studies on the mental health impact of HIV in India remain limited. This systematic review and meta-analysis aim to bridge this gap by providing evidence on the prevalence and comorbidity of anxiety disorders among PLHIV in India.

METHODS AND DESIGN

Study protocol and design

This systematic review and meta-analysis adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines and aimed to determine the prevalence of anxiety among adults living with HIV in India.[10] The study protocol was registered with the International Registration of Systematic Reviews under the registration number CRD42024626815. A comprehensive search strategy was developed to identify primary studies reporting the prevalence of anxiety among PLHIV, regardless of the specific screening tools used.

Inclusion and exclusion criteria

The inclusion criteria encompassed observational studies, including cross-sectional, case-control, and cohort studies, that reported anxiety prevalence using validated screening tools among adults aged 18 years and above living with HIV in India. Studies were restricted to those published in English up to November 31, 2024. Exclusion criteria included studies without full-text availability, qualitative studies, systematic reviews, case series, letters, conference abstracts, and studies lacking validated tools for anxiety assessment [Supplementary Table S1].

Supplementary Table S1.

Inclusion and exclusion criteria

	Inclusion criteria	Exclusion criteria
Population	Diagnosed case of HIV (Irrespective of treatment status)
Outcome	Anxiety complying with diagnostic criteria for anxiety disorders based on the DSM or the ICD	No valid diagnostic measure of anxiety disorders or a symptomatologic measure of anxiety disorders
Timeline	Initiation to November 2024
Language	English	Published in other languages
Type of study	Cross-sectional, baseline data from prospective studies; providing the proof of HIV infection and status	Reviews, case series, case study

Information sources and search strategy

We conducted electronic searches across four databases, namely Excerpta Medica database (EMBASE), PubMed, Scopus, and Web of Science, with the search period extending up to and including November 31, 2024. A thorough search was performed using keywords and Medical Subject Headings (MeSH) related to anxiety, including terms such as anxiety, anxiety disorder, or mental health. Articles were identified by combining terms and utilizing Boolean operators [Supplementary Table S2]. Subsequently, we refined the results to include only the most pertinent ones. The search process was double-blinded and carried out collaboratively by two authors (AM and PH). To ensure comprehensive coverage, we conducted reference checking, hand searches of citations, and scrutinized the reference lists of included studies identified during the search.

Supplementary Table S2.

Search strategy

Pubmed:

#1 (“Anxiety”[Mesh]) OR “Anxiety Disorders”[Mesh] OR Anxiety [Title/abstract] OR “Anxiety disorder” [Title/Abstract] -372,048

#2 “HIV”[MeSH Terms] OR “Acquired Immunodeficiency Syndrome”[MeSH Terms] OR “HIV”[Title/Abstract] OR “Human Immunodeficiency Virus”[Title/Abstract] OR “AIDS”[Title/Abstract] OR “PLHIV”[Title/Abstract] - 499,192

#3 “india”[MeSH Terms] OR “india”[All Fields] OR “india’s”[All Fields] OR “indias”[All Fields] -872,769

#1 AND #2 AND #3 = 109

Web of science

#1 TS=(“Anxiety” OR “Anxiety Disorders” OR “Anxiety disorder”)= 288,018

#2 TS=(“HIV” OR “Acquired Immunodeficiency Syndrome” OR “Human Immunodeficiency Virus” OR “AIDS” OR “PLHIV”)= 503,941

#3 TS=(“India” OR “india s” OR “indias”)= 239,064

#4 #1 AND #2 AND #3= 38

Embase

#1 ‘anxiety’/exp OR ‘anxiety disorder’/exp OR ‘anxiety’:ti,ab OR ‘anxiety disorder’:ti,ab = 716031

#2 ‘hiv’/exp OR ‘acquired immunodeficiency syndrome’/exp OR ‘human immunodeficiency virus’:ti,ab OR ‘aids’:ti,ab OR ‘plhiv’:ti,ab = 895985

#3 ‘india’/exp OR ‘india’:ti,ab OR ‘india s’:ti,ab OR ‘indias’:ti,ab = 277495

#4 #1 AND #2 AND #3 = 182

Scopus:

#1 TITLE-ABS-KEY ( “Anxiety” OR “Anxiety disorder” ) = 595,262

#2 TITLE-ABS-KEY ( “HIV” OR “Acquired Immunodeficiency Syndrome” OR “Human Immunodeficiency Virus” OR “AIDS” OR “PLHIV” ) = 776084

#3 TITLE-ABS-KEY ( “India” OR “india s” OR “indias” ) = 647462

#4 #1 AND #2 AND #3 = 112

Study selection

All citations obtained from the electronic searches were uploaded to the Rayyan software, and duplicate entries were systematically eliminated. Subsequently, two independent researchers (AM and PH) conducted a comprehensive screening of titles and abstracts from the retrieved studies to identify articles eligible for potential inclusion and any disagreement were resolved by discussion and agreement. Any remaining disagreements were consulted with the third reviewer (TA) to assess the inclusion of studies for the next step. The identified articles underwent a thorough review of the full text by the same independent authors (AM and PH), adhering to a pre-defined eligibility criterion to determine relevance for inclusion in the review. In instances where additional information was needed to address queries about eligibility, collaboration with the remaining authors was sought. Any disagreements were resolved through discussion. Furthermore, the reasons for excluding articles were meticulously documented at each stage, and any remaining uncertainties or disagreements were addressed by a third reviewer (TA).

Data extraction

Key details, including the first author’s name, sample size, study regions, design, publication date, utilized assessment scales, anxiety, prevalence, ART Treatment status, presence of comorbidity, constituted the parameters for our data extraction template. A standardized Microsoft Excel spreadsheet served as the data extraction form, ensuring consistency in gathering pertinent information from eligible articles. The authors (AM, PH) independently performed the data extraction from the included articles, and any discrepancies were resolved through discussion and mutual agreement among the authors. In instances where data were insufficient, missing, or the full text was unavailable, the corresponding authors of the original articles were contacted via email to acquire the necessary information.

Statistical analysis

The pooled prevalence of anxiety was calculated using a random-effects model to account for between-study variability. Heterogeneity was assessed using Cochran’s Q statistic and the I² statistic. Subgroup analyses were performed to explore sources of heterogeneity based on gender, study settings and methods used to assess anxiety. Sensitivity analysis was conducted using Baujat plots, leave-one-out analyses, and influence diagnostics to identify studies contributing disproportionately to heterogeneity and to examine the robustness of the pooled estimates. To assess publication bias, funnel plots were visually inspected, and Egger’s regression test was performed to detect small-study effects. Trim-and-fill analysis was applied to adjust for potential publication bias by imputing missing studies and recalculating the pooled prevalence. Meta-regression analyses were conducted to identify potential covariates influencing the prevalence of anxiety, including sample size and the prevalence of depression. The proportion of heterogeneity explained by these covariates was assessed using R², and the relationships were visually represented using bubble plots. Statistical analyses were performed using STATA-18 software (StataCorp LLC, College Station, Texas, United States), with significance set at P < 0.05.

Assessment quality and the risk of bias

Two independent researchers (AM and PH) evaluated the methodological quality and risk of bias in the included studies using the nine item Joanna Briggs Institute Critical Appraisal tools specifically tailored for prevalence studies.[11] Studies that scored 1 to 4 were categorized as poor, 5 to 6 as fair and 7 to 9 as good quality. A higher score indicates a lower risk of bias, while a lower score indicates a higher risk of bias.

RESULTS

A systematic search across four electronic databases yielded 441 records (PubMed = 109, EMBASE = 38, Scopus = 112, Web of Science = 182). After the removal of 221 duplicate records, 220 records were screened based on titles and abstracts, leading to the exclusion of 198 records that did not meet the inclusion criteria. Full-text retrieval was attempted for 22 reports, of which two reports could not be retrieved. The remaining 20 reports were assessed for eligibility, and five reports were excluded due to wrong study population (n = 3), wrong study setting (n = 1), or absence of the outcome variable (n = 1). A total of 15 studies met the eligibility criteria and were included in the final systematic review and meta-analysis [Figure 1].

Figure 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analysis flow diagram

A total of 15 studies were included in this meta-analysis, conducted across various states in India, including Maharashtra, Karnataka, Tamil Nadu, West Bengal, Haryana, Manipur, and Punjab. Of these, 11 studies were hospital-based, while 4 studies were community-based. The sample sizes ranged from 30 to 2,924 participants, with a total of 5,336 participants across all studies. Anxiety was assessed using different tools, with the Hospital Anxiety and Depression Scale (HADS) being the most used, followed by the General Anxiety Disorder-7 (GAD-7) scale, the Depression Anxiety and Stress Scale (DASS), and structured clinical interviews based on ICD-10 and DSM-IV. The proportion of individuals identified as suffering from anxiety ranged widely across studies, from 1% participant to 69.8% [Table 1].

Table 1.

Characteristics of the included studies

Author (last name, et al)	Study site	Study setting (hospital based/community based)	Tools to assess anxiety	Sample size	Number of people suffering from anxiety
Agrawal et al. (2012)[12]	Maharashtra	Hospital based study	Hospital anxiety and depression scale	50	27
Chandra et al. (1998)[13]	Karnataka	Hospital based study	Hospital anxiety and depression scale	51	18
Chandra et al. (2003)[14]	Karnataka	Hospital based study	Hospital anxiety and depression scale	68	17
Chauhan et al. (2013)[15]	Maharashtra	Hospital based study	Clinical interview according to ICD-10	100	7
Ekstrand et al. (2022)[16]	Karnataka	Community based study	General Anxiety Disorder-7	467	77
Ghose et al. (2015)[17]	West Bengal	Hospital based study	Hospital anxiety and depression scale	100	44
Jain et al. (2023)[18]	Haryana	Hospital based study	Hospital anxiety and depression scale	500	298
Joshi et al. (2014)[19]	India	Hospital based study	Hospital anxiety and depression scale	2924	2042
Marbaniang et al. (2020)[20]	Maharashtra	Community based study	General Anxiety Disorder-7	167	41
Naoroibam et al. (2016)[21]	Manipur	Hospital based study	Hospital anxiety and depression scale	44	20
Nebhinani et al. (2011)[22]	Punjab	Hospital based study	Structured Clinical Interview for DSM-IV Clinical Version	100	1
Parsecepe et al. (2024)[23]	Maharashtra	Community based study	General Anxiety Disorder-7	200	7
Peter et al. (2013)[24]	Karnataka	Hospital based study	Hospital anxiety and depression scale	226	117
Sahayam et al. (2022)[25]	Tamil Nadu		Depression Anxiety and Stress Scale	30	16
Stanley et al. (2014)[26]	Tamil Nadu	Community based study	Depression Anxiety and Stress Scale	309	148

In the quality assessment of the study, apart from Sahayam et al. (moderate), all the studies were of good quality. We did not find any studies of poor quality in this meta-analysis [Supplementary Table S3].[25]

Supplementary Table S3.

Quality assessment of the included studies using the JBI critical appraisal tool for prevalence studies

Authors	Q1	Q2	Q3	Q4	Q5	Q6	Q7	Q8	Q9	Score
Marbaniang et al	Yes	Yes	No	Yes	Yes	Yes	Yes	Yes	No	7/9
Nebhinani et al	Yes	No	No	Yes	Yes	Yes	Yes	Yes	Yes	7/9
Parsecepe et al	Yes	No	Yes	No	Yes	Yes	Yes	Yes	Yes	7/9
Ekstrand et al	Yes	Yes	Yes	No	Yes	Yes	Yes	Yes	Yes	8/9
Naoroibam et al	Yes	Yes	No	Yes	Yes	Yes	Yes	Yes	Yes	8/9
Chauhan VS et al	Yes	No	No	Yes	Yes	Yes	Yes	Yes	Yes	7/9
Peter E et al.	Yes	No	Yes	Yes	Yes	Yes	Yes	Yes	Yes	8/9
Agarwal M et al.	Yes	No	No	Yes	Yes	Yes	Yes	Yes	Yes	7/9
Chandra PS et al.	Yes	No	No	Yes	Yes	Yes	Yes	Yes	Yes	7/9
Chandra PS et al.	Yes	No	No	Yes	Yes	Yes	Yes	Yes	Yes	7/9
Jain Deepak et al.	Yes	No	No	Yes	Yes	Yes	Yes	Yes	Yes	7/9
Stanley et al.	Yes	No	No	Yes	Yes	Yes	Yes	Yes	Yes	7/9
Ghose T et al.	Yes	No	No	Yes	Yes	Yes	Yes	Yes	Yes	7/9
Joshi B et al.	Yes	No	Yes	Yes	Yes	Yes	Yes	Yes	Yes	8/9
Sahayam et al	Yes	No	No	No	Yes	Yes	Yes	Yes	Yes	6/9

Q1: Was the sample frame appropriate to address the target population? Q2: Were study participants sampled in an appropriate way? Q3: Was the sample size adequate? Q4: Were the study subjects and the setting described in detail? Q5: Was the data analysis conducted with sufficient coverage of the identified sample? Q6: Were valid methods used for the identification of the condition? Q7: Was the condition measured in a standard, reliable way for all participants? Q8: Was there appropriate statistical analysis? Q9: Was the response rate adequate, and if not, was the low response rate managed appropriately?

A total of 15 studies involving 5,336 PLHIV, of whom 2,880 were identified as suffering from anxiety, were included in this meta-analysis. Using a random-effects model, the overall pooled prevalence of anxiety among PLHIV was 36% (95% confidence interval [CI]: 24%–47%). Substantial heterogeneity was observed across the studies (I² = 99.11%, τ² = 0.05, P < 0.01), indicating significant between-study variability. Cochran’s Q test further supported the presence of heterogeneity (Q = 3865.77, P < 0.01) [Figure 2].

Figure 2.

Forest plot showing the overall pooled prevalence of anxiety among people living with HIV in India. HIV = Human immunodeficiency virus

Subgroup analysis

To explore potential sources of heterogeneity, subgroup analyses were conducted based on study setting and the assessment tools used for diagnosing anxiety. Subgroup analysis by study setting revealed a pooled prevalence of 23% (95% CI: 5%–41%) for community-based studies and 39% (95% CI: 25%–53%) for hospital-based studies. Despite the higher prevalence in hospital-based studies, the difference between the two subgroups was not statistically significant (Q = 1.87, P = 0.17). Heterogeneity remained substantial in both subgroups [Figure 3]. Subgroup analysis by assessment tools demonstrated significant differences in the reported prevalence of anxiety. The pooled prevalence was 48% (95% CI: 43%–54%) for studies using the DASS, 15% (95% CI: 3%–27%) for studies using the GAD-7 scale, and 49% (95% CI: 39%–59%) for studies using the HADS. In contrast, studies using the Structured Clinical Interview for ICD-10 reported a lower prevalence of 7% (95% CI: 2%–12%), and those using the Structured Clinical Interview for DSM-IV reported a prevalence of 1% (95% CI: −1%–3%) [Supplementary Figure S1 ^{(886.4KB, tif)} ]. The difference between subgroups was statistically significant (Q = 334.98, P < 0.01). The subgroup analysis stratified by gender revealed a pooled prevalence of anxiety among females of 46% (95% CI: 28%–63%) with substantial heterogeneity (I² = 92.62%), while for males, the pooled prevalence was 36% (95% CI: 12%–60%) with higher heterogeneity (I² = 97.55%).A test for group differences between males and females showed no statistically significant difference (Q = 0.41, P = 0.52) [Supplementary Figure S2 ^{(551KB, tif)} ].

Figure 3.

Forest plot showing subgroup analysis of the pooled prevalence of anxiety among people living with HIV, stratified by study setting (community-based vs. hospital-based). HIV = Human immunodeficiency virus

Sensitivity analysis

To assess the robustness of the findings and identify influential studies, a series of sensitivity analyses were conducted. The Baujat plot revealed that “Joshi et al. (2014)” and “Ekstrand et al. (2022)” contributed disproportionately to both the pooled prevalence and overall heterogeneity [Figure 4a]. Further influence analysis using Cook’s distance, (Difference in BETAs) DFBETAS, and Q contributions confirmed these two studies as having significant impact [Figure 4b]. A leave-one-out analysis, where each study was systematically excluded, demonstrated that the pooled prevalence remained stable (ranging from 35% to 58%) across all omissions, indicating the robustness of the results. However, omitting “Ekstrand et al. (2022)” slightly reduced heterogeneity (I²) from 99.1% to 97%, suggesting its notable contribution to between-study variability [Figure 4c and d]. Overall, these analyses confirm that while certain studies influence heterogeneity, the pooled prevalence estimate remains reliable and robust.

Figure 4.

Sensitivity analysis and influence diagnostics for the pooled prevalence of anxiety among people living with HIV. (a) Baujat plot showing the contribution of individual studies to overall heterogeneity and influence on the pooled result. (b) Influence diagnostic plots, including standardized residuals, Cook’s distance, Q contribution, and DFBETAS for individual studies. (c) Leave-one-out analysis (sorted by proportion) showing the impact of omitting each study on the pooled prevalence. (d) Leave-one-out analysis (sorted by I²) showing changes in heterogeneity after omitting individual studies. HIV = Human immunodeficiency virus

Publication bias

Publication bias was assessed using funnel plot symmetry, Egger’s regression test, and trim-and-fill analysis. The funnel plot appeared symmetric, suggesting no major evidence of small-study effects. Egger’s test confirmed this, yielding a non-significant result (β = 2.91, P = 0.19). Furthermore, the trim-and-fill analysis did not impute any additional studies, and the pooled prevalence remained unchanged at 36% (95% CI: 24%–47%), supporting the robustness of the findings [Supplementary Figure S3 ^{(178.9KB, tif)} ]. It should be noted, however, that publication bias assessments using funnel plots and Egger’s regression test have recognized limitations when applied to prevalence meta-analyses. In this context, funnel plot asymmetry is more likely to reflect genuine between-study heterogeneity in design, setting, and measurement tool than selective non-publication of studies, and inferential conclusions drawn from these tests should therefore be interpreted with caution.

Meta-regression analysis

To further investigate sources of heterogeneity, meta-regression analyses were performed with sample size and depression prevalence as covariates. Importantly, the prevalence of depression was significantly associated with anxiety prevalence (β = 0.00577, P = 0.006), explaining 46.05% of the between-study variability. For every 1% increase in depression prevalence, the pooled prevalence of anxiety increased by approximately 0.58%. Sample size demonstrated a positive but non-significant association with anxiety prevalence (β = 0.000134, P = 0.071), explaining 15.20% of the observed heterogeneity. Despite this, residual heterogeneity remained high (I² = 94.27%), suggesting that additional unmeasured factors may contribute to variability across studies [Supplementary Figures S4 ^{(180.8KB, tif)} and S5 ^{(178.9KB, tif)} ].

DISCUSSION

The meta-analysis revealed a pooled prevalence of anxiety among PLHIV in India of 36% (95% CI: 24%–47%), significantly higher than the 3.3%–3.5% prevalence reported for neurotic and stress-related disorders in the general population by the National Mental Health Survey (2015–2016) and the Global Burden of Disease Study (2017).[27,28] This highlights a substantial burden of anxiety within this vulnerable population. Similar trends were observed in Brandt et al.’s review, which reported a median prevalence of 22.85%.[4] The higher prevalence in our study may reflect unique sociocultural challenges and systemic barriers to mental health care in India, including stigma, limited access to services, and cultural taboos.

Anxiety and HIV share a bidirectional relationship. HIV-related stressors, such as stigma, poor treatment adherence, and physical health deterioration, can exacerbate anxiety. In turn, anxiety contributes to poorer quality of life, impaired adherence to ART, and worse health outcomes. Mechanistically, anxiety disrupts the hypothalamic-pituitary-adrenal (HPA) and sympathetic-adreno-medullary (SAM) axes, leading to elevated cortisol and norepinephrine levels that impair immune function and enhance HIV viral replication. These physiological disruptions underscore the critical need for addressing anxiety as part of comprehensive HIV care.[29] Most anxiety in PLHIV manifests as PTSD or illness anxiety following diagnosis. Predictive factors include high anxiety sensitivity, poor distress tolerance, and avoidant coping strategies.[4] Although existing studies highlight the association between anxiety and HIV, the exact cause-effect relationship remains unclear due to the cross-sectional nature of most research.[30,31]

Subgroup analysis in this study revealed significant differences in anxiety prevalence based on the choice of assessment tool. Studies utilizing self-report scales, such as DASS and HADS, reported pooled prevalence rates as high as 48% and 49%, respectively, compared to much lower rates reported by structured clinical interviews, such as ICD-10 (7%) and DSM-IV (1%). Clinicians have a range of assessment techniques at their disposal, from semi-structured diagnostic interviews to self- or clinician-administered rating scales based on established criteria.[32] Rating scales offer a time-efficient and easy method of measurement that requires minimal specialized training, with many being zero to low cost, while psychiatric interviews allow for a detailed evaluation of the disorder.[33] Self-report scales, while sensitive to anxiety symptoms, often capture general distress or subclinical symptoms, leading to inflated prevalence estimates.[34,35,36] In contrast, structured interviews, with their rigorous diagnostic criteria, tend to provide more conservative but clinically precise estimates. However, their practical challenges, including implementation difficulties and potential lack of patient acceptability, often limit their use in resource-constrained settings.[32]

Global literature also supports these findings, with studies reporting higher rates of anxiety disorder diagnoses in research using questionnaire-based assessments compared to diagnostic interviews. For example, O’Cleirigh et al. and Brandt et al. noted similar trends, emphasizing the trade-offs between the practicality of self-report tools and the diagnostic precision of structured interviews.[4,29] Furthermore, screening tools such as GAD-7 and DASS-21 demonstrate high sensitivity but lower specificity, often capturing anxiety symptoms that may overlap with other psychiatric or medical conditions. In populations with a high prevalence of HIV, such as Zimbabwe, even validated tools like GAD-7 showed sensitivity of 89% but specificity of only 73% against structured interviews, underscoring the need for context-specific validation.[36] In resource-limited settings like India, self-report scales may provide a practical starting point for screening but should ideally be supplemented by clinician-administered evaluations to improve diagnostic accuracy and avoid overdiagnosis.

Our findings align with existing research showing higher prevalence rates of anxiety disorders in females compared to males, with a male-to-female lifetime prevalence ratio of approximately 1:1.7.[37,38] This disparity emerges early in life and is driven by a complex interplay of psychosocial, biological, and temperamental factors. Psychosocially, societal norms often encourage women to express emotions like anxiety and fear, while men are socialized to suppress these emotions, developing coping mechanisms that align with ideals of strength and resilience.[39] Women also face greater exposure to stressors such as sexual violence, domestic abuse, and relationship challenges, compounded by gender disparities in socioeconomic status and work roles.[40] These pressures increase their vulnerability to anxiety.

Biologically, structural differences in brain regions involved in emotion regulation, such as the prefrontal cortex, hippocampus, and amygdala, play a significant role. Female gonadal hormones, including oestrogen and progesterone, affect neurotransmitter systems and anxiety regulation.[39] Fluctuations in these hormones during the menstrual cycle, puberty, and menopause destabilize homeostatic systems, such as the HPA axis and GABAergic pathways. Genetic factors, such as variations in the Serotonin Transporter (SERT) gene and higher Solute Carrier Family 6 Member 4 (SLC6A4) methylation levels, further amplify susceptibility in women.[41]

Temperamental factors, such as neuroticism and rumination that are strongly linked to anxiety and depression, are more pronounced in women.[39] These traits, shaped by gender socialization, contribute to the internalizing tendencies that underlie anxiety and depressive disorders.

Implications for public health and policy

Integration of mental health into human immunodeficiency virus care

According to a survey, the screening rate for anxiety in HIV clinics is notably lower at 14%, compared to 50% for depression.[42] However, data specific to the Indian context remains unavailable, limiting efforts to develop tailored interventions. Current meta-analysis indicates that one out of every three PLHIV suffers from anxiety, highlighting a significant burden that requires immediate attention. This finding underscores the need for routine anxiety screening alongside depression assessments in ART clinics to enable early identification and timely intervention.[43,44] Tools such as the HADS and DASS, when adapted for cultural contexts, can facilitate large-scale screenings. Furthermore, counselling and peer support groups can play a pivotal role in addressing stigma and providing coping mechanisms, fostering holistic mental health care for PLHIV.

Addressing stigma

Stigma reduction must be a cornerstone of HIV programs in India, as it is critical to fostering a supportive environment for PLHIV.[45] Community education campaigns that address misconceptions about HIV can promote awareness and encourage acceptance, helping to dismantle societal stigma. Simultaneously, interventions that facilitate voluntary disclosure and provide psychological counselling are essential in reducing internalized stigma and alleviating associated anxiety.[46] In this effort, community-based organizations and NGOs play a pivotal role by bridging gaps between PLHIV and healthcare services, ensuring access to care, support, and resources while empowering individuals to live healthier, stigma-free lives.[47]

Gender-sensitive approaches

Given the higher anxiety prevalence among women, interventions must adopt a gender-sensitive approach. Programs that enhance women’s economic independence, such as skill development and microfinance initiatives, can alleviate financial stress.[48] Counselling tailored to address women’s unique challenges, including caregiving burdens and societal stigma, is critical.

A key strength of this review lies in its adherence to PRISMA guidelines and the use of robust statistical analyses, including subgroup analyses, sensitivity testing, and meta-regression, to explore sources of heterogeneity. The study provides region-specific insights into the mental health burden of PLHIV in India, addressing a critical gap in the literature. The inclusion of diverse screening tools and both hospital-based and community-based studies enhances the generalizability of the findings.

However, certain limitations must be acknowledged. First, substantial heterogeneity persisted despite subgroup and meta-regression analyses, indicating the presence of unmeasured confounders. Second, variations in the tools used for anxiety screening and their cutoff thresholds may have influenced prevalence estimates. Third, the exclusion of non-English studies may have introduced language bias. Additionally, the predominance of hospital-based studies limits the generalizability of findings to community settings, where anxiety may be underdiagnosed and undertreated. Fourth, none of the included studies systematically accounted for premorbid psychiatric history or pre-existing anxiety conditions that may have contributed to an overestimation of HIV-attributable anxiety burden. The timing of anxiety assessment relative to HIV diagnosis—whether at the point of acute diagnosis, during chronic stable disease, or at advanced illness—was not consistently reported across included studies, precluding subgroup analysis by disease phase, despite this being a clinically meaningful modifier of anxiety prevalence. The majority of included studies did not report anxiety prevalence stratified by key clinical variables such as CD4 count, WHO disease stage, ART adherence status, or comorbid illness severity, making formal subgroup meta-analysis on these dimensions statistically unjustifiable with the available data. A supplementary search of PubMed and EMBASE conducted up to May 2026 identified one additional eligible study published after the original search cutoff (Mantri et al., 2026), which reported an anxiety prevalence of 19.6% among PLHIV in Western Rajasthan using the DASS-21. This study, from a state not previously represented among the included studies, is consistent with the broader findings of this review and does not alter its conclusions.

Future studies should focus on identifying additional factors contributing to the high heterogeneity observed in this analysis, such as socioeconomic determinants, ART adherence, and regional disparities in healthcare access. Longitudinal studies are needed to assess the impact of anxiety on HIV outcomes, including ART adherence, disease progression, and quality of life. Moreover, community-based interventions and culturally sensitive mental health programs should be evaluated for their effectiveness in reducing anxiety in PLHIV.

CONCLUSION

Our study reveals that the prevalence of anxiety disorders among PLHIV in India is significantly higher (36%) than in the general population. Hospital-based settings reported a higher prevalence than community-based studies, and self-report tools tended to overestimate anxiety compared to structured clinical interviews. These findings underscore the critical need for integrated mental health services within HIV care frameworks, tailored to address gender differences and the influence of diagnostic methods. Interventions that combine routine anxiety screening, stigma reduction, and culturally sensitive care can effectively reduce the mental health burden in this population.

Conflicts of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Supplementary Figure S1

Forest plot showing subgroup analysis of the pooled prevalence of anxiety among PLHIV, stratified by the screening tools used (DASS, GAD-7, HADS, ICD-10, DSM-IV)

Supplementary Figure S2

Forest plot showing subgroup analysis of the pooled prevalence of anxiety among PLHIV, stratified by gender

Supplementary Figure S3

Funnel plot assessing publication bias in the included studies

Supplementary Figure S4

Bubble plot showing the relationship between study sample size and the prevalence of anxiety

Supplementary Figure S5

Bubble plot showing the association between the prevalence of depression and the prevalence of anxiety among PLHIV

Funding Statement

Nil.