Association of Anxiety and Depression with Mortality in Rural Nepal: A 16-Year Longitudinal Community Cohort Study
by
, , ,
Sauharda Rai
1,2,*
,
Ruta Rangel
1,
Isabella Calamari
3, 
Safar Bikram Adhikari
2,4,
Bonnie N. Kaiser
5,
Nanda Raj Acharya
2,
Suraj Koirala
2 and
Brandon A. Kohrt
1,2,6
1
Center for Global Mental Health Equity, Department of Psychiatry and Behavioral Health, George Washington University, Washington, DC 20037, USA
2
Transcultural Psychosocial Organization (TPO-Nepal), Kathmandu 44616, Nepal
3
Columbian College of Arts and Sciences, George Washington University, Washington, DC 20037, USA
4
Child and Adolescent Mental Health Service (CAMHS), Canberra Health Services, Canberra 2605, Australia
5
Department of Anthropology and Global Health, University of California San Diego, CA 92093, USA
6
Duke Global Health Institute, Duke University, Durham, NC 27708, USA
*
Author to whom correspondence should be addressed.
Psychiatry Int. 2026, 7(5), 222; https://doi.org/10.3390/psychiatryint7050222
Submission received: 17 August 2026
/
Revised: 28 September 2026
/
Accepted: 30 September 2026
/
Published: 2 October 2026
(This article belongs to the Section Mental Health)
Abstract
Background: Common mental health conditions are associated with increased mortality, but longitudinal evidence from low- and middle-income countries (LMICs) remains limited. We aimed to examine whether baseline depression and anxiety were associated with all-cause mortality over 16 years in rural Nepal. Methods: We conducted a 16-year prospective cohort study in rural Nepal. At baseline in 2000, 316 adults were assessed using Nepali-validated versions of the Beck Depression Inventory (BDI; cut-off ≥ 20) and Beck Anxiety Inventory (BAI; cut-off ≥ 17). In 2016, 297 participants were successfully followed up; 46 had died and verbal autopsy data were available for 40 deaths. Logistic regression models were adjusted for gender, ethnicity, age, education, and income. Results: Among the 297 participants followed up, 34% of participants scored above the cut-off for depression, 27% for anxiety, and 17% for both at baseline. Forty-six participants (15%) died during follow-up. Among the 40 deaths with completed verbal autopsy, chronic respiratory disease was the leading cause of death (23/40, 57.5%). Adjusted odds of mortality were higher among participants scoring above the cut-off for depression (OR 2.34, 95% CI 1.01–5.53), anxiety (OR 3.49, 95% CI 1.27–10.15), and comorbid depression and anxiety (OR 5.27, 95% CI 1.65–17.61). In the comorbidity-adjusted model, women had lower odds of mortality (OR 0.22, 95% CI 0.07–0.61) and Dalit participants had higher odds (OR 4.62, 95% CI 1.44–16.96). Conclusions: Anxiety and comorbid depression and anxiety were associated with higher mortality over 16 years. Depression alone was associated with mortality in logistic regression, but this association was not statistically significant in sensitivity analyses and should therefore be interpreted cautiously.
1. Introduction
Mental health conditions are a leading contributor to the global burden of disease [1] and are associated with premature mortality [2]. Each year, 8 million, or 14.3% of deaths worldwide, are attributed to mental health conditions with studies showing that serious mental health conditions, such as schizophrenia and bipolar disorder, can reduce life expectancy by 14 years, and by 20 years for those with substance use conditions [2,3]. However, these differences vary not only by mental health condition but also by individual and societal contexts, including gender [2], ethnicity [4], health behavior [5], access to healthcare [6], and lifestyle choices [7]. Recognizing these contextual factors is crucial for developing strategies to mitigate the impact of mental health on mortality.
However, most of the available research on mental health and mortality has been conducted in high-income countries (HICs), with limited data from low-and middle-income countries (LMICs) [2,3]. This is alarming as 85% of the world’s population lives in LMICs, where mental health conditions are a major contributor to the global burden of disease [8], and the treatment gap is widest [9]. One notable longitudinal study, the 10/66 cohort conducted across eight LMICs, found that depression and anxiety were associated with higher mortality among older adults [10]. However, evidence remains limited across younger and broader adult populations and across diverse low-resource settings.
Nepal, a South Asian lower-middle-income country provides an important setting to examine these relationships. Nepal shares many challenges with other LMICs, including high exposure to negative social determinants of mental health such as gender inequity, poverty, social inequality, and a fragile mental health system [11,12]. Prior studies in Nepal have shown that depression and anxiety are higher among women [13], the lower-caste group [4], and individuals exposed to conflict [14]. These social determinants are also associated with differences in physical health, healthcare access and mortality [15]. Despite these intersecting vulnerabilities, the long-term relationship between common mental health conditions and mortality has not been examined in a longitudinal community cohort in Nepal.
This study therefore aimed to determine whether baseline depression and anxiety were associated with all-cause mortality over 16 years in a prospective community cohort of adults, hypothesizing that participants with depression, anxiety, or both would have a higher risk of mortality than those with neither condition.
2. Materials and Methods
2.1. Study Background
In 2000, 316 individuals were recruited from three Village Development Committees (VDCs)—Chandannath, Mahatgaon and Kartikswami in Jumla district of rural Nepal using an nth-household sampling strategy with one randomly selected adult per sampled household [16]. Sampling was not stratified by any participant characteristics. Participants were eligible if they were 18 years or older and provided informed consent.
In 2016, a follow-up study was conducted among these participants, which included conducting verbal autopsies with families of deceased participants to record mortality details. Trained local researchers conducted follow-up interviews by visiting participants’ households, local communities, or nearby cities where participants had migrated. For participants who had died, verbal autopsy interviews were conducted with consenting family members to collect information on the circumstances and likely cause of death. Participants who could not be located or whose vital status could not be confirmed were classified as lost to follow-up. A total of 297 participants (94%) were successfully followed up and comprised the analytical sample for this study.
Before data collection, researchers were trained in obtaining informed consent and administering the study instruments using the data collection software—Open Data Kit Version Collect 1.4.7 (ODK) [17]. Researchers underwent standardized training in consent procedures, instrument administration, and verbal-autopsy interviewing. Practice interviews and pilot verbal-autopsy forms were reviewed by the study team to identify discrepancies and standardize administration before data collection. All other instruments had been tested during the previous phase of the study [16,18].
2.2. Study Settings
The study site, Jumla, is a northern hilly district in mid-western Nepal. In 2000, when the baseline data were collected, Jumla had a population of 69,226 and the most recent census before the 2016 follow-up reported a population of 108,921 with balanced gender distribution. A total of 98% of the population identify as Hindu, followed by Buddhist and Christian. In terms of ethnicity, Chhetri (higher-caste) comprised 60% of the population, followed by 10% Hill-Brahman (higher-caste), 18% Dalit (lower-caste), and 1.5% Janajati (ethnic minorities). The literacy rate has improved from 32% in 2001, to 54% in 2011 [19]. Mental health services in Jumla are scarce. In 2011, Karnali Academy of Health Sciences was established, which now provides basic mental health services in the area [20]. During the 16-year study period, the site also went through a series of socio-political events, importantly the Maoist civil war. The conflict affected many aspects of daily life, including access to healthcare, education, livelihoods and population mobility [18].
2.3. Instruments
Depression and anxiety assessment—Nepali-validated versions of the Beck Depression Inventory (BDI) and Beck Anxiety Inventory (BAI) were used to assess depression and anxiety at baseline [21,22]. Both tools consist of 21 items assessing the participants’ anxiety and depressive symptoms in the past two weeks. Each item is scored from 0–3, resulting in an instrument range between 0–63. These instruments were validated with populations in Jumla and Kathmandu with clinical DSM-IV diagnosis of major depressive disorder and generalized anxiety disorder. Area under the curve (AUC) for the BDI was 0.92 (95% CI 0.88–0.96) and for BAI was 0.85 (95% CI 0.79–0.91). Based on this, BDI score ≥ 20 indicated moderate depression (sensitivity = 0.73, specificity = 0.91) and BAI ≥ 17 suggested moderate anxiety (sensitivity = 0.77, specificity = 0.81). Two-week test–retest reliability (Spearman–Brown correlation) was 0.84 for the BDI and 0.88 for the BAI. In the current analytical sample (n = 297), both instruments demonstrated high internal consistency. Cronbach’s alpha was 0.90 for the BDI and 0.89 for the BAI, while McDonald’s omega was 0.90 for the BDI and 0.90 for the BAI.
Verbal autopsies for mortality data—During the 2016 follow-up, deaths were identified and confirmed through family members or caregivers. For deceased participants, trained study researchers administered a verbal autopsy questionnaire adapted from a previous study conducted in the same study area [23] to a consenting family member or caregiver knowledgeable about the circumstances surrounding the death. The questionnaire collected information on signs and symptoms preceding death, relevant medical history, and circumstances of death. Cause-of-death information was based on the respondent’s accounts, as medical records and death certificates were unavailable for majority of the rural population. During the survey, the respondents were asked an open-ended question regarding the perceived cause of death, and the responses were reviewed and categorized by a physician (BAK). No formal blinding procedure was implemented; however, the physician received the only reported cause of death responses for categorization without any information identifying the participant’s baseline BDI and BAI status.
2.4. Ethics
Approval for the 2000 study was provided by the Department of Psychiatry at Tribhuvan University Teaching Hospital/Institute of Medicine. For the 2016 follow-up study, ethical approval was provided by the Institutional Review Board of Duke University (Pro00052631) and the Nepal Health Research Council (Reg.no. 39/2016). Literate participants provided written consent, while verbal consent with a thumbprint was obtained from those who could not read and write. A similar process was followed with family members of deceased participants who took part in the verbal autopsy.
Any participants who reported high levels of psychological distress and impaired functioning were evaluated by the study team and referred for psychosocial care during the consent process and the interviews.
2.5. Analysis
For this paper, we included participants successfully followed up in 2016 and used their baseline data from 2000 as predictor variables (Table 1). Baseline variables included anxiety, depression, education, income, age, ethnicity, and gender. Anxiety and depression were analyzed as binary variables using the validated cut-offs described above. Comorbidity was categorized as neither condition, either condition, or both conditions. Education (any schooling vs. no schooling), income (any income vs. no income), and gender (male vs. female) were dichotomized. Ethnicity was classified as: (1) high-caste Brahmin, (2) Chhetri/Janajati, and (3) low-caste Dalit. Janajati participants were grouped with Chhetri participants because Janajatis comprised less than 3% of the sample and no deaths occurred in this group; both groups also occupy an intermediate social position relative to Brahmins and Dalits in the local context. Baseline age was categorized as 18–25 years to represent younger adults, followed by approximately 10-year age intervals (26–34 and 35–44 years), with participants aged 45 years and older grouped together. These categories also resulted in reasonably distributed group sizes at baseline. The primary outcome was all-cause mortality (dead vs. living). Baseline exposure and covariate data were complete for the 297 participants in the analytical sample. Among the 46 deceased participants, cause-of-death and time-to-death information were unavailable for six participants because verbal autopsies were not completed. Analyses were conducted in R (version 4.3.2) using RStudio (Version 2023.09.1+494) [24]. Cox proportional hazards models were fitted using the survival package, and Firth penalized logistic regression models were fitted using the logistf package.
We first used Fisher’s exact test to compare baseline characteristics between participants who were successfully followed up and those lost to follow-up, because several expected cell counts were small. We then used logistic regression to estimate crude and adjusted odds ratios (ORs) and 95% confidence intervals (CIs) for the association between baseline common mental health conditions and mortality. Three crude models separately examined depression, anxiety, and comorbidity. Adjusted models included gender, ethnicity, age, education, and income as covariates. Gender and ethnicity were included because both mental health prevalence and mortality may vary across these groups; age was included to account for aging, and education and income were included as socioeconomic covariates. Separate logistic regression models examined baseline associations of gender and ethnicity with depression and anxiety.
As a sensitivity analysis, Cox proportional hazards models were fitted among participants with available time-to-death information to assess the robustness of the primary findings (Supplementary Table S1). The six deceased participants without time-to-death information were excluded from the Cox analysis. The proportional-hazards assumption for the Cox models was assessed using Schoenfeld residuals. Additionally, to assess the potential influence of small-sample bias and model overfitting, we also conducted Firth penalized logistic regression sensitivity analyses using the same exposure and covariate specifications as the primary adjusted logistic regression models (Supplementary Table S2). These analyses included all 297 participants in the analytical sample.
This study is reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines for cohort studies (See Supplementary File S1).
3. Results
3.1. Cohort Characteristics and Follow-Up
Table 1 summarizes baseline characteristics of the study population and follow-up status. Among the 316 participants recruited in 2000, 297 (94%) were successfully followed up in 2016. Fisher’s exact tests showed no statistically significant differences in baseline demographic or mental health characteristics between participants successfully followed up and those lost to follow-up. Of those followed up, 251 (85%) were alive and 46 (15%) had died. Females accounted for 41% of the follow-up sample. In terms of ethnicity, 51% belonged to the Chhetri/Janajati group, 25% to the Brahmin group, and 24% to the Dalit group. Forty-four percent had no education and 45% reported no source of income. At baseline, 34% scored above the cut-off for depression, 27% for anxiety, and 17% for both depression and anxiety.
3.2. Mortality and Causes of Death
Of the 46 participants who had died, 78% were male, 56% belonged to the Dalit group, 61% had no education, and 56% reported no source of income. Verbal autopsies were completed for 40 (87%) of the 46 deceased participants. Among these 40 participants, 27 had at least one baseline mental health condition and 13 had neither depression nor anxiety. Median age at death was 63 years (IQR 53–68) among participants with at least one mental health condition and 65 years (IQR 56–69) among those with neither condition; these comparisons are descriptive and were not formally tested. Among the 16 participants with comorbid depression and anxiety who had a completed verbal autopsy, median age at death was 62 years (IQR 53–66).
Table 2 presents cause of death by baseline mental health status. Among the 40 deceased participants with completed verbal autopsies, chronic respiratory disease was the leading cause of death (23/40, 57.5%), followed by alcohol-related causes (8/40, 20.0%). No deaths were attributed to suicide or self-harm. Six deceased participants did not have a completed verbal autopsy and therefore had no assigned cause of death.
3.3. Association of Depression and Anxiety with Mortality
Table 3 presents the crude associations between depression, anxiety, comorbidity, and mortality. In Model 1, participants with depression scores above the cut-off had more than three times the odds of death compared with those below the cut-off (OR 3.11, 95% CI 1.64–5.98). In Model 2, participants with anxiety scores above the cut-off had nearly three times the odds of death (OR 2.73, 95% CI 1.42–5.24). In Model 3, participants with both anxiety and depression scores above the cut-off had almost five times the odds of death compared with those with neither condition (OR 4.75, 95% CI 2.18–10.44).
Table 4 presents the adjusted models. After adjustment for gender, ethnicity, age, education, and income, depression above the cut-off was associated with higher odds of mortality (OR 2.34, 95% CI 1.01–5.53), as was anxiety above the cut-off (OR 3.49, 95% CI 1.27–10.15). In the comorbidity model, participants scoring above the cut-off for both depression and anxiety had more than five times the odds of mortality compared with those with neither condition (OR 5.27, 95% CI 1.65–17.61). Across adjusted models, women had lower odds of mortality and Dalit participants had higher odds compared with men and Brahmin participants, respectively. Estimates for age were imprecise, particularly for participants aged ≥45 years, because only three deaths occurred in the 18–25-year reference group (compared with 2, 6, and 35 deaths in the 26–34, 35–44, and ≥45-year groups, respectively).
3.4. Baseline Correlates of Depression and Anxiety
To examine baseline demographic correlates of depression and anxiety, we fitted separate logistic regression models with depression and anxiety as outcomes (Table 5; n = 297). Each model included gender and ethnicity simultaneously and did not adjust for additional covariates. Women had significantly higher odds of scoring above the cut-off for both depression and anxiety. Compared with Brahmins, participants in the Chhetri/Janajati group had lower odds of anxiety, whereas Dalit participants had higher odds of scoring above the cut-off for both depression and anxiety.
3.5. Sensitivity Analysis
As sensitivity analyses, Cox proportional hazards models and Firth penalized logistic regression models were fitted. In the Firth models, findings for anxiety and comorbid depression and anxiety were consistent with the primary logistic regression analyses. Anxiety remained associated with mortality (OR 3.21, 95% CI 1.21–8.96, p = 0.019), as did comorbid depression and anxiety (OR 4.72, 95% CI 1.55–14.91, p = 0.006). The estimate for depression was similar in magnitude to the primary logistic regression result but was not statistically significant (OR 2.23, 95% CI 0.99–5.10, p = 0.054) (Table S2). In the Cox analysis, findings for anxiety and comorbid anxiety and depression were also broadly consistent with the primary analyses. For depression, the direction of association was unchanged, but the adjusted hazard ratio was attenuated and not statistically significant (HR 1.57, 95% CI 0.78–3.15, p = 0.20) (Table S1). The proportional-hazards assumption was not violated in any of the Cox models based on Schoenfeld residual tests though the test was borderline for depression (global p = 0.08 for depression, p = 0.38 for anxiety, and p = 0.096 for comorbidity).
4. Discussion
4.1. Principal Findings
This 16-year longitudinal cohort study explored the association between depression, anxiety and mortality in a rural setting in a low-and middle-income country. In adjusted logistic regression analyses, depression, anxiety, and comorbid depression and anxiety were associated with higher odds of mortality, with the strongest association observed for comorbidity. However, the association for depression alone was less robust across sensitivity analyses. Although the effect estimate was similar in magnitude in the Firth penalized logistic regression, its confidence interval included the null, and depression was also not statistically significant in the Cox time-to-event analysis. In contrast, findings for anxiety and comorbidity were broadly consistent across the primary and sensitivity analyses. The difference between the logistic and Cox results for depression may partly reflect lower precision in the time-to-event analysis because only 40 deaths had complete time information, as well as the fact that the Cox model incorporates timing of death rather than mortality status alone. Accordingly, the evidence for an independent association between baseline depression and mortality should be interpreted cautiously. The median age at death differed by only approximately two years between participants with and without a mental health condition, which also warrants caution in interpreting these findings as evidence of substantially shortened lifespan in this cohort.
4.2. Comparison with Existing Evidence
Our findings align with the current literature demonstrating an association between depression, anxiety, comorbid depression and anxiety, and mortality, with studies reporting higher odds of mortality among individuals experiencing these conditions [25,26,27]. These findings are also consistent with results from the 10/66 cohort study conducted across eight low- and middle-income countries, which reported that depression and anxiety were associated with higher mortality among older adults [10].
4.3. Gender, Caste and Social Determinants of Mortality
Consistent with other studies, we found that men had higher odds of mortality than women during the follow-up period [25]. At baseline, women had approximately four times the odds of anxiety and 2.65 times the odds of depression compared with men. The mechanisms underlying these gender differences could not be evaluated in this study. Potential pathways involving gendered coping, tobacco or alcohol use, and household air pollution are plausible contextual hypotheses [28,29,30], but these exposures were not systematically measured and should be examined directly in future research. Alcohol-related causes accounted for 20% (8/40) of deaths with completed verbal autopsies, further supporting the need for future work on behavioral and environmental pathways.
We found that the lower-caste Dalits had approximately five times higher odds of mortality and around three times higher odds of having depression and anxiety compared to higher-caste Brahmins. This is in line with earlier studies which show that Dalits are more prone to have mental health conditions [4] and are less likely to use health services [31]. Interestingly, participants from the Chhetri/Janajati group had lower odds of anxiety compared with Brahmins. Although this study was not designed to examine the mechanisms underlying differences between caste and ethnic groups, the finding highlights the heterogeneity of mental health risk across social groups in rural Nepal [4].
4.4. Chronic Disease Burden and Causes of Death
Among the 40 deceased participants with verbal autopsy, 25 (62.5%) died of chronic health conditions, including asthma, chronic obstructive pulmonary disease (COPD), and cancer. Chronic conditions were the leading cause of death among participants with comorbid anxiety and depression. These findings highlight the substantial burden of chronic disease in this rural Nepali population. However, because chronic health conditions were not systematically assessed among all cohort participants, we were unable to evaluate their role in the association between depression, anxiety, and mortality.
Among participants with neither depression nor anxiety who had a completed verbal autopsy, chronic disease causes accounted for 9 of 13 deaths (69.2%). This pattern suggests that chronic diseases were a major cause of mortality across the study population rather than being specific to individuals with mental health conditions. The high burden of chronic disease-related mortality may reflect broader limitations in access to healthcare in this rural setting. At baseline, the district had only nine health posts, 20 sub-health posts, and one hospital [32], and healthcare resources remain limited. Strengthening access to prevention, diagnosis, and management of chronic diseases may therefore be an important strategy for reducing premature mortality in the region.
4.5. Limitations
This study provides valuable insights into the association between common mental health conditions and mortality in LMICs. However, several limitations should be considered when interpreting the results. First, the relatively small sample of 297 participants and 46 deaths may have increased the risk of overfitting and imprecise estimates. Firth penalized logistic regression sensitivity analyses showed broadly similar findings for anxiety and comorbidity, while the depression association was similar in magnitude but not statistically significant. Second, the depression and anxiety data were based on the assessment in 2000. This approach does not account for potential changes in mental health status over time, which could influence mortality. In addition, the Beck Depression Inventory and Beck Anxiety Inventory are symptom screening instruments and do not represent clinical diagnoses. Third, although the analysis adjusted for several demographic and socioeconomic factors, other potential confounders, such as physical health status, access to healthcare, health behaviors, and especially the direct and indirect effects of Maoist conflict that occurred between the follow-up period, were not accounted for. Baseline physical health and somatic disease were not systematically assessed, preventing adjustment for pre-existing chronic disease burden. This is particularly relevant because chronic respiratory disease accounted for 23 of 40 deaths (57.5%) with completed verbal autopsies. Fourth, logistic regression was used as the primary analytic approach because complete time-to-death information was unavailable for six deceased participants. Cox sensitivity analyses were broadly consistent for anxiety and comorbidity, but depression was no longer statistically significant, underscoring the need for cautious interpretation of the depression finding. Finally, recall biases among family members during the verbal autopsy and lack of proper medical records and death certificates in Nepal might result in potential inaccuracies in accessing the cause of death. Despite these limitations, this study provides important preliminary evidence on the relationship between common mental disorders and mortality in a low-resource setting, highlighting the need for further research in this area.
4.6. Implications for Policy and Practice
These findings contribute to the limited longitudinal evidence on common mental health conditions and mortality in LMICs. They support consideration of mental health alongside broader social and physical-health determinants within routine health services, particularly where access to specialized mental health care is limited. Integration of mental health assessment and care into primary and community-based services may provide opportunities for earlier identification and support, while future research should clarify the mechanisms linking mental health, chronic disease, and mortality.
5. Conclusions
This research contributes longitudinal evidence on mental health and mortality in a rural LMIC setting. Baseline anxiety and comorbid depression and anxiety were associated with higher mortality over 16 years in both the primary logistic regression analyses and the time-to-event sensitivity analyses. Depression alone was associated with mortality in the logistic regression analysis but was not statistically significant in sensitivity analyses and should therefore be interpreted cautiously. Mortality was also disproportionately concentrated among men and Dalit participants, highlighting the importance of considering social and physical-health determinants alongside mental health in low-resource settings.
Supplementary Materials
The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/psychiatryint7050222/s1, Table S1: Cox proportional hazards models examining the association between mental health conditions and mortality. Table S2: Firth penalized logistic regression sensitivity analyses examining the association between mental health conditions and mortality. Supplementary File S1: STROBE Statement—checklist of items that should be included in reports of observational studies.
Author Contributions
Conceptualization, S.R., S.K. and B.A.K.; Methodology, S.K.; Formal Analysis, S.R. and B.A.K.; Investigation, N.R.A.; Data Curation, S.R. and B.A.K.; Writing—Original Draft Preparation, S.R., R.R., I.C., and B.A.K.; Writing—Review and Editing, S.R., R.R., I.C., S.B.A., B.N.K., S.K. and B.A.K.; Visualization, R.R.; Supervision, S.R., S.B.A., B.N.K., N.R.A. and B.A.K.; Project Administration, S.B.A. and N.R.A.; Funding Acquisition, B.A.K. All authors have read and agreed to the published version of the manuscript.
Funding
Funding for this study was provided by Fulbright Fellowship, US National Institute of Mental Health (NRSA F31 MH075584, K01MH104310-01), the Wenner-Gren Foundation, the Graduate School of Arts and Sciences and the Department of Anthropology at Emory University and Duke Global Health Institute, Duke University.
Institutional Review Board Statement
The study was conducted in accordance with the Declaration of Helsinki and approved by the Department of Psychiatry at Tribhuvan University Teaching Hospital/Institute of Medicine, the Institutional Review Board of Duke University (protocol code Pro00052631; approval date: 21 May 2014), and the Nepal Health Research Council (protocol code Reg. No. 39/2016; approval date: 30 March 2016).
Informed Consent Statement
Informed consent was obtained from all subjects involved in the study.
Data Availability Statement
The data supporting the findings of this study are available from the corresponding author upon reasonable request.
Acknowledgments
The authors acknowledge the research staff and participants from Jumla, Nepal, especially Ganesh Rokaya and Indra Subba. We also acknowledge staff at the Center for Global Mental Health Equity, George Washington University for their valuable comments.
Conflicts of Interest
Author Brandon A. Kohrt received funding from the Wenner-Gren Foundation. The funding body played no role in the study design, data collection, analysis or interpretation, manuscript writing, or the decision to publish the results. The remaining au-thors 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.
References
- GBD 2019 Mental Disorders Collaborators. Global, regional, and national burden of 12 mental disorders in 204 countries and territories, 1990–2019: A systematic analysis for the Global Burden of Disease Study 2019. Lancet Psychiatry 2022, 9, 137–150. [Google Scholar] [CrossRef] [Scilit] [PubMed]
- Walker, E.R.; McGee, R.E.; Druss, B.G. Mortality in mental disorders and global disease burden implications: A systematic review and meta-analysis. JAMA Psychiatry 2015, 72, 334–341. [Google Scholar] [PubMed]
- Chan, J.K.N.; Correll, C.U.; Wong, C.S.M.; Chu, R.S.T.; Fung, V.S.C.; Wong, G.H.S.; Lei, J.H.C.; Chang, W.C. Life expectancy and years of potential life lost in people with mental disorders: A systematic review and meta-analysis. EClinicalMedicine 2023, 65, 102294. [Google Scholar] [CrossRef] [Scilit] [PubMed]
- Hawes, A.M.; Axinn, W.G.; Ghimire, D.J. Ethnicity and psychiatric disorders. Ann. Psychiatry Ment. Health 2016, 4, 1072. [Google Scholar] [PubMed]
- Fortuna, K.L.; Williams, A.; Mois, G.; Jason, K.; Bianco, C.L. Social processes associated with health and health behaviors linked to early mortality in people with a diagnosis of a serious mental illness. Perspect. Psychol. Sci. 2022, 17, 183–190. [Google Scholar] [CrossRef] [Scilit] [PubMed]
- Laursen, T.M.; Nordentoft, M.; Mortensen, P.B. Excess early mortality in schizophrenia. Annu. Rev. Clin. Psychol. 2014, 10, 425–448. [Google Scholar] [CrossRef] [Scilit] [PubMed]
- Walsh, R. Lifestyle and mental health. Am. Psychol. 2011, 66, 579. [Google Scholar] [CrossRef] [Scilit] [PubMed]
- Patel, V.; Saxena, S.; Lund, C.; Thornicroft, G.; Baingana, F.; Bolton, P.; Chisholm, D.; Collins, P.Y.; Cooper, J.L.; Eaton, J. The Lancet Commission on global mental health and sustainable development. Lancet 2018, 392, 1553–1598, Correction in Lancet 2018, 392, 1518. [Google Scholar] [CrossRef] [Scilit] [PubMed]
- Thornicroft, G.; Chatterji, S.; Evans-Lacko, S.; Gruber, M.; Sampson, N.; Aguilar-Gaxiola, S.; Al-Hamzawi, A.; Alonso, J.; Andrade, L.; Borges, G. Undertreatment of people with major depressive disorder in 21 countries. Br. J. Psychiatry 2017, 210, 119–124. [Google Scholar] [CrossRef] [Scilit] [PubMed]
- Wu, Y.T.; Kralj, C.; Acosta, D.; Guerra, M.; Huang, Y.; Jotheeswaran, A.T.; Jimenez-Velazquez, I.Z.; Liu, Z.; Llibre Rodriguez, J.J.; Salas, A. The association between, depression, anxiety, and mortality in older people across eight low-and middle-income countries: Results from the 10/66 cohort study. Int. J. Geriatr. Psychiatry 2020, 35, 29–36. [Google Scholar] [CrossRef] [Scilit] [PubMed]
- Lund, C.; Brooke-Sumner, C.; Baingana, F.; Baron, E.C.; Breuer, E.; Chandra, P.; Haushofer, J.; Herrman, H.; Jordans, M.; Kieling, C. Social determinants of mental disorders and the Sustainable Development Goals: A systematic review of reviews. Lancet Psychiatry 2018, 5, 357–369. [Google Scholar] [CrossRef] [Scilit] [PubMed]
- Rai, Y.; Gurung, D.; Gautam, K. Insight and challenges: Mental health services in Nepal. BJPsych Int. 2021, 18, E5. [Google Scholar] [CrossRef] [Scilit] [PubMed]
- Shawon, M.S.R.; Hossain, F.B.; Hasan, M.; Rahman, M.R. Gender differences in the prevalence of anxiety and depression and care seeking for mental health problems in Nepal: Analysis of nationally representative survey data. Camb. Prism. Glob. Ment. Health 2024, 11, e46. [Google Scholar] [CrossRef] [Scilit] [PubMed]
- Axinn, W.G.; Ghimire, D.J.; Williams, N.E.; Scott, K.M. Gender, traumatic events, and mental health disorders in a rural Asian setting. J. Health Soc. Behav. 2013, 54, 444–461. [Google Scholar] [CrossRef] [Scilit] [PubMed]
- Ghimire, U.; Manandhar, J.; Gautam, A.; Tuladhar, S.; Prasai, Y.; Gebreselassie, T. Inequalities in Health Outcomes and Access to Services by Caste/Ethnicity, Province, and Wealth Quintile in Nepal; ICF: Rockville, MD, USA, 2019; Available online: http://dhsprogram.com/pubs/pdf/FA117/FA117.pdf (accessed on 17 August 2026).
- Kohrt, B.A. “Somatization” and “Comorbidity”: A Study of Jhum-Jhum and Depression in Rural Nepal. Ethos 2005, 33, 125–147. [Google Scholar] [CrossRef] [Scilit]
- Hartung, C.; Lerer, A.; Anokwa, Y.; Tseng, C.; Brunette, W.; Borriello, G. Open data kit: Tools to build information services for developing regions. In Proceedings of the 4th ACM/IEEE International Conference on Information and Communication Technologies and Development, London, UK, 13–16 December 2010; pp. 1–12. [Google Scholar]
- Kohrt, B.A.; Hruschka, D.J.; Worthman, C.M.; Kunz, R.D.; Baldwin, J.L.; Upadhaya, N.; Acharya, N.R.; Koirala, S.; Thapa, S.B.; Tol, W.A. Political violence and mental health in Nepal: Prospective study. Br. J. Psychiatry 2012, 201, 268–275. [Google Scholar] [CrossRef] [Scilit] [PubMed]
- Central Bureau of Statistics (Nepal). National Population and Housing Census 2011: National Report; Government of Nepal, National Planning Commission Secretariat, Central Bureau of Statistics: Kathmandu, Nepal, 2012; Volume 1. Available online: https://microdata.nsonepal.gov.np/index.php/catalog/54/related-materials (accessed on 15 August 2023).
- Poudel, S.; Parajuli, A.; Saud, D.; Sharma, P.; Acharya, R. Mental health service delivery in rural Nepal: Opportunities and challenges with insights from Jumla (Karnali Province). J. Karnali Acad. Health Sci. 2025, 8, 57–60. [Google Scholar] [CrossRef] [Scilit]
- Kohrt, B.A.; Kunz, R.D.; Koirala, N.R.; Sharma, V.D.; Nepal, M.K. Validation of a Nepali version of the Beck Depression Inventory. Nepal. J. Psychiatry 2002, 2, 123–130. [Google Scholar]
- Kohrt, B.A.; Koirala, N.; Sharma, V.; Nepal, M. Validation of the Nepali version of beck anxiety inventory. J. Inst. Med. Nepal 2003, 25, 5–9. [Google Scholar] [CrossRef] [Scilit]
- Hagaman, A.K.; Khadka, S.; Wutich, A.; Lohani, S.; Kohrt, B.A. Suicide in Nepal: Qualitative findings from a modified case-series psychological autopsy investigation of suicide deaths. Cult. Med. Psychiatry 2018, 42, 704–734. [Google Scholar] [CrossRef] [Scilit] [PubMed]
- Posit Team. RStudio: Integrated Development Environment for R; Posit Team: Boston, MA, USA, 2025. [Google Scholar]
- Xia, W.; Jiang, H.; Di, H.; Feng, J.; Meng, X.; Xu, M.; Gan, Y.; Liu, T.; Lu, Z. Association between self-reported depression and risk of all-cause mortality and cause-specific mortality. J. Affect. Disord. 2022, 299, 353–358. [Google Scholar] [CrossRef] [Scilit] [PubMed]
- Meier, S.M.; Mattheisen, M.; Mors, O.; Mortensen, P.B.; Laursen, T.M.; Penninx, B.W. Increased mortality among people with anxiety disorders: Total population study. Br. J. Psychiatry 2016, 209, 216–221. [Google Scholar] [CrossRef] [Scilit] [PubMed]
- Bobo, W.V.; Grossardt, B.R.; Virani, S.; St Sauver, J.L.; Boyd, C.M.; Rocca, W.A. Association of depression and anxiety with the accumulation of chronic conditions. JAMA Netw. Open 2022, 5, e229817. [Google Scholar] [CrossRef] [Scilit] [PubMed]
- Garey, L.; Olofsson, H.; Garza, T.; Shepherd, J.M.; Smit, T.; Zvolensky, M.J. The role of anxiety in smoking onset, severity, and cessation-related outcomes: A review of recent literature. Curr. Psychiatry Rep. 2020, 22, 38. [Google Scholar] [CrossRef] [Scilit] [PubMed]
- Puddephatt, J.A.; Irizar, P.; Jones, A.; Gage, S.H.; Goodwin, L. Associations of common mental disorder with alcohol use in the adult general population: A systematic review and meta-analysis. Addiction 2022, 117, 1543–1572. [Google Scholar] [CrossRef] [Scilit] [PubMed]
- Shrestha, I.L.; Shrestha, S.L. Indoor air pollution from biomass fuels and respiratory health of the exposed population in Nepalese households. Int. J. Occup. Environ. Health 2005, 11, 150–160. [Google Scholar] [CrossRef] [Scilit] [PubMed]
- Thapa, R.; van Teijlingen, E.; Regmi, P.; Heaslip, V. Uptake of health services by people from the dalit community. J. BP Koirala Inst. Health Sci. 2018, 1, 1–6. [Google Scholar] [CrossRef] [Scilit]
- Kohrt, B.A.; Worthman, C.M. Gender and anxiety in Nepal: The role of social support, stressful life events, and structural violence. CNS Neurosci. Ther. 2009, 15, 237–248. [Google Scholar] [CrossRef] [Scilit] [PubMed]
Table 1.
Baseline characteristics of study participants.
| Characteristic | Full Sample 2000 (n = 316) | Follow-Up Status 2016 (n = 316) | ||||
|---|---|---|---|---|---|---|
| Follow-Up Completed (n = 297) | Lost to Follow-Up (n = 19) | Fisher’s Exact p-Value * | ||||
| All Follow-Up Sample (n = 297) | Living (n = 251) | Deceased (n = 46) | ||||
| Gender | ||||||
| Male | 183 (58%) | 176 (59%) | 140 (56%) | 36 (78%) | 7 (37%) | 0.09 |
| Female | 133 (42%) | 121 (41%) | 111 (44%) | 10 (22%) | 12 (63%) | |
| Ethnicity | ||||||
| Brahmin | 79 (25%) | 75 (25%) | 70 (28%) | 5 (11%) | 4 (21%) | 0.61 |
| Chhetri/Janajati | 162 (51%) | 150 (51%) | 135 (54%) | 15 (33%) | 12 (64%) | |
| Dalit | 75 (24%) | 72 (24%) | 46 (18%) | 26 (56%) | 3 (15%) | |
| Age | ||||||
| 18–25 years | 92 (29%) | 82 (28%) | 79 (31%) | 3 (7%) | 10 (53%) | 0.18 |
| 26–34 years | 81 (26%) | 77 (26%) | 75 (30%) | 2 (4%) | 4 (21%) | |
| 35–44 years | 65 (21%) | 63 (21%) | 57 (23%) | 6 (13%) | 2 (10%) | |
| 45 and above | 78 (24%) | 75 (25%) | 40 (16%) | 35 (76%) | 3 (16%) | |
| Education | ||||||
| No education | 137 (43%) | 130 (44%) | 102 (41%) | 28 (61%) | 7 (37%) | 0.64 |
| Any education | 179 (57%) | 167 (56%) | 149 (59%) | 18 (39%) | 12 (63%) | |
| Income | ||||||
| No income | 142 (45%) | 134 (45%) | 108 (43%) | 26 (56%) | 8 (42%) | 1.00 |
| Any income | 174 (55%) | 163 (55%) | 143 (57%) | 20 (44%) | 11 (58%) | |
| Depression | ||||||
| Below cut-off | 213 (67%) | 197 (66%) | 177 (71%) | 20 (44%) | 16 (84%) | 0.13 |
| Above cut-off | 103 (33%) | 100 (34%) | 74 (29%) | 26 (56%) | 3 (16%) | |
| Anxiety | ||||||
| Below cut-off | 230 (73%) | 217 (73%) | 192 (77%) | 25 (54%) | 13 (68%) | 0.61 |
| Above cut-off | 86 (27%) | 80 (27%) | 59 (23%) | 21 (46%) | 6 (32%) | |
| Comorbidity | ||||||
| None | 180 (57%) | 168 (57%) | 152 (61%) | 16 (35%) | 12 (63%) | 0.85 |
| Either | 83 (26%) | 78 (26%) | 65 (26%) | 13 (28%) | 5 (26%) | |
| Both | 53 (17%) | 51 (17%) | 34 (13%) | 17 (37%) | 2 (11%) | |
* Between follow-up completed vs. lost to follow-up participants in 2016.
Table 2.
Distribution of causes of death by baseline mental health status among the 46 deceased participants. Rows represent baseline mental health groups and columns indicate cause of death. Verbal autopsy was completed for 40 deaths. “No verbal autopsy” (n = 6) indicates participants for whom cause of death could not be determined because a verbal autopsy was not completed.
Table 2.
Distribution of causes of death by baseline mental health status among the 46 deceased participants. Rows represent baseline mental health groups and columns indicate cause of death. Verbal autopsy was completed for 40 deaths. “No verbal autopsy” (n = 6) indicates participants for whom cause of death could not be determined because a verbal autopsy was not completed.
| Baseline Mental Health Group | Chronic Respiratory | Alcohol | Stroke | Accident/ Injury | Cancer | No Verbal Autopsy | Total |
|---|---|---|---|---|---|---|---|
| No anxiety/depression | 7 | 3 | 1 | 0 | 2 | 3 | 16 |
| Depression only | 3 | 3 | 1 | 0 | 0 | 2 | 9 |
| Anxiety only | 2 | 1 | 1 | 0 | 0 | 0 | 4 |
| Comorbid anxiety/depression | 11 | 1 | 2 | 2 | 0 | 1 | 17 |
| Total | 23 | 8 | 5 | 2 | 2 | 6 | 46 |
Table 3.
Association of depression and anxiety with mortality (crude models).
| Characteristic | OR (95% CI) | p Value |
|---|---|---|
| Model 1—Depression (ref below cut-off) | ||
| Above Cut-Off | 3.11 [1.64, 5.98] | <0.001 |
| Model 2—Anxiety (ref below cut-off) | ||
| Above Cut-Off | 2.73 [1.42, 5.24] | 0.002 |
| Model 3—Comorbidity (ref none) | ||
| Either | 1.90 [0.85, 4.17] | 0.11 |
| Both | 4.75 [2.18, 10.44] | <0.001 |
Table 4.
Association of depression and anxiety with mortality (adjusted models).
| Model 4 | Model 5 | Model 6 | ||||
|---|---|---|---|---|---|---|
| Characteristic | OR (95% CI) | p Value | OR (95% CI) | p Value | OR (95% CI) | p Value |
| Depression (ref below cut-off) | ||||||
| Above Cut-Off | 2.34 [1.01, 5.53] | 0.049 | ||||
| Anxiety (ref below cut-off) | ||||||
| Above Cut-Off | 3.49 [1.27, 10.15] | 0.018 | ||||
| Comorbidity (ref none) | ||||||
| Either | 1.65 [0.60, 4.51] | 0.331 | ||||
| Both | 5.27 [1.65, 17.61] | 0.006 | ||||
| Gender (ref male) | ||||||
| Female | 0.30 [0.11, 0.77] | 0.016 | 0.21 [0.07, 0.60] | 0.005 | 0.22 [0.07, 0.61] | 0.005 |
| Ethnicity (ref Brahmin) | ||||||
| Chhetri and Janajati | 1.80 [0.58, 6.44] | 0.330 | 1.82 [0.59, 6.40] | 0.318 | 1.96 [0.63, 7.01] | 0.266 |
| Dalits | 4.97 [1.57, 18.22] | 0.009 | 4.31 [1.35, 15.79] | 0.018 | 4.62 [1.44, 16.96] | 0.014 |
| Age (ref 18–25 years) | ||||||
| 26–34 years | 0.65 [0.08, 4.26] | 0.655 | 0.97 [0.12, 6.47] | 0.972 | 0.70 [0.09, 4.62] | 0.708 |
| 35–44 years | 2.53 [0.58, 13.32] | 0.232 | 3.77 [0.84, 20.68] | 0.095 | 2.78 [0.62, 15.03] | 0.196 |
| 45 years and above | 17.54 [5.30, 80.97] | <0.001 | 26.60 [7.63, 130.72] | <0.001 | 19.89 [5.87, 94.20] | <0.001 |
| Education (ref no education) | ||||||
| Any education | 0.76 [0.31, 1.91] | 0.564 | 0.71 [0.28, 1.77] | 0.461 | 0.73 [0.29, 1.85] | 0.506 |
| Income (ref no income) | ||||||
| Any income | 1.09 [0.48, 2.49] | 0.844 | 1.38 [0.58, 3.38] | 0.476 | 1.47 [0.61, 3.64] | 0.397 |
OR = Odd Ratio, CI = Confidence Interval.
Table 5.
Association of Gender and Ethnicity with Depression and Anxiety at Baseline (n = 297).
| Model 7 | Model 8 | |||
|---|---|---|---|---|
| Characteristic | OR (95% CI) | p Value | OR (95% CI) | p Value |
| Gender (ref male) | ||||
| Female | 2.65 [1.56, 4.56] | <0.001 | 4.11 [2.24, 7.87] | <0.001 |
| Ethnicity (ref Brahmin) | ||||
| Chhetri and Janajati | 0.66 [0.35, 1.23] | 0.191 | 0.38 [0.19, 0.77] | 0.007 |
| Dalits | 2.70 [1.35, 5.53] | 0.005 | 3.63 [1.74, 7.88] | <0.001 |
OR = Odd Ratio, CI = Confidence Interval. Model 7 = Logistic regression model with depression as the outcome and gender and ethnicity as predictors; Model 8 = Logistic regression model with anxiety as the outcome and gender and ethnicity as predictors.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2026 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license.
Share and Cite
MDPI and ACS Style
Rai, S.; Rangel, R.; Calamari, I.; Adhikari, S.B.; Kaiser, B.N.; Acharya, N.R.; Koirala, S.; Kohrt, B.A. Association of Anxiety and Depression with Mortality in Rural Nepal: A 16-Year Longitudinal Community Cohort Study. Psychiatry Int. 2026, 7, 222. https://doi.org/10.3390/psychiatryint7050222
AMA Style
Rai S, Rangel R, Calamari I, Adhikari SB, Kaiser BN, Acharya NR, Koirala S, Kohrt BA. Association of Anxiety and Depression with Mortality in Rural Nepal: A 16-Year Longitudinal Community Cohort Study. Psychiatry International. 2026; 7(5):222. https://doi.org/10.3390/psychiatryint7050222
Chicago/Turabian StyleRai, Sauharda, Ruta Rangel, Isabella Calamari, Safar Bikram Adhikari, Bonnie N. Kaiser, Nanda Raj Acharya, Suraj Koirala, and Brandon A. Kohrt. 2026. "Association of Anxiety and Depression with Mortality in Rural Nepal: A 16-Year Longitudinal Community Cohort Study" Psychiatry International 7, no. 5: 222. https://doi.org/10.3390/psychiatryint7050222
APA StyleRai, S., Rangel, R., Calamari, I., Adhikari, S. B., Kaiser, B. N., Acharya, N. R., Koirala, S., & Kohrt, B. A. (2026). Association of Anxiety and Depression with Mortality in Rural Nepal: A 16-Year Longitudinal Community Cohort Study. Psychiatry International, 7(5), 222. https://doi.org/10.3390/psychiatryint7050222
