District-level differences in the distribution of mental and substance use disorders in Assam Authors: Vijay Gogoi, Priyadarshee Abhishek, Soumik Sengupta, Indrajeet Banerjee, Samir Sarma, H. Sobhana, Sonia P. Deuri, Kangkan Pathak, Gautham M. Sukumar, Vivek Benegal, Girish N. Rao, Mathew Varghese, Gopalkrishna Gururaj Published: 2023 ABSTRACT Background:The lifetime prevalence of mental morbidity in Assam is estimated at 8% (NMHS 2015–16). Understanding the distribution patterns of different types of mental disorders among persons with mental morbidity in different districts would facilitate evidence-driven district mental health programming in Assam. Given the varied socio-geopolitical situation across districts in Assam, significant variations in the distribution of mental disorders are expected.Aims:To assess interdistrict differentials in common mental disorders (CMDs), severe mental disorders (SMDs), socioeconomic impact, healthcare utilization, and mental disability across three districts sampled in NMHS in Assam.Materials and Methods:This cross-sectional study used stratified random cluster sampling to identify and study eligible adult participants in Dibrugarh, Barpeta, and Cachar districts. Standardized scales and validated questionnaires were used to assess mental morbidity, disability, socioeconomic impact, and healthcare utilization. The distribution of different mental disorders among persons with mental disorders and their interdistrict differentials were tested using the Chi-square test of significance.Results:Among persons with mental morbidity, the most common disorder was CMDs (79%). The proportional distribution of CMDs among persons with mental morbidity was significantly higher in the Dibrugarh district (79%), whereas the distribution of SMDs was higher in the Cachar district (55%). The distribution of alcohol use disorder was the highest in the Dibrugarh district (71.6%). Significant differences in disability and healthcare utilization were observed between the districts.Conclusions:NMHS 2015–16 Assam indicates significant differentials in the distribution of CMDs and SMDs, healthcare utilization, and associated disability between the three districts. The differentials necessitate further research to understand socio-ethnocultural, religious, geopolitical, and other factors influencing the distribution. These differences need to be accounted for during the implementation of mental health programs in the state. INTRODUCTION Mental health problems are widespread and significantly affect individuals, families, and the community. However, the lack of epidemiological data and knowledge on the burden of mental health problems and the lack of availability of resources to handle them poses a significant threat to the effective implementation of policies and programs. There are few psychiatric epidemiological studies available, and these studies report varying prevalence rates, have inadequate sample sizes, and have differing methodologies, limiting their representativeness at the national level. The National Mental Health Survey of India (NMHS), 2015–16, conducted in 12 states of India, was conceptualized to represent the national population.[1] NMHS focused on estimating the prevalence and pattern, treatment gap, service utilization, socioeconomic impact, and disability of mental disorders. The weighted lifetime and current prevalence of any mental morbidity at the national level were 13.67% and 10.56%, respectively.[2] Among the 12 states surveyed, Assam and Manipur represented the North-Eastern region. The weighted lifetime and current prevalence of any mental morbidity in the state of Assam were 8.01% and 5.85%, respectively. Assam, known for its tea gardens and silk, connects the other North-Eastern states to mainland India. It has a unique and varying demography due to its multiethnicity and long-standing immigration. In a population of approximately 31 million in 78,500 km2, there are around 115 identified ethnic groups[3] with a decadal growth rate of 17.07%. The North-Eastern region of India is often resource-strapped economically as well as in mental health personnel and services. Socioeconomic status is known to influence the utilization of mental health services.[4] Delay in seeking treatment and inconsistency or lack of utilization of mental health services slows recovery and can result in long-term disability. Against the backdrop of a poor socioeconomic context with few psychiatric services available, this paper aims to explore the distribution of mental morbidity, socioeconomic status, health utilization pattern, and disability among persons with mental illness in three selected districts of Assam. A brief overview of these districts, namely Dibrugarh, Barpeta, and Cachar, in the Statistical Handbook of Assam, 2016, shows the demographic differences in population density,[5] religion, and percentage of the urban and tribal population [Table 1]. Table 1Demographic differences in the three districtsCharacteristicsDistrictsBarpetaDibrugarhCacharPopulation Density742392459Decadal Growth rate21.4311.9220.19Urban Population (%)8.7018.3818.17Hindu (%)29.1190.3559.83Muslim (%)70.744.8637.71Scheduled tribes (%)0.090.310.06 Demographic differences in the three districts Dibrugarh has a higher representation of Hindu population, while Cachar and Barpeta have a primarily Muslim population. Dibrugarh also has a higher representation of tea tribes and other tribal populations. The population density and decadal growth rate are higher in Barpeta and Cachar. Therefore, these three districts vary in socioeconomic and demographic backgrounds, indicating their ethnocultural and religious variations. In this context, it is essential to understand differences in the distribution of mental morbidities between the districts and explore the role of macro-level determinants for the prevalence of mental morbidity. The present study, a part of the larger NMHS, 2015–16, assesses the interdistrict differentials in the distribution of common mental disorders (CMDs) and severe mental disorders (SMDs) among persons with mental morbidity. It also aims to understand differences in socioeconomic impact, healthcare utilization, and associated disability in the three selected districts of Assam. MATERIALS AND METHODS The study was conducted by Lokopriya Gopinath Bordoloi Regional Institute of Mental Health (LGBRIMH) in Assam as part of NMHS, India, 2015–16, coordinated by the National Institute of Mental Health and Neurosciences (NIMHANS), Bengaluru. Detailed methodology of NMHS was published by Pradeep et al. in 2016,[6] and a brief Assam-specific description is provided in this paper. Study design This multisite, community-based, cross-sectional study used a multistage, stratified, random cluster sampling technique, with random selection based on probability proportion to size at each stage (MSRS-PPS). Sampling procedure Four-stage sampling was adopted (District › Taluka/CDB › Village/Ward › Household). At the time of planning the study, Assam had 27 districts, which later increased after August 15, 2015. All 27 districts were stratified, rank-ordered, and trifurcated based on poverty estimates. One district was randomly selected from the three strata. Using the PPS strategy, two talukas were randomly identified in each district, followed by urban and rural clusters within the selected talukas, proportional to the state urban–rural proportions. The selected districts and talukas were Dibrugarh (Dibrugarh East, Tengakhat), Barpeta (Kalgachia, Bajali), and Cachar (Katigora, Silchar). Initially, Kokrajhar was selected for the survey; however, due to the prevailing ethnic clashes and social tension, Kokrajhar was replaced by Barpeta.[7] Households within each cluster, the final sampling unit, were selected using systematic random sampling. Based on the prevalence rates of 7.5% for any mental morbidity and a non-response rate of 30% in the Kolar pilot study conducted in 2014, the final sample size was calculated as 3000 individuals above the age of 18 years. All eligible individuals within the selected households were interviewed. Those not available even after three visits were considered non-responders. Study tools The study tools included a sociodemographic questionnaire adapted from the household questionnaire of the Census of India 2011, MINI 6.0 for the assessment of mental morbidity, Fagerstrom questionnaire for tobacco use, WHO-SEARO screening questionnaire for epilepsy, a brief screening instrument developed by NIMHANS for intellectual disability, and Sheehan's disability scale. The healthcare-seeking behavior and utilization pattern were assessed by questions adapted from the WHO-Pathways Interview Schedule, and the socioeconomic impact was assessed using modified WHO DAS 2.0. Assamese, Bengali, and Hindi-translated versions of the tools were loaded into tablets used for data collection. Focused group discussions and key informant interviews were also conducted with community members, healthcare providers, and others to understand the treatment gap in the state adequately. Training and quality control Eight field data collectors (FDCs) were recruited based on mental health and community survey experience, fluency in reading, and communication in Assamese, Bengali, and Hindi. Four FDCs had an M.Phil. degree in Psychiatric Social Work, while the rest had a Masters degree in Social Work. The FDCs were extensively trained using a uniform training schedule for 8 weeks. By the end of the training, each FDC had observed nearly 80 interviews and conducted 40 interviews. Refresher training was also provided midway after the completion of data collection in 30 clusters. The primary objective of the training was to ensure that all data collectors understood the requirements and components of data collection, similarly administered the study tools, and elicited an unbiased response. In addition, daily and weekly monitoring by the supervisor and review meetings were held to keep track of the progress and troubleshoot problems. Finally, investigators did six supervisory field visits and 115 re-interviews to monitor the quality of the interviews. Statistical analysis We specifically looked for differential distribution in CMDs, SMDs, alcohol use disorders, tobacco use disorders, other substance use disorders, and suicide risk between the three districts (Barpeta, Dibrugarh, and Cachar) selected for the NMHS survey. The Chi-square test of significance was applied to test for significant interdistrict differentials in the proportion of studied variables mentioned earlier (significant at P < 0.05). The interdistrict percentages are shown row-wise in the tables; hence, the mentioned percentages were computed out of the total number of absent and present cases, respectively. Data were analyzed using IBM SPSS Statistics for Windows version 25.0 (IBM Corp., Armonk, New York, USA). Ethical clearance for the study was obtained from the Ethics Committee of LGBRIMH. RESULTS Sociodemographic characteristics A total of 2603 individuals were interviewed in the survey. The sociodemographic characteristics of the three selected districts (Barpeta, Dibrugarh, and Cachar) are depicted in Table 2. Overall, 212 (8.14%) individuals out of 2603 were found to have some form of mental morbidity across all three districts. Table 2Sociodemographic characteristics of the sample in the three districts (Assam)CharacteristicDMHPDistrictsBarpetaNoDibrugarhNoCacharNo n % n % n %Gender of Respondents Male37042.8%37943.4%37042.8% Female49557.2%49456.6%49557.2%Age Group (in years) 18–2927231.4%32937.7%31035.8% 30–3919522.5%19322.1%18721.6% 40–4914416.6%15317.5%15117.5% 50–5912214.1%10512.0%839.6% 60 and above13215.3%9310.7%13415.5%Education Illiterate20023.1%26930.8%18821.7% Primary9811.3%11813.5%19722.8% Secondary13115.1%16518.9%14717.0% High School18921.8%16118.4%21625.0% Pre-University/Vocational13015.0%799.0%839.6% Graduate/Post-Grad/Professional11713.5%758.6%343.9% Not Applicable/NK0060.7%00Occupation Working28833.3%44050.4%31136.0% Not Working57065.9%41547.5%54663.1% Others70.8%182.1%80.9%Marital Status Never Married13916.1%16318.7%17119.8% Married69880.7%64974.3%65575.7% Widowed/Divorced/Separated111.3%222.5%242.8% Others172.0%394.5%151.7%BPL Card Availability50758.6%62972.1%57666.6%Major Sources of Treatment Public Sources71282.3%59880.0%60870.3% Private Sources14616.9%16619.0%20223.4% Others70.8%91.0%556.4% Sociodemographic characteristics of the sample in the three districts (Assam) Common mental disorders Lifetime CMDs accounted for 79% (n = 169) of all persons with mental morbidity (n = 212) across the three districts [Table 3]. The proportional distribution of lifetime CMDs among persons with mental morbidity was significantly higher in Dibrugarh district (55%) as against Barpeta (21.9%) and Cachar (23.1%) districts. The current CMDs accounted for 65.57% of mental morbidity and varied significantly between the three districts. Dibrugarh (58.3%) had the highest proportional distribution of current CMDs, followed by Cachar (22.3%) and Barpeta (19.4%). Hence, among persons identified to have mental morbidity in Assam, proportion of lifetime CMDs was the highest in Dibrugarh and lowest in Barpeta. Table 3District-wise distribution of common and severe mental disorders, substance use disorders, and suicide risk among persons with mental morbidityMental morbidityStatusDistrictsTotal Cases^ (n=212)X2 (df)BarpetaDibrugarhCacharCommon Mental Disorders CMD_LifetimePresent37 (21.9%)93 (55.0%)39 (23.1%)169 (79.72%)23.81 (2)* CMD_CurrentPresent27 (19.4%)81 (58.3%)31 (22.3%)139 (65.57%)20.39 (2)*Severe Mental Disorders SMD_LifetimePresent12 (26.1%)10 (21.7%)24 (52.2%)46 (21.69%)18.25 (2)* SMD_CurrentPresent2 (11.1%)6 (33.3%)10 (55.6%)18 (8.49%)6.44 (2)†Substance Use Disorders Alcohol Use DisorderPresent8 (9.9%)58 (71.6%)15 (18.5%)81 (3.11%)55.29 (2)* Tobacco Use DisorderPresent182 25.2%)177 (24.5%)364 (50.3%)723 (27.77%)132.28 (2)* Other Substance UsePresent3 (60.0%)2 (40.0%)0 (0%)5 (0.19%)2.78 (FE)Suicide Risk LowPresent25 (23.8%)66 (62.9%)14 (13.3%)105 (4.04%)43.87 (2)* ModeratePresent3 (17.6%)7 (41.2%)7 (41.2%)17 (0.65%)1.87 (2) HighPresent2 (10.0%)14 (70.0%)4 (20.0%)20 (0.7%)12.28 (2)†CMD=Common Mental Disorders, SMD=Severe Mental Disorders, FE=Fisher’s exact. *=P<0.001; †= P<0.01; ^In the column titled ‘total cases’ percentages provided are column percentages District-wise distribution of common and severe mental disorders, substance use disorders, and suicide risk among persons with mental morbidity CMD=Common Mental Disorders, SMD=Severe Mental Disorders, FE=Fisher’s exact. *=P<0.001; †= P<0.01; ^In the column titled ‘total cases’ percentages provided are column percentages Severe mental disorders The proportion of lifetime and current SMDs among persons with mental morbidity was 21.69% (n = 46) and 18 (8.49%), respectively [Table 3]. The distribution of SMDs varied across the three districts. Proportion of lifetime and current SMDs among persons with mental morbidity was higher in Cachar (lifetime = 52.2%; current = 55.6%) than in Dibrugarh and Barpeta [Table 3]. Substance use disorder This section describes substance use disorders among the individuals found eligible for the study (N = 2603). Around 3% of persons with mental morbidity had alcohol use disorder (AUD), of which nearly 71.6% was distributed in Dibrugarh district, significantly higher than that in Cachar (18.5%) and Barpeta (9.9%) [Table 3]. However, the distribution of tobacco use disorder was significantly higher in Cachar (50.3%) than in Dibrugarh (24.5%) and Barpeta (25.2%). Five participants were found positive for other substance use (0.19%) across all the districts. Suicide risk For suicide risk (N = 2599), a considerable number of individuals (4.04%) were found to be at low risk for suicide. The distribution of high and moderate risk for suicide was higher in Dibrugarh district (70% of total cases with high suicide risk and 41% of cases with moderate risk). The distribution of low, moderate, and high risk for suicide significantly varied across the three districts [Table 3]. Epilepsy and intellectual disability Epilepsy and intellectual disability were found in four and nine individuals out of 2603 participants, respectively. All the epilepsy cases were found only in the Dibrugarh district, whereas cases with intellectual disability were found in all the districts [Barpeta: 4 (44.44%); Dibrugarh: 2 (22.22%); Cachar: 3 (33.33%)]. Disability with mental morbidity A total of 173 participants were found to have responded to disability-related questions [Table 4]. Dibrugarh had the highest number of individuals experiencing mild (54.17%) and moderate-to-severe (42.86%) work-related disability. Similarly, a greater number of individuals from Dibrugarh reported moderate-to-severe levels of disability in social life (44%). In addition, Dibrugarh had the highest number of cases of disability related to home and family life. Table 4Disability levels across the three districts (n=173, missing 39)DisabilityBarpetaDibrugarhCacharTotal casesAt work or school No disability25 (20.66%)64 (52.89%)32 (26.44%)121 (69.94%) Mild6 (25.00%)13 (54.17%)5 (20.83%)24 (13.87%) Moderate to severe10 (35.71%)12 (42.86%)6 (21.43%)28 (16.18%)In Social Life No disability26 (20.80%)67 (53.60%)32 (25.60%)125 (72.25%) Mild10 (40.00%)11 (44.00%)4 (16.00%)25 (14.45%) Moderate to severe5 (21.74%)11 (47.83%)7 (30.43%)23 (13.29%)At home and in family life No disability24 (20.87%)59 (51.30%)32 (27.83%)115 (66.47%) Mild12 (41.38%)14 (48.27%)3 (11.11%)29 (16.76%) Moderate to severe5 (17.24%)16 (55.17%)8 (27.59%)29 (16.76%) Disability levels across the three districts (n=173, missing 39) Healthcare utilization As the data was significantly less for interdistrict comparison, we present comprehensive data for healthcare utilization in the state. The minimum duration between the onset of symptoms to formal consultation was 9.44 months (SD = 16.29, range = 47 months). The range of treatment providers between the first consultation and the current treatment consultation was 9, with a mean of 2.45 (SD = 2.21). Most respondents approached a local faith healer or a primary care doctor for their mental health problems. The average distance to seek care was 14.88 km (SD = 20.39, range = 79 km). People spent an average of ₹1545.28 (SD = 2389.91, range = 9985) during one visit to a healthcare provider. Socioeconomic impact The average number of days in the past month during which respondents with mental morbidity faced difficulties due to health conditions was 15.84 days (SD = 10.82 days; range = 28 days). The average number of days in the past 1 month during which patients were completely unable to work or perform their usual activities was 15.83 days (SD = 11.50, range = 29 days), whereas the average number of days during which reduction in usual activities or work was experienced was 18.50 days (SD = 10.89, range = 26 days). Five respondents reported that the average number of days during which family members were unable to go to work was 29.80 days (SD = 35.003, range = 86 days). Patients missed an average of 48.93 days (SD = 84.45, range = 299 days) of family, social, and leisure activities because of illness. The average money spent on a patient's mental illness was ₹2381.25 (SD = 4849.74; range = 19,950). DISCUSSION The NMHS provided a unique opportunity to explore the pattern of mental morbidity and other crucial factors (healthcare utilization pattern, socioeconomic impact, and disability) in Assam.[15] The careful selection of staff, followed by extensive training, reinforcement training, and regular monitoring, contributed to a valid and reliable estimation of mental morbidity.[7] The current paper explored the differences in the three districts selected for the NMHS, considering their variability in the demographics and socioeconomic conditions. Mental morbidity Out of the 2603 individuals screened, 212 were found with some form of mental morbidity. District-wise, CMD (lifetime and current) cases were the highest in Dibrugarh, whereas Cachar had a higher distribution of SMDs (lifetime and current). As determinants, severity and care strategies differ for CMDs and SMDs; as such, it is imperative to understand the reasons for this peculiar distribution. Furthermore, as the two districts differ in socio-economic-cultural variables, including the migrant population, district-specific strategies may be required for need-based mental health programming. Income quintile and mental morbidity across the three districts Previously, poor socioeconomic factors have been consistently reported as significant predictors of CMDs.[8910] Economic inequality is important, with lower income being a major risk factor for stress resulting in CMDs such as depression and anxiety. The NMHS findings for Assam also reflect that mental morbidity (both lifetime and current) was the highest in the lowest income quintile. Dibrugarh had a greater number of cases in the lowest (52.63%) and second-lowest (50.00%) income quintiles [Table 5]. This finding partly explains why a greater number of cases with some form of mental morbidity was seen in Dibrugarh. However, the relationship between low income and CMDs is equivocal, and lower income possibly contributes to mental illness through various mediating factors such as education, social class, food insecurity, and similar variables.[11] Nevertheless, these findings have significant implications for policymakers and clinicians in the state. It requires policymakers to consider the mental health consequences of income inequality and highlight the need for social initiatives that can somewhat buffer low-income people's effects. Furthermore, as Compton and Shim[9] suggested, mental healthcare providers should consider income status while treating patients and take appropriate steps to link people with relevant social services. Table 5Household income quintile distribution across the three districtsIncome quintileDistrictTotal casesBarpetaDibrugarhCacharLowest6 (10.53%)30 (52.63%)21 (36.84%)57Second7 (17.50%)20 (50.00%)13 (32.50%)40Middle8 (30.77%)14 (53.85%)4 (15.38%)26Fourth11 (26.83%)19 (46.34%)11 (26.83%)41Highest17 (35.41%)17 (35.41%)14 (29.17%)48 Household income quintile distribution across the three districts Substance use Alcohol use disorder across the three districts was 3.1%, with Dibrugarh reporting the highest percentage of alcohol use disorder. The distribution of tobacco users was higher in Cachar compared to Barpeta and Dibrugarh. Among the three districts, Dibrugarh has a higher representation of tea tribes, with 177 tea gardens.[12] The prevalence of alcohol use disorder varied from 32.2% to 59.2% in earlier epidemiological studies in tea tribes.[13] In addition, brewing and consumption of alcohol are culturally accepted in this population; hence, these factors could be attributed to the higher percentage of alcohol use disorder in Dibrugarh compared to the other two districts. The variation in tobacco use could be linked to the religious affiliations of the participants. Barpeta and Cachar have more participants belonging to the Islam religion, which prohibits alcohol use, while the use of tobacco has a cultural sanction. Suicide risk The NMHS assessed the risk for suicide in the past month through direct interviews based on the MINI questionnaire. High and moderate risks for suicide were 0.75% and 0.6%, respectively. Furthermore, the risk was higher in the age group of 18–29 years, with a prevalence of 1.12% (1.00–1.23 at 95% CI). Males (0.93%) had a considerably higher risk than females (0.51%). The district-wise data suggest that Dibrugarh had significantly more respondents (n = 87) with low, moderate, or high risk for suicide [Table 3]. This could partly be explained by the fact that Dibrugarh had a higher percentage of illiterate individuals, widowed/separated/divorced, alcohol use, and CMDs. These correlates of suicide risk were also observed in the national representative population of the NMHS[14] and highlighted in the policy paper on national suicide prevention strategy in India.[15] Some of the leading psychosocial reasons mentioned in the 2020 NCRB data for suicide in Assam were family problems, mental illness, love affairs, poverty, and unemployment, apart from several other factors. Disability in mental morbidity Disability was measured in terms of disability at work or school, social life, family life, and at home. In all three domains, disability was the highest in Dibrugarh [Table 4] in comparison to the other two districts. CMDs and alcohol use disorders were higher in Dibrugarh. Given that CMDs and AUDs are associated with disability, data available from Dibrugarh indicates higher disability levels among the participants. Treatment gap Mental health services are integrated with primary health care services in India.[216] Arguably, this integration is expected to reduce mental health stigma and improve care accessibility cost-effectively.[17] However, the assessment of healthcare utilization, treatment gap, and morbidity between the three districts indicates the need for integrating mental health services to be more meaningful, functional, and tangible at the district level. Furthermore, the NMHS showed that the availability of mental health professionals per one lakh population during the study period was alarmingly poor in the state. This paper brings out the differential distribution of mental disorders in three districts of Assam and recommends further investigation to identify reasons for this distribution. The reasons could range from differentials in macro-level factors (literacy, economy, occupation, culture, migration, and others) to differences in the districts' mental health systems and services. As the district is the unit of mental health program implementation, our study indicates a requirement for need-based, district-specific strategies in India's District Mental Health Program, especially in vulnerable districts. Limitations and future directions Though the survey had uniform study instruments, quality assessments, and monitoring, there was limited scope for the inclusion of additional questionnaires. Gathering information on migration, tea plantation workers, social security measures, ethnicity, and other variables that could influence the distribution of mental morbidities would have been helpful. Considering the ethnic diversity of Assam, the scope of exploring the cultural influences on mental morbidity in the context of varying cultural practices in the North-East region was limited in this study. There is no adequate information on the type of tobacco use (smoking and smokeless use). Having access to information about the specific methods of tobacco consumption could have contributed to a better grasp of the patterns associated with tobacco use. In the context of the absence of district mental health programs in these districts, the findings of this study have implications for mental health services and their utilization. Early identification and treatment can prevent or reduce disability. Financial support and sponsorship The National Mental Health Survey (NMHS) was funded by the Ministry of Health and Family Welfare, Government of India and was implemented and coordinated by the National Institute of Mental Health and Neurosciences (NIMHANS), Bengaluru, INDIA in collaboration with state partners. NMHS phase 1 (2015–16) was undertaken in 12 states of India across six regions and interviewed 39,532 individuals (http://indianmhs.nimhans.ac.in). Funder had no role in the implementation, data acquisition, data analysis and interpretation, and write-up of the paper. Conflicts of interest There are no conflicts of interest.