Current prevalence and determinants of generalized anxiety disorder from a nationally representative, population-based survey of India

Abstract

Introduction:

Generalized anxiety disorder (GAD) is one of the common anxiety disorders leading to impairment and burden. However, GAD remains the least studied anxiety disorder. There is a need for nationally representative epidemiological data of GAD to understand the current burden and plan the mental health policies and programs to attain their unmet needs. Hence, this study focuses on epidemiology, socio-demographic correlates, disability, and treatment gap of GAD from India's National Mental Health Survey (NMHS) 2016.

Materials and Methods:

NMHS 2016 was a nationally representative epidemiological survey of adult respondents from 12 states of India. NMHS is a multi-stage, stratified, random cluster sampling with random selection based on probability proportional to size at each stage. The Mini-International Neuropsychiatric Interview 6.0.0 used to diagnose psychiatric disorders. Sheehan disability scale was used to assess the disability. The current weighted prevalence of GAD was estimated. Association between GAD and socio-demographic factors was done using Firth's penalized logistic regression. The treatment gap and disability in GAD also calculated.

Results:

The current weighted prevalence of GAD is 0.57%. The male gender and higher education groups have significantly lesser odds with current GAD. Urban metro and the married group have significantly higher odds with current GAD. The most common comorbid psychiatric disorders are depression (15.8%) followed by agoraphobia (9.4%). Among respondents with current GAD in the past 6 months across three domains, around 2/5^th has mild and moderate disability, 1/10^th has a severe disability, and 1/20^th has an extreme disability. The overall treatment gap of current GAD is 75.7%.

Conclusion:

NMHS 2016 has provided valuable insights into the epidemiology and burden of GAD among the general population. The available findings provide a glimpse of the current scenario in GAD to aid policymakers in targeting interventions.

INTRODUCTION

Generalized anxiety disorder (GAD) is characterized by excessive anxiety and worry about several events or activities. It is also associated with other symptoms such as restlessness, irritability, easy fatiguability, lack of concentration, muscle tension, and sleep disturbances. It causes significant impairment in functioning irrespective of comorbid disorders.[1] Though studies on anxiety disorders are gaining interest in the recent past,[2] GAD remains the least studied among them.[3]

The 12-month prevalence of GAD across countries from the general population varies from 1.2-1.9% among European countries[4] to 0.6% in Japan. Epidemiological studies specific to a particular country are very much required in anxiety disorders like GAD as prevalence rates differ from Western countries,[5] and cultural variations exist. The epidemiology of GAD from a representative sample of India has been very scarce. The diagnosis of GAD itself was added only in the diagnostic and statistical manual of mental disorders – 3^rd edition (DSM-III) in 1980. This could have led to notably fewer epidemiological studies. Other reasons could be methodological limitations, the absence of a nationwide survey system, and less sample size. The World Mental Health Survey (WMHS) in 2005 estimated the 12-month prevalence of GAD was 0.34% from 24371 Indian sample size from eight sites.[6] However, we need epidemiology of GAD from a better national representative population after a decade gap. Hence, this study focuses on epidemiology, socio-demographic correlates, disability, and treatment gap of GAD from India's National Mental Health Survey 2016.

MATERIALS AND METHODS

Sample

The study methodology of India's National Mental Health Survey (NMHS) 2016 has been described in earlier articles in detail.[7,8,9] Briefly, the NMHS 2016 was a nationally representative epidemiological survey of respondents aged 18 years and above from 12 states, 2 each in six parts of the country (East, West, North, South, Central, and North-eastern). The study was conducted as a multi-stage, stratified, random cluster sampling with random selection based on probability proportional to size at each stage. The respondents were approached at their households, and informed consent was obtained before conducting face-to-face interviews. All data collection instruments were translated and conducted in local languages. The Institutes Ethics committees from the National Institute of Mental Health and Neurosciences, Bengaluru, and respective participating states approved the protocol.

Measures

Socio-demographic variables

The socio-demographic variables such as age, gender, education, occupation, marital status, and socioeconomic details of the respondents are collected in a specially designed proforma prepared exclusively for NMHS 2016.

Mini-International Neuro-Psychiatric Interview 6.0.0 (MINI 6.0.0)

The computer-generated Mini-International Neuropsychiatric Interview (M.I.N.I 6.0.0),[10] a short structured diagnostic instrument, was used to estimate the prevalence of psychiatric disorders. It has established psychometric properties and has been widely used in estimating anxiety disorders.[11] The GAD module in M.I.N.I 6.0.0 estimates the current prevalence of GAD in the last 6 months. Other disorder modules were used to assess comorbidity with GAD. The M.I.N.I instrument was administered by trained interviewers using tablets, and the time taken was anywhere between 15 and 40 minutes.

Sheehan disability scale (SDS)

The SDS[12] was used to assess functional impairment associated with GAD. The extent to which work, social, and family life are impaired due to GAD was assessed using a 10-point visual analogue scale. The scores were recategorized into no disability, mild, moderate, severe, and extreme disability.

Treatment gap and related variables

The treatment gap was estimated from the number of respondents with GAD who were not on treatment from a formal healthcare provider. In addition, information on the duration of treatment, the time between the onset of illness and treatment, and duration of illness were collected.

Statistical analysis

To ensure that findings are representative of the Indian population, all estimates were weighted to adjust for unequal probability of selection and non-response rates. Descriptive analysis was performed to describe the current weighted prevalence of GAD, its socio-demographic characteristics, comorbidity with other psychiatric disorders, functional impairment, and treatment gap. Modelling the data for binomial outcome [persons with GAD and without GAD] was essential to generate an odds ratio from logistic regression analysis. Since the number of persons with GAD was disproportionate to persons without, it led to the problem of statistical separation or monotone likelihood. Hence, Firth bias reduced logistic regression was chosen in concurrence with the published literature.[13,14,15] This statistical procedure reduces first-order bias in maximum-likelihood estimates of beta and small sample size and provides finite, consistent estimates. P <.05 was kept statistically significant. R software and IBM Statistical Package for Social Sciences (SPSS) version 28 were used to perform statistical analysis.

RESULTS

A total of 34,802 respondents were interviewed, with an individual response rate of 88% among eligible adults. The current weighted prevalence of GAD is 0.57% (95% CI 0.56-0.59) [Table 1] and socio-demographic profile of GAD patientsa are provided in Table 1.

Table 1.

The current weighted prevalence estimates of GAD and socio-demographic characteristics of the GAD

Socio-demographic parameter	n (%)
Total number diagnosed with GAD	173 (0.57)
Age group
18 to 29	42 (27)
30 to 39	36 (22.9)
40 to 49	38 (22.4)
50 to 59	30 (15.8)
60 and above	27 (11.9)
Gender
Male	58 (31.7)
Female	115 (68.3)
Education
Illiterate	62 (31.4)
Primary	44 (26.7)
Secondary	18 (14)
High school	15 (9.5)
Pre-university/vocational	21 (12.1)
Graduate/postgraduate/professional	12 (5.7)
Others*	1 (0.5)
Occupation
Working	78 (38)
Not working	95 (62)
Not known	0
Marital status
Never married	14 (8.7)
Married	144 (84.6)
Widowed/divorced/separated	14 (6.6)
Others*	1 (0.2)
Residence
Rural	94 (47.3)
Urban non-metro	33 (10.1)
Urban metro	46 (42.6)

Among the socio-demographic factors [Table 2], the male gender and higher education groups have significantly lesser odds with Firth Reduced Logistic Regression for current GAD. Urban metro and the married group have significantly higher odds with the current GAD.

Table 2.

Firth penalized logistic regression of socio-demographic correlates and current GAD

	Adjusted OR	95% CI	P
Gender
Male	0.59	0.4-0.9	0.007
Female (ref)
Age
18-29 (ref)
30-39	1.04	0.6-1.7	0.8
40-49	1.15	0.7-1.9	0.5
50-59	1.16	0.7-1.9	0.5
>60	0.8	0.4-1.3	0.4
Education
Illiterate (ref)
Primary	0.97	0.6-1.4	0.8
Secondary	0.43	0.2-0.7	0.001
High school	0.31	0.1-0.5	<0.001
Pre-university	0.82	0.5-1.4	0.5
Graduate	0.38	0.1-0.7	0.002
Occupation
Employed (ref)
Unemployed	0.87	0.6-1.3	0.5
Marital status
Single (ref)
Married	1.92	1.1-3.7	0.02
Divorced/separated	1.87	0.8-4.4	0.15
Residence
Rural (ref)
Urban non-metro	1.48	1-2.2	0.06
Urban metro	3.45	2.4-5	<0.001

Table 3 shows the comorbidity of current GAD with other psychiatric disorders. The most common comorbid psychiatric disorder of GAD is depression (15.8%) followed by agoraphobia (9.4%).

Table 3.

Comorbidity of current GAD with other psychiatric disorders

Comorbidity with GAD	n (%)
Panic disorder	14 (5)
Social phobia	6 (2.3)
Agoraphobia	24 (9.4)
Obsessive Compulsive Disorder	16 (8.3)
Post-Traumatic Stress Disorder	6 (2.7)
Depression	36 (15.8)
Bipolar disorder	2 (1.1)
Alcohol use disorder	14 (5.6)

Table 4 shows the severity of disability of GAD. Among respondents with current GAD across three domains, around 2/5^th had mild and moderate disability, 1/10^th had severe disability, and 1/20^th had an extreme disability.

Table 4.

Severity of disability of current GAD

Self-reported disability (n=170*)	Any disability n (%)				No disability n (%)
	Mild disability	Moderate disability	Severe disability	Extreme disability
Disability at work	74 (43.5)	35 (20.6)	23 (13.5)	9 (5.3)	29 (17.1)
Disability at social life	64 (37.6)	41 (24.1)	21 (12.4)	7 (4.1)	37 (21.8)
Disability at family life	57 (33.5)	56 (32.9)	19 (11.2)	9 (5.3)	29 (17.1)

*Missing data (3)

Table 5 provides the treatment gap and care characteristics of GAD population. The overall treatment gap of current GAD is 75.7%. The treatment gap in females and males is 77.4% and 72.4%, respectively. The treatment gap in urban metro (67.40%) is lesser compared to rural (78.7%) and urban non-metros (78.8%).

Table 5.

Treatment gap and care characteristics in respondents with current GAD

Treatment gap and care characteristics	Overall
Treatment gap	75.70%
The median duration of illness in months	48
Median interval between onset of illness and consultation in months	10
Median duration of being on treatment in months	36

DISCUSSION

Prevalence rate and comparison

This is the first largest epidemiological study from India from a nationally representative sample with a good response rate[16] (88%). The current weighted prevalence of GAD is 0.57% (95% CI 0.56-0.59). This figure is higher than the 12-month prevalence rates of GAD in Indian World Mental health Survey (WMHS) conducted in 2005, a decade earlier (0.34%). Despite being the current prevalence (6 months of prevalence), the higher rates showed that the prevalence of GAD has increased over the decade. In comparison with other Asian countries, the rates are similar to the 12-month prevalence rates of GAD in Japan[17] (0.6%), and South Korea[18] (0.8%). Other national surveys, such as Singapore mental health study,[19] could not be compared as the estimates were in lifetime prevalence. However, the prevalence rates of GAD in Western countries have been observed to be consistently higher.[20] Whether these differences reflect true difference in prevalence or cross-cultural variations in the presentation of anxiety disorders remains a debate.[21] This is similar to epidemiological differences noted in other anxiety disorders, which calls for a Determinants of Outcome of Severe Mental Disorders (DOSMED) kind of multinational study to understand these cross-national differences.[22]

Socio-demographic correlates

Males have lesser odds of having current GAD even after controlling for other socio-demographic variables. This has been observed in other epidemiological surveys as well.[23] Among education, respondents with higher educational status have lesser odds of having a current GAD. Married people had higher odds of having current GAD [Table 2]. This finding contrasts with other studies that have found single and separated people to have increased risk.[23] Also, our study found that respondents residing in urban metros had increased odds of having a current GAD. The impact of urbanization on the increasing prevalence of psychiatric disorders, especially depressive and anxiety disorders, has been well documented in the literature.[24] However, from prospective studies, none of the socio-demographic variables seems to predict the course of GAD.[25]

Comorbid psychiatric disorders

This study found that depressive disorder (15.8%) is the common comorbid psychiatric disorder with current GAD. This has been time-tested in various studies that depressive disorders and GAD at times look inseparable.[26] The presence of depression can complicate the course and outcome of GAD.[27] Among other anxiety disorders, agoraphobia is the most common, followed by panic disorder (5%). Comorbidity among anxiety disorders has been a rule rather than an exception.

Disability and treatment gap in GAD

This study found that around 80% of GAD have a disability similar to the disability estimated in Singapore Mental Health Survey (around 90%).[18] However, disability in GAD can be transient and improve once adequate treatment is given. Alarmingly, the treatment gap in GAD is comparable with that of severe mental illness (73.6% in severe mental disorders).[9] The overall treatment gap is 75.7%, higher in rural and urban non-metros, showing the prime target areas for intervention. This points toward the scaling up of efforts to promote help-seeking in individuals with mental health concerns.[28] This huge treatment gap in one of the common mental disorders points out the current lag in mental health interventions reaching out to the community.

CONCLUSION

The current study estimated the current prevalence of GAD to be 0.57%. Depression and agoraphobia are most often comorbid with GAD. A majority have disabilities due to GAD (80%) and the treatment gap is 75.7%, similar to severe mental illnesses. The results suggest the need of appropriate policies, programs, and service developments to reduce the treatment gap and reverse the disability due to GAD.

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 the six regions and interviewed 39,532 individuals (http://indianmhs.nimhans.ac.in). Funder had no role in implementation, data acquisition, data analysis, and interpretation and write up of the study.

Conflicts of interest

There are no conflicts of interest.