Background
Anxiety and depression embrace a range of mental conditions occurring frequently in primary health care, usually in the form of overt disorders. They are acknowledged as common mental disorders (CMDs) [
1,
2] and, in terms of their ubiquity and the burdens they impose, as major disorders of the brain [
3]. In the recent Global Burden of Disease Study 2013 (GBD 2013), while mental and substance use disorders collectively accounted for 21.2 % of all years of life lost to disability (YLDs) [
4], depression and anxiety were ranked second and ninth highest specific causes of YLDs in both developed and developing countries. These disorders are therefore of considerable public-health importance [
5] in high-income [
6,
7] and low- and middle-income (LAMI) countries [
8] alike. They are also highly comorbid [
9].
Extrapolations from GBD 2013 data indicate that depression and anxiety are among the top ten causes of YLDs in South Asia, which includes Nepal [
4]; however, no research has been undertaken to make direct national estimates in this Region. Available studies were limited to a few scattered health centres [
10,
11], districts [
12‐
14], villages [
15] or cities [
16], and could not comprehensively describe the prevalence of CMDs, let alone provide an account of the burdens attributable to them.
With these factors under consideration, our principal aim was to assess health-care needs by estimating the prevalences of anxiety and depression in Nepal, using a well-validated and culturally-adapted screening instrument [
17]. Also, we wished to establish their degree of comorbidity, as well as their associations with sociodemographic characteristics, social behaviours and health-related factors. This would fill current knowledge gaps regarding these CMDs in the South Asian Region, and serve our overall purpose of guiding public-health policy towards better mental health in Nepal.
There were methodological issues to be considered. Most earlier South Asian studies utilized self-report scales with caseness of anxiety [
10,
12] or depression [
13‐
15] being indicated by summed scores at or above defined thresholds. Such scales are useful in areas with inadequate or unevenly distributed resources that greatly limit epidemiological surveys [
18]. This was certainly true of Nepal, one of the poorest nations of the world [
19]. Furthermore, its geographical and sociocultural diversities posed unique logistic and methodological difficulties which our survey had to overcome [
20,
21]. We considered important the relationships between mental wellbeing and behaviours typical of life and culture in the Nepalese community [
20], such as the use of alcohol and marijuana during festivals, and the common method of carrying heavy loads on the back, suspended by a tumpline around the forehead. We also considered personality traits associated with psychopathology (neuroticism), and measures of burden in the face of hardship (“life toughness”), including impairment of quality of life (QoL), since these might be pertinent. In selecting covariates for analysis of associations, we had no previous research to draw upon. However, the data were gathered in the context of a nationwide survey of headache disorders [
22,
23], which incorporated a range of demographic, environmental and health variables.
Discussion
HADS-A was more prevalent among the Nepalese than HADS-D, while the two conditions were highly comorbid with each other. HADS-cAD showed significant associations with widowhood, urban and high-altitude dwelling and neuroticism. HADS-A, like HADS-cAD, was more prevalent among urban dwellers. All three types of caseness were associated with poorer QoL. However, comorbid cases containing all elements of both anxiety and depression were associated, more than cases of HADS-A or HADS-D only, with life complications such as those of urban or high-altitude dwelling, and widowhood. We say more about these later.
Owing to the high comorbidity between anxiety and depression, psychiatric research tends to report their combined prevalence: one recent review found the collective worldwide 1-year prevalence of these disorders to be almost 20 % [
5]. However, other recent global reviews have revealed prevalences separately of depression in the range 4.4–5.0 % [
33] and anxiety in the range 4.8–10.9 % [
34]. Our finding for HADS-D (5.2 %) was at the upper limit of the global range, but for depression (
ie, adding those cases included among HADS-cAD), at 11.7 % (95 % CI: 10.3–13.1), it was more than double. Our finding of 16.2 % for HADS-A was already well above the global range, and for anxiety (adding those included in HADS-cAD), at 22.7 % (95 % CI: 20.9–24.5), it was again more than double. These findings are in keeping with the WHO Mental Health (WMH) survey [
35], which showed anxiety disorders to be the most prevalent of all mental disorders, but suggest that, in Nepal, both depression and anxiety are excessively prevalent.
We have exercised caution here, in using the word “suggest”. It is the case that most of these reviews, as well as cross-national epidemiological studies [
36], found both depression and anxiety to be more prevalent in the Western world than in less developed regions such as South Asia. While genetic, sociocultural, environmental and other factors might contribute to real differences, there are important methodological factors to consider that influence prevalence estimates. Commonly these relate to sampling methods, but of specific concern here are the instruments used. Most Western studies utilized diagnostic interviews, while surveys in the less-affluent world used symptom-based scales to screen for psychiatric caseness. Accordingly, HADS, which we used, is a screening instrument for estimating the point prevalences of anxiety and depression, and as such it has limitations. It detects the subjective manifestations of anxiety and depression [
24], while vegetative or somatic symptoms of distress forming parts of the diagnostic classifications (DSM [
37] or ICD [
38]) may not be sufficiently captured. Surveys dependent on HADS and similar instruments may therefore
underestimate actual prevalences, as has been discussed both in the review on CMDs [
5] and in a WMH survey from China [
39].
The Chinese study [
39] emphasized the relevance of sociocultural protective factors (family structure, neighbourhood), which are believed to play a buffering role in most Asian countries [
5], including Nepal, against the distress associated with anxiety or depression. In a different vein, stigma associated with the widespread belief that disclosure of mental illness might lead to embarrassment and discrimination is more common in underdeveloped societies, and may contribute towards underreporting of mental as opposed to physical symptoms; this too was evidenced in a WMH multicentre study [
40]. Because these issues were likely to apply to our study, our findings of excessively prevalent depression and anxiety in Nepal appear even more remarkable since they were unlikely to be overestimates.
From the public-health perspective, the importance of their very high prevalences lies in the associations of both depression and anxiety with substantial disability. There is a wealth of evidence of this, including the data from GBD 2013 [
4]: globally, major depressive disorders are the 2
nd highest cause of YLDs (51.8 million per year), and dysthymia, which is also expected to be captured by HADS-Depression, is 16
th (another 9.8 million YLDs per year); anxiety disorders are 9
th (24.4 million YLDs per year). These GBD estimates are based on the global mean prevalences—well below those we have found in Nepal. In other words, the disability these disorders give rise to globally [
4], great though it is, may not at population level match that in Nepal. Important also are our findings that all of HADS-A, HADS-D and HADS-cAD were associated with low QoL and high neuroticism, illustrations of their major effects on functioning at individual level.
With regard to associations, damage to family or social functioning is linked to mental ill health [
41]. As evidence of this, we found HADS-cAD to be more prevalent among widows. Similarly, a cross-national survey [
36], as well as two Asian studies—one from Iran [
16] and one from China [
42]—showed a high prevalence of CMDs among the widowed. Beyond the stress precipitated by a major family life event, widowhood entails substantial deviance in the societal role as well as in self-perception: widows perceive a lack of social support compared with those who are married [
43].
Society is made up of households that are characteristic of the habitation where they stand. We found substantial associations between mental health and the location of the home: HADS-A and HADS-cAD were significantly more prevalent among the urban population, as was seen for the anxiety disorders in one of the global reviews [
34]. Similarly, in India [
44], multiple effects of unplanned urbanization including fast population growth, environmental degradation and sociocultural conflicts were cited as possible contributors of escalating mental-health problems, particularly depression and anxiety, among city populations. These may be applicable also in Nepal: decade-long political conflicts resulted in rapid migration of villagers into nearby cities, thus swiftly expanding the urban population [
45]. In addition, other factors may lead to an increase in the prevalence of psychiatric disorders in the cities [
46]: for example, the tendency of some mentally ill people to settle in towns rather than in the countryside, possibly to protect them from social stigma, to be away from the difficulties of rural life, to obtain proper care from social welfare institutions or better treatment, in search of jobs, or just to beg.
Habitation in Nepal also includes high hills and mountains, which cover almost one third of the total land area of the country [
47]. So far, no study has explored psychiatric illnesses among the occupants of these territories. This was the first research in the South Asian Region to demonstrate the possible effect of geographical elevation on mental health. It showed HADS-cAD to be more prevalent above 2000 m. Two studies in Peru [
48,
49] and a US study [
50] suggested the role of hypoxia and mitochondrial dysfunction as the possible link between altitude and depression. There are also studies on high-altitude ascenders from China [
51] and among porters and trekkers in Nepal [
52] that found anxiety to be one of the most recorded medical symptoms. But further work is necessary to elucidate whether biological conditions or psychosocial factors related to life adversity, isolation or the limited access to mental-health facilities in these areas are responsible for the mental-health problems.
Our findings of a negative association between alcohol use and HADS-A (as well as HADS-cAD), and of a positive relationship between marijuana use and HADS-D, need cautious interpretation. It is not so simple to capture the impact of alcohol use in a country like Nepal where drinking is considered an integral part of social functioning in most of the so-called
Matwali community, which traditionally is prone to drinking [
53]. Similarly, marijuana, which is also culturally accepted among the Nepalese, especially during certain religious functions [
54], may not show its true relationship with mental health. We were unable fully to evaluate the potential relationships between substance dependence and mental health because the length of the questionnaire restricted us from adding more questions. Further studies are necessary in this area.
Contrary to global [
5,
33,
34] and local studies [
10,
15,
16], we found no gender associations. Perhaps the somatic symptoms that generally are reported more frequently by depressed females [
55] were not sufficiently captured by HADS, which was originally constructed to detect emotional symptoms among hospital populations, and therefore pays less attention to bodily symptoms [
24].
Although many studies have shown associations between poverty and CMDs in the LAMI countries [
56], we did not find direct support for these. It may be that our questions on annual household consumption and monthly expenditure did not estimate well the socioeconomic status of the household; the responses may have been vaguely reported by the participants, many of whom were illiterate [
30]. In Nepal, where people generally are poor [
19], it is challenging to find a suitable measure of socioeconomic status in relation to CMD caseness. Income is not a reliable indicator, consumption cannot be assessed using direct monetary measures and proxy measures such as educational status have limited usefulness in this country with so much illiteracy; all these issues have been discussed in an earlier publication [
20]
. Nevertheless, our observation of high prevalence of both anxiety and depression among the Nepalese with relatively low socioeconomic status may well establish the link between poverty and mental ill-health. As concluded in a WMH survey, income inequality is a possible factor promoting chronic illnesses like depression [
36], but more so in high-income countries than in the LAMI countries.
Our study was built upon tried and tested methodology [
20], a large sample size, a very high participation rate achieved through careful sampling methodology [
21], completeness of data and representativeness of the population’s geographical and cultural diversities. These were the strengths of this study. The cross-sectional nature of the study obviously could not capture the longitudinal, relapsing and remitting course of depression and anxiety, or illustrate the temporal direction of associations with sociodemographic factors [
36]. The limitations in the use of HADS rather than expert clinical interview have been discussed above. However, these are considerably offset: we believe culturally validated study instruments account for socially acceptable outcomes better than the ethnically insensitive diagnostic classifications relied upon in various cross-national [
35] and global reviews [
33].
Competing interests
The author declares that they have no competing interests.
Authors’ contributions
Conception and design: AR, KM, ML, TS, AH. Acquisition of data: AR, KM, ML, TS, AH. Analysis and interpretation of data: AR, KM, ML, TS, AH. Drafting the article: AR. Revising it critically for important intellectual content: AR, KM, ML, TS, AH. Giving final approval of the version to be submitted: KM, ML, TS, AH. All authors read and approved the final manuscript.