Depression and anxiety among war-widows of Nepal

0 downloads 0 Views 2MB Size Report
Jun 19, 2017 - well described in literatures (Balkir et al., 2013; Hadley, Brewis, & Pike, 2010). For instance,. Hadley and colleagues found that freedom of ...
Psychology, Health & Medicine

ISSN: 1354-8506 (Print) 1465-3966 (Online) Journal homepage: http://www.tandfonline.com/loi/cphm20

Depression and anxiety among war-widows of Nepal: a post-civil war cross-sectional study Syaron Basnet, Pragya Kandel & Prabhat Lamichhane To cite this article: Syaron Basnet, Pragya Kandel & Prabhat Lamichhane (2018) Depression and anxiety among war-widows of Nepal: a post-civil war cross-sectional study, Psychology, Health & Medicine, 23:2, 141-153, DOI: 10.1080/13548506.2017.1338735 To link to this article: https://doi.org/10.1080/13548506.2017.1338735

Published online: 19 Jun 2017.

Submit your article to this journal

Article views: 83

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=cphm20

Psychology, Health & Medicine, 2018 VOL. 23, NO. 2, 141–153 https://doi.org/10.1080/13548506.2017.1338735

Depression and anxiety among war-widows of Nepal: a postcivil war cross-sectional study Syaron Basneta,b, Pragya Kandelc and Prabhat Lamichhaned§ a

Department of Public Health, University of Helsinki, Helsinki, Finland; bDepartment of Public Health Solutions, National Institute for Health and Welfare (THL), Helsinki, Finland; cResearch and Action in Public Health, Kathmandu, Nepal; dDepartment of Epidemiology and Preventive Medicine, Monash University, Melbourne, Australia

ABSTRACT

Thousands of Nepalese women were widowed as a consequence of a decade (1996–2006) long civil war in Nepal. These women are at grave risk of mental health problems due to both traumatic experiences and violation of natural order of widowhood. The present study explores the  depression and anxiety among war-widows. In 2012, a cross-sectional study was designed to interview 358 war-widows using validitated Beck Depression Inventory and Beck Anxiety Inventory in four districts of Nepal – Bardiya, Surkhet, Sindhupalchowk and Kavrepalanchowk with history of high conflict intensity. The prevalence of depression and anxiety was 53% and 63% respectively. Financial stress was significantly associated with depression (2.67, 95% CI: 1.40–5.07) and anxiety (2.37, 95% CI: 1.19–4.72). High autonomy of women as compared to low autonomy, high social support as compared to low social support and literacy as opposed to illiteracy was associated with less likelihood of depression and anxiety. Our results suggest high magnitude of depression and anxiety among warwidows in Nepal. Future policy efforts should be directed at providing mental health services to identify mental health issues among conflict affected individuals with focus on education, employment and activities to promote social support and autonomy at community.

ARTICLE HISTORY

Received 5 August 2016 Accepted 24 May 2017 KEYWORDS

War-widows; women; conflict; depression; anxiety; Nepal

Abbreviations: WHO: World Health Organisation; DALYs: Disabilityadjusted life years; PTSD: Post-Traumatic Stress Disorder; BDI: Beck-Depression Inventory; BAI: Beck Anxiety Inventory; WHR: Women Health Rights; DDC: District Development Committee; SNV: Netherlands development organisation; SIRF: Social Inclusion Research Fund; ORs: Odds ratios; CIs: Confidence intervals; UnOR: Unadjusted odds ratios; AOR: Adjusted odds ratios; HGRF: Harka Gurung Research Fellowship

Background It is estimated that 10% of the people who experience traumatic events will develop serious mental health problems and another 10% will develop behaviour that hinders their CONTACT  Syaron Basnet  [email protected] All authors have equally contributed to this work. § Present Address: School of Medicine, Deakin University, Geelong, Victoria. © 2017 Informa UK Limited, trading as Taylor & Francis Group

142 

 S. BASNET ET AL.

ability to function effectively. In recent years’ mental health as an indirect consequence of traumatic events such as civil war has been recognized as a serious public health challenge (Collins et al., 2011; Murray et al., 2002; Murthy, 2007). The estimated global prevalence of the post-traumatic stress disorder (PTSD) is 3 million (Collins et al., 2011). The association between war and mental health disorders such as PTSD is well recognised in literatures (Cardozo et al., 2004; Murthy, 2007). Depression, anxiety, acute stress disorder, bereavement reactions, fear, somatisation, anger control, functional disability, and arrest or regression of childhood developmental progression have also increasingly received recognition (Murthy, 2007). However, evidence on vulnerable population such as women, children, displaced, widows, orphans and elderly is still limited (Murthy, 2007; Raphael, Taylor, & McAndrew, 2008; Usta, Farver, & Zein, 2008). These populations are at the heightened risk of mental health problems due to the traumatic experiences (Corey LMK, 2006; Horowitz et al., 1997; Prigerson et al., 1997; Prigerson & Jacobs, 2001; Raphael & Martinek, 1997) The Nepalese civil war between 1996 and 2006 claimed 13,227 lives, 1007 disappearances, 785 disabilities, 150,000 internal displacement, and undocumented burden of mental health problems (Centre(INSEC), 2016; Douglas, 2005; Murshed & Gates, 2005; Singh, Dahal, & Mills, 2005). The, unnatural sudden death or disappearance of the partner made thousands of women be widowed to face restrictive social norms and values and exclusively endure the burden of elderly and child care (Surkan, Broaddus, Shrestha, & Thapa, 2015). The restriction and stigma faced by widows in Nepal is highlighted by the fact that women use non-disclosure of widowhood as a coping strategy (Surkan et al., 2015). The patriarchal social structure further harbours an unfavourable biased environment to these women. They are vulnerable to mental health problems owing to difficult socio-economic situation, violence, abuse and bereavement (Sabri et al., 2016; Surkan et al., 2015)

Methods Participants A cross-sectional study was conducted in four districts of Nepal in 2012. Semi-structured interview was conducted with 358 women. The participants were identified as the survivors of conflict who were married women and whose husband was killed or made to disappear during the civil war period (1996–2006) in Nepal by either rebel groups or government forces. The list of participants was collected from locally available committee (conflict affected society or local peace committee) and updated with the reference from Women Health Rights (WHR, a Non-Government Organisation) local office or from government office (District Development Committee Office). The number of interviews per district was based on the probability proportionate to size of the number of war-widows in other districts. Participants were randomly selected. The response rate was 94%. Study setting Number of deaths or disappearance per 1000 inhabitants was used to measure the conflict intensity. Conflict intensity was calculated based on the number of deaths and disappearance per 1000 inhabitants in the districts based on 1991 census data (Centre(INSEC),

PSYCHOLOGY, HEALTH & MEDICINE 

 143

Figure 1. Conflict intensity and study districts, Nepal.

2016; ‘Districts of Nepal’, 2000). The districts were divided into four groups based on conflict intensity quartile. The highest intensity districts had mean of 2.37 deaths/ disappearance per 1000 population (Standard Deviation (SD) 1.42), high intensity districts had 1.00 deaths/disappearances per 1000 population (.15), medium intensity had .58 deaths/ disappearances per 1000 population (SD .11) and low intensity district had .27 deaths/ disappearances per 1000 population (SD .08). We purposively selected two districts from highest intensity (Surkhet and Bardiya) and two from high intensity (Kavre and Sindhupalchowk) (Figure 1). The purposive selection was done to ensure that the districts from all three ecological regions of Nepal (Hill- Kavre and Surkhet, Southern lowland or terai – Bardiya; Mountain- Sindhupalchowk) and districts with war-widows who had husbands disappeared in the conflict were included in the study. Two districts were from central development region (Kavre and Sindhupalchowk) and two from mid-western development region (Surkhet and Bardiya). Measurements A semi-structured questionnaire was developed to collect information on socio-demographic characteristics, financial difficulty, social norms and support, decision-making, depression and anxiety. Socio-demographic measurements were completed age, women’s age at husband’s death or disappearance, current marital status, caste/ethnicity, district of residence, death or disappearance of husband, affiliation of husband, current number of family members, educational status, employment status, smoking and drinking habit. To measures financial

144 

 S. BASNET ET AL.

difficulty binary variable ‘stressful finance’ was created to explore women’s feeling stress due to financial loans or to pay tuition fees of their children. Another variable ‘conflict over property’ was created to explore women who were legally engaged over property dispute. Participants were asked about restrictions on the choice of clothes, foods, social function attendance, friends, gender of friends and places to visit. The responses were combined to form (i) low autonomy (women restricted for three items) (ii) medium autonomy (women restricted for two to one item), (iii) high autonomy (no restriction in any items). A ‘social support’ variable was created based on three questions asked to women (i) ‘when you need help or have a problem, can you usually count on member of your maternal family for support’ (ii) ‘do you have any close friends to who you can rely on in case you need help or have a problem’, and (iii) ‘are you a member of any community groups where you can discuss in case you need help or have a problem’. The responses were coded to create three groups of women ‘low social support’ (women with positive response to none or one question), ‘medium social support’ (women with positive response to two questions) and ‘high social support’ (women with positive response to all questions). Depression and anxiety was measured with Beck-Depression Inventory (BDI-21) and Beck Anxiety Inventory (BAI-21). These are psychometric questionnaire, which measures depression, and anxiety from 21 item scales which reflect intensity of a particular item of mood containing four statements each (Beck, 1961). The range of BDI-21 and BAI-21 total score is from 0 to 63 and that of each item is from 0 to 3. The scales were validated and previously used in Nepal (Kohrt et al., 2009; Kohrt & Worthman, 2009). The validated cutoff for BDI and BAI was 20 (sensitivity = .73 and specificity = .91) and 17 (sensitivity = .77 and specificity = .81) respectively (Kohrt et al., 2009, 2012; Kohrt & Worthman, 2009; Luitel et al., 2013). Based on the cut-off point, depression and anxiety score was categorised to no or minimal and moderate or severe category. The cut-off point is used to indicate the need for clinical intervention as in Luitel et al. (2013) and not to make diagnoses of psychological disorder (Kohrt et al., 2012). Statistical analysis Descriptive analysis was performed to assess the prevalence of depression and anxiety and its distribution by various socio-demographic characteristics. Kernel density estimators were used to reveal the underlying skewness and multi-modality of total depression and anxiety score as opposed to histogram (Salgado-Ugarte et al., 1994). Regression models were used to test the strength of association between women’s characteristics and depression and anxiety severity. The association of socio-demographic was examined with depression using three models of logistic regression (Table 3). First, we report the unadjusted odds ratios (UnOR) using univariate logistic regression for participant’s characteristics. Second, findings from two multivariable logistic regression are reported – (i) adjusted for demographic characteristics (ii) adjusted for demographic as well as socio-economic characteristics (AOR) (see Table 2). Third, socio-economic characteristics were added to the second model. Current age of women was dropped from the adjusted logistic regression due to correlation with age at which women lost her husband. Odds ratios (ORs) with 95% confidence intervals (CIs) have been presented along with the strength of statistical significance. Data analysis was performed with SPSS v.20 and STATA v13.1.

PSYCHOLOGY, HEALTH & MEDICINE 

 145

Ethics Verbal informed consent was obtained from the participants because of the literacy status and as deemed appropriate by ethical committee. Ethical committee of Netherlands development organisation (SNV)/Social Inclusion Research Fund (SIRF) approved the research protocol.

Results Demographic characteristics The demographic characteristics of the participant were measured. Median age was 37 years with more than three-fifth above 31 years of age. Less than 1 in 10 (6%) had remarried or had partner at the time of interview (see Table 1). Depression and anxiety The mean score of depression was 21.2 (SD, 11.6) with highest score of 49. More than half participants (53, 95% CI: 48–58%) had depression score of 20 or higher. The kernel density plot (Epanechnikov function) revealed that the scores clustered around 0–7 and 28–40 revealing some degree of multimodality in the distribution (Figure 2). The mean anxiety score was 20.95 (SD, 12.7) and a maximum score was 52. Nearly two-thirds (63, 95% CI: 58–68%) had >17 BAI score indicating moderate severity anxiety. Kernel density plot revealed skewness and bimodality with cluster around 7–10 score for anxiety (Figure 2). Socio-economic characteristics and mental health Three-quarter (75%) of the women were stressed by their financial situation and 15% had conflict over property (Table 2). One in four (24%) women had high autonomy. 51% had low social support. The proportion of widows with moderate severity depression was greater among those in financial stressful condition (60%), had conflict over property (78%), had low autonomy (70%) and had medium social support (63%). The pattern was similar for anxiety as widows in financial stressful condition (66%), having conflict over property (78%), with low autonomy (80%) and low social support (76%) had moderate severity anxiety score. Factors associated with moderate severity depression Univariate logistic regression showed being single, resident in Surkhet, had financial stressful condition and conflict over property had significantly higher odds (2.07–3.64) of having moderate severity depression score. Literate women had medium autonomy, had strong social support and whose husbands were ‘Maoists’ were significantly less likely (.60–.15) to have moderate severity depression in the univariate model. Even after adjustment of individual characteristics the significant results were retained with increases in odds values for all variables except husband’s affiliation to Maoists party (which became non-significant).

146 

 S. BASNET ET AL.

Table 1. Demographic characteristics and mental health. All women Women characteristics Current age  Less than 31  31−45   46 and above Age at which women lost her husband  Less than 31  31−45   46 and above Current marital status   Married/has partner  Single Caste   Brahmin/Chhetri   Janajatis/Dalits/Others District   Kavre   Sindhupalchowk   Bardiya   Surkhet Husband’s status  Death  Disappearance Husband’s affiliation  Government security force  Rebel/Maoist  General public/politician Number of family members   2 or less   More than 2 Educational status  Illiterate  Literate Employment status  Unemployed  Employed Smoking habit  No  Yes Drinking habit  No  Yes  Total

%

Depression No/minimal

Moderate/severe

No/minimal

%

%

%

n

17.9 60.6 21.5

64 217 77

62.3 29.6 8.1

223 106 29

6.4 93.6

23 335

40.5 59.5

145 213

15.4 24.9 38.3 21.5

55 89 137 77

68.7 31.3

246 112

21.5 21.5 57

77 77 204

12 88

43 315

33.5 66.5

120 238

26.5 73.5

95 263

80.7 19.3

289 69

76.8 23.2 100.0

275 83 358

Anxiety

 

 

     

   

         

54.7 43.3 49.4 43.9 51.9 48.3 69.6 45.1 42.8 49.3 52.7 49.4 48.9 35.1 43.1 54.5 40.3 57.1 45.1 48.8 46.3 38.3 50.8 50.5 45.2 47.1 44.9 46.5 47.0 46.6

 

 

     

   

         

45.3 56.7 50.6 56.1 48.1 51.7 30.4 54.9 57.2 50.7 47.3 50.6 51.1 64.9 56.9 45.5 59.7 42.9 54.9 51.2 53.7 61.7 49.2 49.5 54.8 52.9 55.1 53.5 53.0 53.4

 

 

     

   

         

53.1 32.7 35.1 37.7 40.6 17.2 65.2 34.9 32.4 39.9 29.1 25.8 46.0 39.0 29.7 52.7 37.7 32.5 38.2 34.9 37.1 20.0 45.4 29.5 39.5 39.1 27.5 37.1 36.1 36.9

Moderate/severe %  

 

     

   

         

46.9 67.3 64.9 62.3 59.4 82.8 34.8 65.1 67.6 60.1 70.9 74.2 54.0 61.0 70.3 47.3 62.3 67.5 61.8 65.1 62.9 80.0 54.6 70.5 60.5 60.9 72.5 62.9 63.9 63.1

The results however changed when adjusted for socio-economic characteristics as the significant association with marital status and husband’s status (death or disappearance) became non-significant. The final model showed that war-widows from Bardiya as compared to Kavre district (2.84, CI: 1.12–7.218), war-widows with stressful financial condition (2.67, CI: 1.40–5.07) and conflict over property (2.91, CI: 1.29–6.58) had significantly higher odds of moderate severity depression score. On the other hand, being literate (.49, CI: .26–.91), having high as compared to low autonomy (.30, CI: .15–.612) and strong social support as compared to low support (.16, CI: .07–.37) significantly reduced the odds of moderate severity depression score.

.01

.02

Density

.03

.03 .02

0

0

.01

Density

 147

.04

.04

PSYCHOLOGY, HEALTH & MEDICINE 

0

10

20

30

40

50

0

Depression score

20 40 Anxiety Score

60

Kernel density estimate Normal density

Kernel density estimate Normal density

Figure 2. Kernel density of depression and anxiety scores among war-widows, Nepal. Table 2. Socio-economic characteristics and mental health. Socio-economic characteristics Financial stressful condition  No  Yes Conflict over property  No  Yes Women’s autonomy  Low  Medium  High Social support  Low  Medium  Strong  Total

All women %  

24.9 75.1 84.9 15.1 39.1 36.6 24.3 51.1 28.5 20.4 100

Depression No/mild

n      

 

89 269 304 54 140 131 87 183 102 73 358

     

 

66.3 40.1 51.0 22.2 29.9 36.6 66.4 39.3 35.3 80.8 46.6

Moderate/severity   33.7 59.9   49.0 77.8   70.1 63.4 33.6   60.7 64.7 19.2 53.4

Anxiety No/mild      

 

47.2 33.5 39.5 22.2 19.5 26.7 57.1 23.5 31.4 78.1 36.9

Moderate/Severe      

 

52.8 66.5 60.5 77.8 80.5 73.3 42.9 76.5 68.6 21.9 63.1

Factors associated with moderate severity anxiety Association of war-widows characteristics with anxiety score was examined using univariate logistic regression and two multivariable regressions (Table 2). Univariate logistic regression showed that war-widows who were 46 years or older, single, had stressful financial condition and conflict over property had significantly higher odds (1.78–3.50) of moderate severity anxiety score. The odds was significantly lower and ranged from .48 to .08 for war-widows resident in Bardiya district, literate, whose husbands had disappeared, had high autonomy and strong social support. When adjusted for individual characteristics, the odds of moderate severity anxiety score was significantly higher for single war-widows (3.40, CI: 1.26–9.12) while it was lower for war-widows of 31–45 years of age (.52, CI: .29–.93), widows whose husbands had disappeared (.35, CI: .18–.69) and war-widows who were literate (.23, CI: .12–.43). After adjustment for both individual and socio-economic characteristics, the score was significantly associated for women with age 31–45 years as compared to women less

 S. BASNET ET AL.

40 30

BAI score

0

0

10

20

30 10

20

BDI score

40

50

50

148 

Kavrepalanchowk

Sindhupalchowk

Kavrepalanchowk

Sindhupalchowk

Bardiya

Surkhet

Bardiya

Surkhet

Figure 3. Depression and anxiety score distribution of war-widows by study districts.

than 31 years (.50, CI: .26–.96), literate women (.29, CI: .15–.58), financial stressful condition (2.37, CI: 1.19–4.72), women in high autonomy group as opposed to low autonomy (.31, CI: .14–.66) and strong social support as opposed to low support (.17, CI: .08–.39).

Discussion Bereavement is associated with stress and mental health problems. A study from Pakistan found that family loss was positively related with depression and anxiety (Mujeeb & Zubair, 2012). Another study from Bosnia and Herzegovina reported increasing depression, anxiety and somatization among women with child bereavement (Baraković, Avdibegović, & Sinanović, 2013). A systematic review assessing mood and anxiety disorders in widowhood found nine times elevated levels of anxiety after bereavement compare to controls (Onrust & Cuijpers, 2006). Our study results found that one in two war-widows had depression while two in every three had anxiety. Internally displaced people in Nepal reported up to 80% prevalence of depression in Nepal (Thapa & Hauff, 2005). The present findings were in consistent with other studies conducted in post-conflict in Nepal (Kohrt et al., 2012; Luitel et al., 2013; Thapa & Hauff, 2005). About one-third of general population, prior to conflict had depression (33%) and anxiety (27%) in a study conducted in a mountain district of Nepal (Kohrt et al., 2009). Another study reported double likelihood of anxiety in women than men (Kohrt & Worthman, 2009). A study conducted before and after conflict, reported increasing prevalence of depression from 30 to 40% and anxiety from 26 to 47% (Kohrt et al., 2012). Another study, in post conflict, reported depression, anxiety and PTSD to be 23, 22 and 9% respectively (Luitel et al., 2013). The findings were also similar to results from other countries in a post-conflict setting. A study conducted in the post war Afghanistan found 83% of women had depression and anxiety (Cardozo et al., 2004). Another study among widows from Kosovo war after 10 years found that 96% of widows had depression, anxiety or a substance use disorder (Morina et al., 2012). Our findings largely relates to social milieu which has been described by Trivedi and colleagues as one of the three underlying issues of depression besides reproductive and

Notes: Confidence Intervals in parentheses. *p