Abstract
Background:
Depression is emerging as a highly prevalent psychiatric condition among people living with HIV/AIDS (PLWHA). Perceived family support (PFS) buffers depression among chronic disease patients. However, a similar relationship among PLWHA is unexplored.
Objective:
To examine the relationship between PFS and depression among PLWHA in the Kathmandu Valley, Nepal.
Methods:
In this cross-sectional study, depression was measured by Beck Depression Inventory Ia. Perceived family support was measured by Nepali Family Support and Difficulties Scale. The status of depression was compared between 208 PLWHA and 208 HIV-negative participants. The relationship between PFS and depression was examined only among PLWHA.
Results:
Among each of the 208 participants, the number of depressed PLWHA (n = 61,29.3%) was higher than that of HIV-negative participants (n = 13,6.2%; P < .001). Perceived family support had a negative association with depression in PLWHA (adjusted odds ratio [AOR] = 0.20, 95% confidence interval [CI] = 0.08-0.53).
Conclusion:
In Nepal, PLWHA display a higher level of depression than HIV-negative people, and a lower level of PFS is associated with depression among PLWHA. Improved family support might be helpful in reducing depression among Nepalese PLWHA.
Introduction
Depression is emerging as a highly prevalent psychiatric condition among people living with HIV/AIDS (PLWHA). The prevalence of depression has been reported as high as 50% among the HIV-positive population. 1,2 People living with HIV/AIDS are 2 to 4 times more at risk of depression compared with HIV-negative population. 3,4
Depression has a wide range of negative effects on the lives of PLWHA and on efficient treatment and prevention of HIV infection. The PLWHA with depression have increased risk of HIV disease progression and nonadherence to antiretroviral therapy (ART). 5 They are also less likely to follow care providers’ advice and engage in high-risk sexual behaviors 6,7 and develop suicidal ideation. 8 Likewise, depressed HIV-positive individuals face higher expenditures on health care and poorer health-related quality of life than their nondepressed counterparts. 9,10
In Nepal, a small country in South Asia, the number of PLWHA has been increasing. The first case of HIV was reported in 1988. Then HIV infection gradually spread and at the end of 2010, the estimated number of PLWHA reached 55 626, 11 of which 47 645 were adults. The distribution of estimated HIV infection in the adult population shows that 28% are labor migrants, especially to India, 26% are low-risk females, 15% are low-risk males, 7% are men who have sex with men (MSM), 5% are the clients of female sex workers (FSWs), 3% are injecting drug users (IDUs), and 1% are FSWs. 11 Out of 68 000 estimated adult PLWHA in Nepal in 2006, 15.7% were residing in the Kathmandu Valley. 12
To our knowledge, only 2 studies have been conducted on the mental health of PLWHA in Nepal. In 1 study among 98 HIV-positive former CSWs, 13 3% of the participants were identified as depressed using Center of Epidemiologic Studies Depression Scale (CES-D). Lack of validation of CES-D in Nepal might have underestimated the rate of depression. In another study among 164 survivors of human trafficking, 14 13 were HIV-positive former sex workers. All of the HIV-positive sex workers were identified as depressed using Hopkins Symptom Checklist 25 (HSCL-25). Distress from trafficking as well as HIV-positive status and low sample size might have caused a higher depression rate in this study. HSCL-25 is not a properly validated scale in Nepal; a validation of the scale was performed against Composite International Diagnostic Interview (CIDI) diagnoses on a small subpopulation (n = 25) of a study. 15 Despite these limitations, both studies highlighted depression as a prevalent mental condition among PLWHA.
A number of factors place PLWHA at risk of depression. Such risk factors include higher HIV symptom burden, stressful life events, 6,16 current alcohol abuse and dependence, 17 and current drug use. 18 Likewise, varieties of factors are also correlated with depression among PLWHA. Such factors include stigma, internalized stigma, 19,20 AIDS in spouse, 17 living alone, and lack of social support. 19,21
Family support is a very essential social support in everyone’s life. However, within the family, PLWHA are stigmatized in many countries, and Nepal is not an exception. In Nepal, PLWHA are physically restricted within a household. 22 Such restrictions include refraining from using the kitchen, common bathroom, and same bed and from regular religious activities. In extreme cases, PLWHA are excluded from family, too. 22 They fear to disclose HIV status to family members considering that they might lose family support. 23 Female PLWHA as compared to male PLWHA get less support from their family. Although the women are most often infected with HIV by their husbands, they are not accepted and get less support from the family compared with their husbands. 23 Families of PLWHA are afraid of infection and of losing their social prestige in their communities. 23
Family support is also an important factor to ameliorate depression among chronic disease patients. In cancer and end-stage renal disease patients, higher level of family support was associated with lower level of depression. 24,25 Similarly, in psychiatric patients, family support played a major role in recovering from depression. 26,27
However, little data are available about the effect of family support on depression among PLWHA. In the United States, high perceived family support (PFS) was associated with lower levels of depression among HIV-positive women 28 as well as among HIV-positive men of diverse sexual orientation. 29 Although sociodemographic variables were adjusted for in these studies, other potential confounders of depression, such as stigma, ART use, and HIV-related symptoms, were not controlled.
Only 1 study in India 17 has examined the effect of family relationship on depression in South Asian countries. In this study, the family relationship variable was measured only by 2 questions: whether the respondents were currently residing with their family and a self-evaluation of their relationship with their family as either good or bad. In addition, the perception of support from family was not measured.
People living with HIV/AIDS perceiving lack of family support might be vulnerable to social, psychological, and financial problems. Therefore, this study was conducted to examine the association between PFS and depression among PLWHA in the Kathmandu Valley, Nepal.
Methods
Selection and Description of Participants
The study site for this cross-sectional study was the Kathmandu Valley. We selected PLWHA from baseline data of “sexual risk reduction intervention study among HIV-positive people” overseen by the second author (KCP). In this intervention study participants were recruited through 5 different nongovernmental organizations (NGOs) working with PLWHA in the Kathmandu Valley. The participants were recruited conveniently as a detail database was not available to select them randomly. Out of the 323 from the intervention study, we took a subsample of 210 PLWHA for this study. We selected the participants in such a way that the participants’ distributions in 5 NGOs in the main sample and in our subsample were similar.
The inclusion criteria were age 18 to 55 years, self-reported diagnosis of HIV infection at least 3 months prior to the date of interview, and written informed consent to participate in the study voluntarily. We set adult population as selection criteria because they constitute nearly 60% of the total HIV-positive population of this country. 11 The exclusion criteria were people being physically and mentally unable to give an interview.
For comparing depression between PLWHA and the normal population, we selected 210 HIV-negative participants using the matching criteria age ± 3, sex, and ethnicity. Ethnicity also served as a risk factor for depression among Nepalese people. 30 HIV-negative participants were recruited from the same community as that of PLWHA. The staff members of 5 NGOs identified the matched participants while conducting home visits under the Community Home Based Care (CHBC) program. In this study, the HIV status of PLWHA and HIV-negative participants was self-reported.
The number of participants was 210 PLWHA and their matched 210 HIV-negative people, but 2 HIV-positive participants and their matched HIV-negative individuals were excluded in statistical analysis due to missing information.
Measures and Variables
We used the following different variables and measures for this study and data were collected from February to May 2010.
Depression
We used the Beck Depression Inventory Ia (BDI-Ia) to assess depression. It is a 21-item and a 4-point Likert-type scale. Beck Depression Inventory Ia is composed of items relating to symptoms of depression such as hopelessness, irritability, cognition such as guilt, or feelings of being punished occurring in the past 2 weeks. For all analyses, we used a cutoff score of 20 or higher to distinguish depressed participants from the nondepressed. Cronbach α for the combined sample of PLWHA and HIV-negative participants was .90 and for PLWHA it was .89. The Nepali version of the scale has been validated for use in Nepal. 31
Perceived family support
Family in this study means a unit typically composed of a mated couple and their children (who may themselves be married) in coresidence. We measured PFS using the Nepali Family Support and Difficulty Scale which is a 10-item and a 4-point Likert-type scale. The items rank the respondent’s perception of receiving different forms of essential support from family in the past year (eg, How much do you feel disliked by your family? How much do you feel distant from your family?). The scale was developed for use specifically in Nepal. 32 The total score for the scale, ranging from 0 (no support) to 30 (high support), was dichotomized into low and high support by the median (23.0). Cronbach α of the scale in this study among the PLWHA was .85.
Internalized stigma
The measure for internalized stigma in this study was the 7-item Internalized AIDS Stigma Scale having dichotomous response (agree/disagree; eg, It is difficult to tell other people that I am HIV positive, I am ashamed that I am HIV positive). This instrument has been previously used for PLWHA in Cape Town, South Africa. 19 For analysis, we dichotomized the score into low and high stigma by the mean (4.00). In this study, the reliability coefficient was 0.70.
HIV symptom burden
We used a 20-item HIV symptom index (HSI) questionnaire 33 to measure the HIV symptom burden among PLWHA. The HIV-related symptoms assessed with response based on a 1-month recall period. There were 5 response options ranging from 0 (I don’t have this problem) to 4 (It bothers me a lot). For analysis, we dichotomized the total score into higher or lower symptoms by the median (34.0). The Cronbach α for this scale was .90 in this study.
Sociodemographic and other characteristics
Demographic characteristics assessed were age, gender, ethnicity, income, education, house type, marital status, and current employment. Ethnicity designation was based on the last name of the participants which indicate ethnicity and caste in Nepal. Based on hierarchy perceived by society, ethnicity was divided into 3 categories: Bahun/Chhettri as high class, Janajati as medium, and Dalit as low class. The items measuring substance use included alcohol use in the past month, current smoking status, and life time history of drug use. Different HIV-related characteristics included for PLWHA were months since testing HIV positive, current ART use, and HIV disclosure to at least 1 family member.
Procedure and Data Collection
We first prepared the questionnaire in English, translated it into Nepali, and then back translated into English. After revising the Nepali version questionnaire, we pretested it on 30 HIV-positive participants and made additional modifications based on the results. Then we hired 4 interviewers and provided them with a 1-day orientation on questionnaire content and interviewing techniques. The interviewers conducted face-to-face interviews for about 30 to 45 minutes for each HIV-negative participant and for 30 to 60 minutes for each HIV-positive participant.
The Ethical Committee of the University of Tokyo and the Nepal Health Research Council reviewed and approved the study protocol. The participants were informed about the study objectives and procedures at the beginning and assured of their anonymity and confidentiality. Skipping questions as well as withdrawal were allowed at any time of the study. We made arrangements to provide psychosocial support through NGO for participants of both groups who were depressed and who reported suicidal ideation.
Statistical Analysis
To determine the necessary sample size for PLWHA and their matched HIV-negative population, 3 we used a power of 80%, 2-sided significance level 95%, ratio of unexposed/exposed as 1, a percentage of unexposed with outcome 28% 30 and odds ratio 2. The calculated sample size was 210 for each group, assuming a 75% response rate.
In this analysis, we obtained data about frequencies and proportions for categorical variables, mean and standard deviation for normally distributed continuous variables, and median for skewed variables.
We used McNemar test to compare demographic and substance use characteristics and depression between PLWHA and HIV-negative comparison group. We used generalized estimating equations (GEEs) to examine the association between depression and HIV status controlling for the clustering within each pair of PLWHA and HIV-negative people. Other variables included in the model were marital status, job status, education, drug use, alcohol use, smoking status, owning a house, and PFS. These variables were selected based on the literature review. Age, sex, and ethnicity were not included in the model as these variables were matched while selecting the participants. We tested interaction between HIV status and all the variables included in the model. We evaluated the statistical significance of a first-order cross product term between HIV status and each variable in separate models. We included all the potential confounders in each model.
We examined the association between PFS and depression only among PLWHA using multivariate logistic regression while controlling other risk factors. Other factors included sociodemographic and substance use characteristics, HIV disclosure, HIV symptom burden, and internalized stigma. These variables were selected based on the literature review. Here also, we tested the interaction between PFS and all the variables included in the model. We evaluated the statistical significance of a first-order cross product term between PFS and each variable in separate models. We included all the potential confounders in each model. Furthermore, we used univariate logistic regression analysis to identify the factors associated with PFS among PLWHA. For all analytical procedures, we set statistical significance at P value less than .05. For all the statistical analyses, we used Statistical Package for the Social Sciences (SPSS) version 16.0.
Results
We analyzed 208 PLWHA with a mean age of 33.7 (SD, 6.7) years and 208 HIV-negative participants with a mean age of 33.2 (SD, 7) years. Within each group, 129 (62%) were men and 110 (52.9%) were of Janajati ethnicity. Demographic and substance use characteristics of PLWHA and their matched HIV-negative participants are presented in Table 1. No difference was detected between PLWHA and HIV-negative participants regarding demographic characteristics except education. The number of HIV-negative participants who had a higher level of education (>8 years of education) was higher compared with that of PLWHA (P < .001). The 2 groups were different regarding substance use characteristics (P < .001). Surprisingly, more HIV-negative participants were engaged in current alcohol use compared with PLWHA (P < .001). Out of the total PLWHA, 169 (81.2%) had disclosed their HIV status to at least 1 family member. Other HIV-related characteristics are given in Figure 1.
HIV-related characteristics of people living with HIV/AIDS (PLWHA).
Demographic and Substance Use Characteristics of PLWHA and Matched HIV-Negative Participants.
a Age, sex, and ethnicity are the matched criteria.
b Mean ± standard deviation (SD); 33.23 ± 7.04 (HIV negative) and 33.75 ± 6.66 (people living with HIV/AIDS [PLWHA]).
In this study, out of 208 participants in each group, 61 (29.3%) of PLWHA and 13 (6.2%) of HIV-negative participants were depressed (McNemar test, P < .001; Figure 2). In the GEEs, PLWHA were more likely to have depression compared with HIV-negative comparison group (adjusted odds ratio [AOR] = 8.11, 95% confidence interval [CI] = 4.03-16.33, P value < .001; Figure 2). No statistically significant interaction was detected between HIV status and other variables such as sociodemographic and substance use characteristics, and PFS.
Comparison of depression in PLWHA and in HIV-negative participants.
Out of the PLWHA (N = 208), 103 (49.5%) had a higher level of PFS (Table 2). Among the total PLWHA who reported lower levels of PFS, 51 (48.6%) were depressed. But among total PLWHA who reported higher levels of PFS, only 10 (9.7%) were depressed. In another way, among 61 PLWHA with depression, 51 (83.6%) reported lower levels of PFS. When we performed multiple logistic regression analysis, we adjusted demographics, substance use, and HIV-related characteristics, current ART status, internalized stigma, and HIV symptom burden. Then we found PLWHA who had a higher level of PFS were less likely to be depressed compared with those who had a lower level of PFS (AOR = 0.20, 95% CI = 0.08-0.53). No statistically significant interaction was detected between PFS and other variables such as sociodemographic and substance use characteristics, HIV symptom burden, internalized stigma, and HIV disclosure.
Factors Associated with Depression among PLWHA.
Abbreviation: PLWHA, people living with HIV/AIDS.
a P value < .01.
b P value < .001.
c P value < .05.
Besides PFS, other factors were also associated with depression among PLWHA. They were HIV symptom burden (AOR = 8.05, 95% CI = 3.18-20.35), internalized stigma (AOR = 4.17, 95% CI = 1.76-9.83), and having their own house in the Kathmandu Valley (AOR = 0.36, 95% CI = 0.14-0.95). The details of these factors are given in Table 2.
The factors associated with PFS were female gender (OR = 0.39, 95% CI = 0.22-0.70), education up to 8 years (OR = 5.00, 95% CI = 1.88-13.26) and higher than 8 years of education (OR = 8.33, 95% CI = 3.13-22.13), and having their own house in the Kathmandu Valley (OR = 2.03, 95% CI = 1.16-3.57; Table 3).
Factors Associated with Perceived Family Support (PFS) among PLWHA.
Abbreviation: PLWHA, people living with HIV/AIDS.
a P value < .01.
b P value < .05.
c P value < .001.
Discussion
In this study, around 29% of PLWHA were depressed compared with 6% of HIV-negative participants. This displays a substantial burden of mental health among PLWHA. PLWHA who had a lower level of PFS were more likely to be depressed than those who had a higher level.
The proportion of the depressed PLWHA in this study was different from other studies from different parts of the world. Our result was higher than that reported by some 4,13 while lower than that reported by others. 7,8,17 –19 Such discrepancies might be due to the differences in the selection criteria for participants and to the use of different measurement tools.
In this study PLWHA were 8 times more likely to be depressed than their HIV-negative counterparts; the association was not mediated by the sociodemographic and drug and alcohol use characteristics and PFS. A meta-analysis of 10 studies, which were conducted in different locations, 3 found PLWHA were only twice more likely to have depression than the HIV-negative comparison group. Therefore, our results underscore the urgent need of mental health interventions among PLWHA in the Kathmandu Valley, Nepal.
A study conducted on people without HIV in a very rural area of Nepal 30 used BDI-Ia as the depression scale and reported 33.7% depression. They speculated that their study population has a high depression rate because their study site has the lowest human development index in Nepal. Our study in the capital city of Nepal found a 29.3% depression rate among PLWHA. It shows the existence of substantial burden of mental health among PLWHA in the study site.
In this study, PLWHA who had lower levels of PFS were more likely to be depressed than those who had higher levels of PFS. This association was not mediated by sociodemographic and substance use characteristics and HIV-related characteristics. HIV symptoms in PLWHA and in those who meet the criteria for AIDS have worse physical functioning and emotional well-being compared with those with asymptomatic HIV infection or chronic diseases. 34 People living with HIV/AIDS who are facing such deteriorating conditions are likely to need caregivers. In the family-oriented society like South Asia, family is a primary source of support for everyone including PLWHA. 35,36 Therefore, family members might be expected to serve as an important source of support for PLWHA when necessary. Accordingly, PLWHA who have lower PFS might be worried about their limited physical functioning or emotional well-being in the future.
This study indicates that higher PFS can improve mental health status of PLWHA. In this study, women, the illiterate, and those who did not own a house in the Kathmandu Valley perceived lower support from their families. Therefore, the PFS program for PLWHA should particularly focus on such subgroups. For example, we need more programs that encourage family members of PLWHA to make them more supportive to the sufferers.
People living with HIV/AIDS who had higher internalized stigma were 4 times more likely to have depression than those with lower internalized stigma. Similar results were reported from the United States, 37 South Africa, 19 and India. 20 In Nepal, the HIV infection was portrayed as a fatal disease transferred by FSWs and drug use, at the beginning of the HIV epidemic. 38 Both FSWs and drug users are stigmatized groups in Nepal. So, this concept of HIV/AIDS triggered and compounded internalized stigma among PLWHA.
Moreover, PLWHA who had higher internalized stigma were less likely to attend an HIV support group and had less PLWHA friends as compared with those who had lower internalized stigma. 37 If PLWHA have many PLWHA friends and are engaged in an HIV support group, they might realize alternative sources of support. Thus, such support may reduce psychological distress in PLWHA.
In this study, the PLWHA who had higher HIV-related symptom burden were 8 times more likely to be depressed than those who had a lower HIV-related symptom burden. Chandra 17 and Knowlton 18 also reported similar findings but in different research settings. Antiretroviral therapy extended survival and improved the health-related quality of life of PLWHA 39 but it cannot cure AIDS. Therefore, although 70.7% of the depressed HIV-positive participants of this study were taking ART, they might be afraid of death if they develop a few HIV-related symptoms. Therefore, addressing the symptoms associated with HIV/AIDS might improve the mental health status of PLWHA.
The PLWHA who owned a house in the Kathmandu Valley were less likely to be depressed compared with those who did not. Those who did not own their houses in the Kathmandu Valley might be internal migrants. The health care utilizing behaviors of internal migrants might be poorer than their counterparts 40 due to limited social networks in the host community. Furthermore, in the Kathmandu Valley, owning a house is an indicator of higher income. 41 So the PLWHA with higher income may have better access to medical and mental health care and treatment and thus may have a lower risk of being depressed.
Our findings should be interpreted in the light of certain study limitations. First, our HIV-positive participants were the members of HIV-related NGOs. Therefore, our findings specifically represent the subpopulation of PLWHA who belong to the network of HIV-related organizations in the Kathmandu Valley. Second, all the data were self-reported, which can result in over- or underreporting. However, we attempted to minimize such bias by using HIV-positive interviewers and by assuring confidentiality of the information they had provided. Third, in this study the BDI-Ia was used as a screening tool for depression rather than as a diagnostic tool. Therefore, caution should be taken while interpreting the findings of this study. Fourth, the cross-sectional study design of this study limited the possibility of causal effect of PFS on depression. However, our study results are in line with a longitudinal study in the United States 42 that revealed the causal relationship of social support on depression. Prospective studies will be helpful to better understand the relationship observed among our participants. Finally, our findings may not be applicable to HIV-positive MSM as they were not included in this study group. Despite such limitations, this study has uncovered important findings regarding the mental health condition of PLWHA in the Kathmandu valley, Nepal.
In conclusion, PLWHA who had lower PFS were more likely to be depressed compared with those who had higher PFS. Our study thus underscores the need to improve PFS among those PLWHA whose family members are available and accessible. Moreover, it also underscores the need to reduce internalized stigma and to address HIV symptom burden among PLWHA. Women, the illiterate group, and those who do not own a house in the Kathmandu Valley should be a focus in particular for the intervention. In addition, we need to focus on the role of individual family members to identify key persons who can provide primary support.
Footnotes
Acknowledgments
We would like to acknowledge all the participants of this research for their participation and cooperation. We thank the interviewers who put forth their effort to carry out the interviews and also thank concerned NGOs for recruitment of the participants and managing an appropriate place for interviewing.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Part of this study was supported by the Grant for Research on Global Health and Medicine (No. 21A-2) from the Ministry of Health, Labour, and Welfare and by the Grant-in-Aid for Young Scientists (B) (22790581), Japan Society for the Promotion of Science, The Ministry of Education, Culture, Sports, Science and Technology, Japan.