Background Globally, HIV remains a major public health challenge, having claimed an estimated 44.1 million lives to date [ 1 , 2 ]. At the end of 2023, approximately 39.9 million people were living with HIV, and an estimated 630,000 people died from HIV-related causes that year. In the African region, an estimated 26.0 million people were living with HIV in 2023, of whom 90% knew their status, 82% were receiving treatment, and 76% had suppressed viral loads.
Sub-Saharan Africa accounted for roughly 390,000 HIV-related deaths in 2023 [ 1 ]. In Ghana, 354,927 people were diagnosed with HIV, with an estimated 16,574 new HIV infections in 2022. It has made significant progress in strengthening its HIV response through the combined efforts of the Government of Ghana, the Ghana AIDS Commission, and the National AIDS Control Program, supported by the WHO and other partners [ 3 ].
However, the gaps which remain between actual progress and complete success still exist for HIV prevention and treatment efforts. The complete achievement of HIV program results remains blocked by structural and health system and socio-cultural challenges, which require better comprehension and resolution for people living with HIV. People living with HIV (PLWH) face numerous challenges - including social stigma, solitude, and anxiety, which heighten their susceptibility to mental disorders [ 4 , 5 ].
Depression is the most common mental condition among PLWH, with a meta-analysis showing that its prevalence is twice that of the general population [ 6 ]. Another meta-analysis reported an overall depression prevalence of 39% among PLWH, with individual studies across countries ranging from 12.8% to 78% [ 7 ]. Given the high burden and serious health consequences of depression in this population, identifying contributing factors is critical to guiding effective interventions.
In sub-Saharan Africa, approximately 3.63 million PLWH, have major depression while an estimated 1.57 million Disability-Adjusted Life Years (DALYs) are lost due to depression among PLWH [ 8 ]. Depression among PLWH is linked with poorer health-related quality of life, reduced viral suppression, and poor adherence to antiretroviral treatment [ 9 – 11 ]. In Ghana, the prevalence of depressive symptoms among PLWH varies by region.
The prevalence of depression in the Greater Accra region was 40.7% [ 12 ], in the Central region (50%) [ 13 ], and in the Volta region (20%) [ 14 ]. Previous studies have shown that depressive symptoms are associated with HIV viral load non-suppression among PLWH [ 15 , 16 ]. Research indicates that viral non-suppression among PLWH with major depressive disorder ranges from 12.2–13.5% [ 17 , 18 ].
A recent review also highlighted the complex influence of socioeconomic determinants of health on the relationship between depression and HIV outcomes, emphasizing the need for further research in this area [ 19 ]. Food insecurity and water insecurity are now considered as significant social determinants of health for PLWH. In particular, food insecurity among HIV infected individuals has been found to contribute to poor antiretroviral therapy (ART) adherence, virologic failure, poor immunologic recovery, and loss to follow up [ 20 – 24 ].
The lack of appropriate food resources can also negatively affect disease progression and ART success rates. On the other hand, water insecurity is also a crucial social determinant that affects HIV treatment and disease progression. It is reported that water insecurity can be associated with decreased ART adherence, physical illness, and wellbeing among PLWH as water is essential for taking medications, preparing food, and maintaining general hygiene in addition to preventing and controlling opportunistic infections [ 25 – 27 ].
It is therefore evident that both food and water insecurity are integral determinants that should be considered when managing HIV infection and disease. Apart from their impacts on HIV-related outcomes, food insecurity and water insecurity have significant effect on mental health. Both issues contribute to ongoing uncertainty, feelings of hopelessness, social stigma and stress for those struggling to obtain the essentials for life.
Individuals experiencing food insecurity may exhibit symptoms of depression as they may fear loss food, be embarrassed about their situation and withdrawn from society. Similarly, water insecurity can lead to depression due to the stress, loss of opportunity, worries and social stigma involved. Water insecurity also raises concerns such as spending too much time fetching water, worries about water quality, expenditure on water purchases and inadequate personal care amongst those involved.
Systematic reviews carried out, as well as surveys in different settings, show that both food insecurity and water insecurity are risk factors for mental health issues such as anxiety, distress, and depression [ 28 – 34 ]. Water-related problems are growing in magnitude and complexity globally, with adverse impacts on people in affected areas, such as water scarcity, too much availability from flooding, unreliability, inaccessibility or pollution [ 35 ]. Water insecurity is a situation in which people’s ability, or capacity, to access reliable, affordable, adequate and safe water is reduced such that their physical and mental health, and their ability to participate in social and cultural activities in a productive manner are impaired [ 36 ].
Water insecurity is an increasing challenge for global health and a major contributor to the global burden of disease [ 37 , 38 ]. More than 4 billion people in the world experience water stress at least once every month [ 37 ] and only 24% of the population in sub-Saharan Africa (SSA) has access to safe water [ 37 , 39 ]. PLWH are uniquely affected by not only water insecurity but also food insecurity, which is associated with poor mental health [ 22 , 24 , 26 , 40 – 45 ].
Food insecurity is widespread among PWLH in West Africa and impacts poor HIV outcomes, including virologic failure, noncompliance with ART, and loss to follow-up [ 22 , 24 , 44 ]. Prior studies indicate that depression is more prevalent among PWLH in West Africa [ 44 , 46 ] and might be up to three times as common in PLWH as in the general population [ 8 , 47 , 48 ]. Studies conducted within Ethiopia [ 30 ] and Lesotho [ 31 ] have shown an association between food insecurity, water insecurity and increased psychosocial distress.
In a longitudinal study of mothers living in western Kenya, researchers found a syndemic relationship between food insecurity, water insecurity and HIV infection and probable depression [ 32 ]. In Haiti, researchers found that household water insecurity has a direct association with increased anxiety and depression in addition to an indirect association with these outcomes through household food insecurity [ 28 ]. Additionally, researchers in India documented that the individual experiences of food insecurity and water insecurity varied from household to household and had differential associations with mental health outcomes [ 29 ].
Collectively, these studies point to the value of considering both food and water insecurity in understanding the root causes of mental health issues. Although recent literature increasingly recognizes HIV, food insecurity, water insecurity, and depression as interconnected syndemics [ 32 , 49 , 50 ], research simultaneously evaluating the role of all these factors is limited. There is also a lack of understanding of how food and water insecurity relate to depression among PLWH.
Although several studies have evaluated the association between either food or water insecurity and depression, none of them have combined both. Additionally, despite the recognized importance of addressing HIV, food, and water insecurities remain significant public health challenges in Ghana. Therefore, there is a need for further investigation of the link between food and water insecurities and depression among PLWH in Ghana.
Understanding this relationship is critical for informing public health policy, improving water access initiatives, and strengthening HIV care strategies. This study assessed the associations between water insecurity and food insecurity and how they relate to depression among PLWH in Ghana. Methods Study design This was a cross-sectional analysis of baseline data collected for a convergent mixed-methods study, which collects quarterly (every 3-month) survey data and annual biomarker data among a longitudinal cohort.
In this analysis, we only include the baseline data collected between November 2024 and January 2025. This analysis is based only on the baseline data. Study setting The study is conducted in three regions of Ghana: The Eastern, Greater Accra, and Upper East regions.
Ghana’s tropical climate is strongly influenced by the West African monsoon winds and is generally warm, with variable temperatures masked by seasons and elevation. There are five geographical territories that experience varied climates: northern, eastern coast, west coast, southern, and central territories. Ghana receives a mean annual rainfall of about 1200 millimeters, and the mean annual temperature ranges from 26°C along the coast to about 28.6°C in the driest northern regions, with two seasons: the dry season (November – February) and the rainy season (April – June) [ 51 ].
Northern Ghana typically records one rainy season (May – September), while Southern Ghana records two rainy seasons: a major season (April – July) and a minor season (September - November). The data analyzed in this study was collected in October-November. We enrolled participants from five facilities in the Greater Accra (Ashaiman Polyclinic, Ga West Municipal Hospital, Kaneshie Polyclinic, Madina Polyclinic - Kekele, and Weija Gbawe Municipal Hospital), three facilities in Eastern region (Akuse Government hospital, Atua Government Hospital, and Somanya Polyclinic), and three facilities from Upper East regions (Bongo Hospital, Paga District Hospital, and War Memorial Hospital).
Study population We enrolled patients who were receiving care at the 11 health facilities in Eastern, Greater Accra, and Upper East regions of Ghana, and had been initiated on ART at least 2 months prior to enrollment. We included PLWH who were 1) receiving care at the selected study facilities, 2) had initiated ART at least 2 months prior to enrollment, 3) capable and willing to provide informed consent, 4) willing to provide urine and blood samples for adherence and viral load measures, respectively, and 4) were aged 18 years or above. We excluded 1) patients who are too ill to participate or do not live within a feasible location if interviews are scheduled to be conducted in the community, and 2) patients who were unwilling to provide urine and/or blood samples.
Data collection Survey data was collected in-person at clinics or preferred participant locations using a programmed questionnaire in REDCap software with logic and consistency checks, and stored on a secure cloud-based server [ 52 ]. Data were collected in English, Twi, Ewe, Gurene, Kusaal, and Krobo. Research assistants who are fluent in English and the local languages were used to collect the data.
Data collection took place between November 10, 2024, and January 31, 2025. Outcome measure The outcome was a depression score generated from data collected using the patient health questionnaire (PHQ)-9 depression scale [ 53 ]. This scale is the 9-item depression module from the full PHQ, which assesses depressive symptoms over the last 2 weeks.
These items had the response options: not at all, several days, more than half the days, and nearly every day, recoded as ‘0’,’1’,’2’, and ‘3’ respectively. A generated summation score ranged between 0–27, and was then categorized as ‘None-minimal’ for ‘0-4’, Mild for ‘5-9’, moderate for ‘10-19’, and severe for ’20 or more [ 53 ]. The scale has been previously used in other studies in Ghana, but given that the ‘severe’ category had few observations, we combined the ‘moderate’ and ‘severe’ into ‘moderate/severe’ category Key independent variables Our exposure variables were food insecurity and water insecurity.
Food insecurity was measured by Household Food Insecurity Access Scale (HFIAS) [ 54 ]. WI was collected using the Individual WI Experiences (IWISE) Scale [ 55 ]. The IWISE Scale consists of 12 questions that query the frequency with which individuals experienced life-disrupting water-related problems in the previous 4 weeks.
Response options are ‘never’ (scored as 0), ‘1–2 weeks’ (1), ‘in some but not every week’ (2), or ‘almost every week’ (3). Scores for the 12 items are summed; the IWISE-12 scores range from 0 to 36 We created 4 categories as no to marginal ‘0-2’, Low ‘3-11’,moderate ’12–23’ and high’24–36’ [ 55 ]. The HFIAS is a sum of nine 4-point Likert scale questions, ranging from No (0), rarely (1), often (2), and (3) always.
Data on perceived stress was collected using perceived stress scale; consisting of 10 questions that widely used to assess perceptions of recent stress [ 56 ]. Response options were never (0), almost never (1), sometimes (2), fairly often (3), and very often (4). Total perceived stress score was generated, with higher scores representing higher levels of stress Covariates The following variables were used as control variables: age (in years), place of residence (peri-urban/rural and urban), marital status (single/never married, married/cohabiting, widowed, divorced/separated and dating), region of residence (Greater Accra, Eastern, and Upper East), gender (male and female), highest educational attainment (none, primary, O level, A level and tertiary), current occupation (agriculture, professional, sales and services [including petty trade], skilled manual, unskilled manual and unemployed), monthly income (1000 or less cedis, Above 1000 cedis), financial status (poor, average, not enough and enough) drinking water (piped water, well water and sachet/bottle water).
Anticipated HIV stigma was assessed using a 9-item scale adapted from the Anticipated Stigma subscale of the HIV Stigma Mechanisms Scale (HIV-SMS) [ 57 ]. Participants rated each item on a 5-point Likert scale ranging from 1 (“very unlikely”) to 5 (“very likely”), with higher scores indicating greater anticipated HIV stigma. Enacted HIV stigma was measured using a 9-item scale adapted from the Enacted Stigma subscale of the HIV Stigma Mechanisms Scale (HIV-SMS) [ 57 ].
Participants rated how often they experienced each behavior in the past six months using a 5-point Likert scale ranging from 1 (“never”) to 5 (“very often”), with higher scores indicating more frequent experiences of experienced stigma. Ethical considerations The study was conducted according to the Declaration of Helsinki. Approval for our study protocol was sought from the University of California, San Francisco Institutional Review Board (Approval number 24–41806) and Ghana Health Service Ethics Review Committee (Approval number GHSERC:012/07/24).
Written informed consent was obtained from all participants and witnessed by a healthcare provider or another individual chosen by the participant. No minors were included in the study. “Additional information regarding the ethical, cultural, and scientific considerations specific to inclusivity in global research is included in the Supporting Information ( S1 Checklist ) Data analysis Data were analyzed using Stata version 19.5 (StataCorp, College Station, TX, USA).
We present frequencies alongside percentages for categorical variables and means with standard deviations for continuous variables to describe the study population. We used the ordered logistic regression model with a health facility random intercept for both bivariable and multivariable analysis to investigate whether water insecurity is associated with depression. The measure of association is presented as the adjusted Odds ratio (aOR); statistical significance is determined using the 95% confidence intervals.
Missing data resulted in the loss of only one case in multivariable analyses and was thus ignored in the analyses. To test for proportional odds assumption, we used a generalized latent class modeling approach using gllamm, by fitting both the models that support the proportional odds assumption and the one that relaxes the assumption and then checking the best fitting model using a likelihood ratio test. This test showed the ordered model was a better fit, indicating a non-violation of the proportional odds assumption.
The final results presented here after are for the ordered logistic regression model with a health facility random intercept. Additionally, we conducted a stratified ordered logistic regression model by food insecurity levels, gender, one’s marital status, and perceived stress levels with a health facility random intercept. Results Description of health and sociodemographic characteristics of the study population Of the 503 participants enrolled, 68.8% were female, and 31.2% were male.
Almost a third (33.2%) of them were aged between 40–49, 23.9% aged between 50–59, 13.9% aged 60 or more, 18.5% aged between 30–39, and 10.5% aged 30 or less years. Nearly half (47.3%) of them experienced no to marginal water insecurity, 38.6% had low water insecurity, and 14.1% had high water insecurity. Over half (57.1%) of them had moderate food insecurity, 22.1% had low, and 20.9% had high food insecurity.
Almost two-thirds (64.0%) of them had high hygiene and 36.0% had moderate hygiene. Most (78.1%) of the participants had a monthly income of 1000 or less cedis. A more detailed description of the health and sociodemographic characteristics of the study population is presented in Table 1 .
10.1371/journal.pwat.0000546.t001 Table 1 Socio-demographic characteristics of the participants (N = 503). Characteristics Frequency (%) Participant’s gender Male 158 (31.4) Female 345 (68.6) Age in years 45.63 (12.06) Age in years categorized 30 or less 53 (10.5) 30-39 93 (18.5) 40-49 167 (33.2) 50-59 120 (23.9) 60 or more 70 (13.9) Religion Christianity 444 (88.3) Islam 40 (8.0) Traditional 15 (3.0) No religion 3 (0.6) Other (specify) 1 (0.2) Ethnicity Akan 70 (13.9) Ewe 79 (15.7) Ga/Dangme 154 (30.6) Guan 2 (0.4) Gurma 1 (0.2) Kusaasi 1 (0.2) Mole-Dagbani 5 (1.0) Kassem 101 (20.1) Frafra/Nankani 62 (12.3) Buli 5 (1.0) Mampruli 1 (0.2) Hausa 7 (1.4) Other (specify) 15 (3.0) Highest level of school completed None (no formal education) 130 (25.8) Elementary/ primary school 124 (24.7) Junior high school/ O level 135 (26.8) Senior high school/ A level 92 (18.3) Tertiary level (post senior high training/ polytechnical/ University) 22 (4.4) Currently work outside of regular housework No 80 (15.9) Yes 423 (84.1) Current main occupation Agriculture 77 (18.2) Professional/technical/managerial 38 (9.0) Clerical 3 (0.7) Sales sand services (including petty trading) 149 (35.1) Skilled manual 74 (17.5) Unskilled manual 70 (16.5) Other (specify) 13 (3.1) Description of financial situation Poor 239 (47.5) Average 178 (35.4) Not enough 78 (15.5) Enough 8 (1.6) Place of residence Rural 146 (29.0) Peri urban 188 (37.4) Urban 169 (33.6) Monthly money from all sources < =1000 cedis 390 (77.5) > 1000–5000 cedis 107 (21.3) > 5000–10000 cedis 3 (0.6) Refused to answer 3 (0.6) Current marital status Single/Never married 107 (21.3) Married/Cohabiting 183 (36.5) Widowed 100 (19.9) Divorced/Separated 97 (19.3) Dating 15 (3.0) Number people live in the household 4.61 (2.79) Head of household No 152 (30.2) Yes 351 (69.8) Distribution of depression among adults living with HIV Fig 1 provides the distribution of individual items. Overall, participants mostly recorded the option of “not at all” for each item, followed by “several days”, then “more than half the days”, and the least recorded option was “nearly every day”.
10.1371/journal.pwat.0000546.g001 Fig 1 Study participants responses to the individual items. The average depression score among study participants was 5.88 (SD = 4.82). Of which 47.1% (237/503) of them having none-minimal depression, 32.2% (162/503) having mild depression, 19.9% (100/503) having moderate depression, and 0.8% (4/503) of them having severe depression.
Fig 2 presents the different depression levels. 10.1371/journal.pwat.0000546.g002 Fig 2 Distribution of levels of depression among the study population. Among those who had none-minimal depression, almost two-thirds (62.9%) of them had no-to-marginal water insecurity, 51.5% had moderate food insecurity, most (70.9%) had high hygiene, and 79.3% had moderate perceived stress.
Of those who had mild depression, 45.7% of them had low water insecurity, 66.7% had moderate food insecurity, most (66.7%) had high hygiene, and 82.1% had moderate perceived stress. Among those who had moderate/severe depression, nearly half (48.1%) of them had low water insecurity, 54.8% had moderate food insecurity, most (55.8%) had moderate hygiene, and 63.5% had moderate perceived stress ( Table 2 ). 10.1371/journal.pwat.0000546.t002 Table 2 Characteristics of adults living with HIV among different depression levels.
None-minimal (N = 237) Mild (N = 162) Moderate/Severe (N = 104) Water insecurity categorized No-to-marginal ‘0-2’ 149 (62.9) 62 (38.3) 27 (26.0) Low ‘3-11’ 70 (29.5) 74 (45.7) 50 (48.1) Moderate ‘12-23’ 18 (7.6) 26 (16.0) 27 (26.0) Household food insecurity in past 3 months Low FI 82 (34.6) 18 (11.1) 11 (10.6) Moderate FI 122 (51.5) 108 (66.7) 57 (54.8) High FI 33 (13.9) 36 (22.2) 36 (34.6) Hygiene score in 3 levels Moderate hygiene 69 (29.1) 54 (33.3) 58 (55.8) High hygiene 168 (70.9) 108 (66.7) 46 (44.2) Age in years 46.24 (12.20) 44.76 (11.77) 45.61 (12.21) Age in years categorized 30 or less 25 (10.5) 17 (10.5) 11 (10.6) 30-39 41 (17.3) 31 (19.1) 21 (20.2) 40-49 80 (33.8) 53 (32.7) 34 (32.7) 50-59 54 (22.8) 41 (25.3) 25 (24.0) 60 or more 37 (15.6) 20 (12.3) 13 (12.5) Participant’s gender Male 94 (39.7) 46 (28.4) 17 (16.3) Female 143 (60.3) 116 (71.6) 87 (83.7) Highest level of school completed None (no formal education) 62 (26.2) 40 (24.7) 28 (26.9) Elementary/ primary school 48 (20.3) 48 (29.6) 28 (26.9) Junior high school/ O level 70 (29.5) 46 (28.4) 19 (18.3) Senior high school/ A level 46 (19.4) 22 (13.6) 24 (23.1) Tertiary level (post senior high training/ polytechnical/ University) 11 (4.6) 6 (3.7) 5 (4.8) Current main occupation Agriculture 39 (16.5) 32 (19.8) 6 (5.8) Professional 21 (8.9) 14 (8.6) 7 (6.7) Sales sand services 77 (32.5) 51 (31.5) 32 (30.8) Skilled manual 40 (16.9) 19 (11.7) 15 (14.4) Unskilled manual 26 (11.0) 22 (13.6) 22 (21.2) Unemployed 34 (14.3) 24 (14.8) 22 (21.2) Monthly money from all sources <=1000 cedis 180 (75.9) 131 (80.9) 82 (78.8) More than 1000 cedis 57 (24.1) 31 (19.1) 22 (21.2) Description of financial situation Poor 102 (43.0) 81 (50.0) 56 (53.8) Average 88 (37.1) 54 (33.3) 36 (34.6) Not enough 41 (17.3) 25 (15.4) 12 (11.5) Enough 6 (2.5) 2 (1.2) 0 (0.0) Current marital status Single/Never married 47 (19.9) 29 (17.9) 31 (29.8) Married/Cohabiting 92 (39.0) 60 (37.0) 31 (29.8) Widowed 51 (21.6) 27 (16.7) 22 (21.2) Divorced/Separated 39 (16.5) 39 (24.1) 19 (18.3) Dating 7 (3.0) 7 (4.3) 1 (1.0) Region Greater Accra 81 (34.2) 50 (30.9) 40 (38.5) Eastern Region 64 (27.0) 53 (32.7) 49 (47.1) Upper East Region 92 (38.8) 59 (36.4) 15 (14.4) Place of residence Rural 65 (27.4) 57 (35.2) 24 (23.1) Peri urban 93 (39.2) 59 (36.4) 36 (34.6) Urban 79 (33.3) 46 (28.4) 44 (42.3) MOS social support score, mean (SD) 64.75 (24.85) 53.39 (23.19) 39.68 (21.57) Perceived stress score categorized Low stress 37 (15.6) 5 (3.1) 0 (0.0) Moderate stress 188 (79.3) 133 (82.1) 66 (63.5) High stress 12 (5.1) 24 (14.8) 38 (36.5) ISAT score categorized Low 23 (9.7) 6 (3.7) 0 (0.0) Moderate 137 (57.8) 67 (41.4) 27 (26.0) High 77 (32.5) 89 (54.9) 77 (74.0) Anticipated Stigma score categorized Low 37 (15.6) 23 (14.2) 11 (10.6) Moderate 159 (67.1) 104 (64.2) 53 (51.0) High 41 (17.3) 35 (21.6) 40 (38.5) Socioeconomic standing of people in the community, mean (SD) 6.38 (2.34) 6.03 (2.79) 7.42 (2.33) Water situation of people in the community, mean (SD) 4.76 (2.54) 4.73 (2.90) 5.45 (2.62) Main source of drinking water Piped water 90 (38.0) 53 (32.7) 46 (44.2) Well water 54 (22.8) 47 (29.0) 18 (17.3) Sachet/bottled water 93 (39.2) 62 (38.3) 40 (38.5) Factors associated with depression among Ghanaian adults living with HIV In bivariable analysis, the factors associated with depression included: water insecurity, food insecurity, gender, attainment of O-level education, money from all sources, stigma, and having Sacket/bottled water as a main source of drinking water. Participants with lower or moderate water insecurity had higher odds of depression compared with those experiencing no-to-marginal water insecurity (low water insecurity OR=2.72, 95%CI: 1.98,3.74), and moderate water insecurity (OR=4.48, 95%CI: 2.75,7.30). Those with moderate or high food insecurity had greater odds of depression than participants with low food insecurity (moderate food insecurity; OR=3.49, 95%CI: 2.03,6.01), and high food insecurity (OR=7.14, 95%CI: 4.43,11.51).
Female participants had higher odds of depression compared with males (OR=2.07, 95%CI: 1.42,3.02). Participants who had attained O-level education had reduced odds of depression compared to those who had attained none (OR=0.56, 95%CI: 0.33,0.95). Participants whose monthly money from all sources was above 1000 cedis (equivalent to USD 100) had lower odds of depression than those whose monthly money was 1000 cedis or less (OR=0.65, 95%CI: 0.45,0.95).
Participants with any form of high stigma had greater odds of depression compared to those with a low form of stigma (High anticipated stigma, OR=2.61, 95%CI: 1.49,4.58 and high enacted stigma, OR=7.80, 95%CI: 2.13,28.53) ( Table 3 ). 10.1371/journal.pwat.0000546.t003 Table 3 Bivariate and multivariable analysis factors associated with depression. Characteristic Bivariate analysis (N = 503) Multivariable analysis (N = 502) cOR (95% CI) p aOR (95% CI) p Has water insecurity No to marginal ‘0-2’ Reference – Reference – Low ‘3-11’ 2.72 (1.98,3.74) <0.001 1.96 (1.30,2.95) 0.001 Moderate’12–23’ 4.48 (2.75,7.30) <0.001 2.64 (1.45,4.82) 0.001 Food Insecurity Low Reference – Reference – Moderate 3.49 (2.03,6.01) <0.001 2.22 (1.27,3.87) 0.005 High 7.14 (4.43,11.51) <0.001 3.32 (2.22,4.95) <0.001 Age in years 30 or less 30-39 40-49 50-59 60 or more Reference – – – 1.25 (0.65,2.39) 0.508 – – 1.17 (0.44,3.14) 0.756 – – 1.27 (0.50,3.23) 0.615 – – 0.98 (0.34,2.78) 0.967 – – Participant’s gender Male Female Reference – Reference – 2.07 (1.42,3.02) <0.001 1.73 (1.23,2.44) 0.002 Highest education level None Reference – Reference – Primary 1.06 (0.63,1.76) 0.832 – – O level 0.56 (0.33,0.95) 0.031 – – A level 0.74 (0.36,1.49) 0.398 – – Tertiary 0.77 (0.17,3.52) 0.739 – – Current main occupation Agriculture Reference – – – Professional 0.83 (0.40,1.73) 0.618 – – Sales sand services (including petty trade) 0.81 (0.43,1.52) 0.513 – – Skilled manual 0.75 (0.39,1.48) 0.411 – – Unskilled manual 1.87 (0.88,3.98) 0.104 – – Unemployed 1.28 (0.49,3.31) 0.616 – – Monthly money from all sources 1000 or less cedis Reference – Reference – Above 1000 cedis 0.65 (0.45,0.95) 0.026 1.18 (0.88,1.60) 0.270 Description of financial situation Poor Reference – – – Average 0.66 (0.39,1.10) 0.108 – – Not enough 0.54 (0.36,0.82) 0.004 – – Enough 0.18 (0.04,0.71) 0.014 – – Current marital status Single/Never married Reference – Reference – Married/Cohabiting 0.70 (0.47,1.05) 0.088 0.67 (0.42,1.06) 0.089 Widowed 0.86 (0.50,1.49) 0.590 0.61 (0.43,0.86) 0.006 Divorced/Separated 1.06 (0.69,1.63) 0.786 0.81 (0.54,1.20) 0.292 Dating 0.57 (0.30,1.11) 0.097 0.65 (0.27,1.56) 0.337 Region Greater Accra Reference – – – Eastern Region 1.47 (0.94,2.31) 0.090 – – Upper East Region 0.70 (0.31,1.58) 0.388 – – Place of residence Rural Reference – – – Peri urban 0.70 (0.37,1.32) 0.272 – – Urban 0.78 (0.50,1.22) 0.274 – – Perceived stress score categorized Low stress Reference – Reference – Moderate stress 9.06 (2.98,27.50) <0.001 6.22 (1.85,20.86) 0.003 High stress 41.11 (14.95,113.06) <0.001 21.07 (6.75,65.80) <0.001 Anticipated Stigma score categorized Low Reference – Reference – Moderate 1.26 (0.68,2.36) 0.460 0.98 (0.56,1.70) 0.937 High 2.61 (1.49,4.58) 0.001 1.53 (0.77,3.03) 0.226 Enacted Stigma score categorized Low Reference – Reference – Moderate 1.51 (0.76,3.02) 0.238 – – High 7.80 (2.13,28.53) 0.002 – – Main source of drinking water Piped water Reference – – – Well water 1.45 (0.87,2.40) 0.151 – – Sachet/bottled water 0.66 (0.48,0.90) 0.008 – – Footnote: the bold text within the table indicates statistical significance.
In the multivariable analysis, water insecurity, food insecurity, gender, current marital status, and perceived stress were significantly associated with depression. Participants with lower or moderate water insecurity had higher odds of depression compared with those experiencing no-to-marginal water insecurity (low water insecurity OR=1.96, 95%CI: 1.30,2.95), and moderate water insecurity (OR=2.64, 95%CI: 1.45,4.82). Those with moderate or high food insecurity had greater odds of depression than participants with low food insecurity (moderate food insecurity; OR=2.22, 95%CI: 1.27,3.87, and high food insecurity; OR=3.32, 95%CI: 2.22,4.95).
Female participants had higher odds of depression compared with males (OR=1.73, 95%CI: 1.23,2.44). Widowed participants had lower odds of depression than those who were single or never married (OR=0.61, 95%CI: 0.43,0.86). Participants with moderate or high perceived stress had substantially higher odds of depression compared with those with low perceived stress (moderate perceived stress; OR=6.22, 95%CI: 1.85, 20.86, and high perceived stress; OR=21.07, 95%CI: 6.75,65.80) ( Table 3 ).
Results from univariable stratified analysis of water insecurity on depression by food insecurity, gender, marital status, and perceived stress Among participants with low food insecurity and with none-minimal water insecurity, over two-thirds (77.2%) of them had none-to-marginal depression, similar patterns were observed among those with low food insecurity and with low water insecurity (65.5%), and low food insecurity and with moderate water insecurity (66.7%. Among participants with moderate food insecurity and had none-minimal water insecurity, most (56.4%) of them had none-to-marginal depression, in those with moderate food insecurity and with low water insecurity, nearly half them (42.7%) had mild depression, and for those with moderate food insecurity and moderate water insecurity,37.0% had mild depression. Among female participants with no to marginal water insecurity, 58.6% had none to minimal depression ( Table 4 ).
Significant univariable associations between water insecurity and depression were recorded among those with moderate, high food insecurity, males or females, single, married, widowed, moderate or high perceived stress ( Table 4 ). 10.1371/journal.pwat.0000546.t004 Table 4 Measure of associations of water insecurity on depression when stratified by food insecurity, gender, marital status, and perceived stress. Group Characteristic Depression level Univariable analysis (N = 503) None* Mild Moder* OR (95% CI) p Low food insecurity Has water insecurity No to marginal ‘0-2’ 61 (77.2) 12 (15.2) 6 (7.6) Reference – Low ‘3-11’ 19 (65.5) 6 (20.7) 4 (13.8) 1.80 (0.77,4.19) 0.175 Moderate’12–23’ 2 (66.7) 0 (0.0) 1 (33.3) 2.39 (0.09,65.06) 0.605 Moderate food insecurity Has water insecurity No to marginal ‘0-2’ 66 (56.4) 38 (32.5) 13 (11.1) Reference – Low ‘3-11’ 43 (34.7) 53 (42.7) 28 (22.6) 2.09 (1.52,2.88) <0.001 Moderate’12–23’ 13 (28.3) 17 (37.0) 16 (34.7) 3.20 (1.85,5.55) <0.001 High food insecurity Has water insecurity No to marginal ‘0-2’ 22 (52.4) 12 (28.6) 8 (19.0) Reference Low ‘3-11’ 8 (19.5) 15 (36.6) 18 (43.9) 4.10 (1.88,8.91) <0.001 Moderate’12–23’ 3 (13.6) 9 (40.9) 10 (45.5) 4.77 (1.58,14.36) 0.006 Has water insecurity No to marginal ‘0-2’ Low ‘3-11’ Male gender 57 (70.4) 21 (25.9) 3 (3.7) Reference – 30 (51.7) 16 (27.6) 12 (20.7) 2.64 (1.33,5.23) 0.006 Moderate’12–23’ 7 (38.9) 9 (50.0) 2 (11.1) 3.21 (1.56,6.58) 0.002 Female gender Has water insecurity No to marginal ‘0-2’ 92 (58.6) 41 (26.1) 24 (15.3) Reference – Low ‘3-11’ 40 (29.4) 58 (42.7) 38 (27.9) 2.78 (1.76,4.38) <0.001 Moderate’12–23’ 11 (20.7) 17 (32.1) 25 (47.2) 5.34 (2.44,11.69) <0.001 Single/ Never married Has water insecurity No to marginal ‘0-2’ 27 (62.8) 10 (23.3) 6 (13.9) Reference – Low ‘3-11’ 17 (36.2) 15 (31.9) 15 (31.9) 3.10 (1.33,7.20) 0.009 Moderate’12–23’ 3 (17.7) 4 (23.5) 10 (58.8) 8.46 (2.38,30.07) 0.001 Married/ Cohabiting Has water insecurity No to marginal ‘0-2’ 65 (69.1) 20 (21.3) 9 (9.6) Reference – Low ‘3-11’ 23 (34.3) 28 (41.8) 16 (23.9) 4.08 (2.36,7.05) <0.001 Moderate’12–23’ 4 (18.2) 12 (54.5) 6 (27.3) 6.43 (4.17,9.90) <0.001 Widowed Has water insecurity No to marginal ‘0-2’ 35 (64.8) 10 (18.5) 9 (16.7) Reference – Low ‘3-11’ 14 (41.2) 13 (38.2) 7 (20.6) 1.81 (1.07,3.05) 0.027 Moderate’12–23’ 2 (16.7) 4 (33.3) 6 (50.0) 6.12 (1.65,22.78) 0.007 Divorced/ Separated Has water insecurity No to marginal ‘0-2’ 19 (48.7) 17 (43.6) 3 (7.7) Reference – Low ‘3-11’ 11 (29.0) 16 (42.0) 11 (29.0) 2.72 (0.94,7.90) 0.066 Moderate’12–23’ 9 (45.0) 6 (30.0) 5 (25.0) 1.58 (0.67,3.69) 0.294 Dating Has water insecurity No to marginal ‘0-2’ 3 (37.5) 5 (62.5) 0 (0.0) Reference – Low ‘3-11’ 4 (57.1) 2 (28.6) 1 (14.3) 0.66 (0.02,18.28) 0.807 Moderate’12–23’ 0 (0.0) 0 (0.0) 0 (0.0) – – Low stress Has water insecurity No to marginal ‘0-2’ 27 (90.0) 3 (10.0) 0 (0.0) Reference – Low ‘3-11’ 9 (81.8) 2 (18.2) 0 (0.0) 2.12 (0.50,9.08) 0.311 Moderate’12–23’ 1 (100.0) 0 (0.0) 0 (0.0) – – Moderate stress Has water insecurity No to marginal ‘0-2’ 115 (61.2) 52 (27.6) 21 (11.2) Reference – Low ‘3-11’ 56 (38.1) 61 (41.5) 30 (20.4) 2.35 (1.71,3.25) <0.001 Moderate’12–23’ 17 (32.7) 20 (38.5) 15 (28.8) 3.18 (1.76,5.76) <0.001 high stress Has water insecurity No to marginal ‘0-2’ 7 (35.0) 7 (35.0) 6 (30.0) Reference – Low ‘3-11’ 5 (13.9) 11 (30.5) 20 (55.6) 3.40 (1.12,10.34) 0.031 Moderate’12–23’ 0 (0.0) 6 (33.3) 12 (66.7) 6.40 (1.48,27.64) 0.013 None*, none to minimal, moder* = moderate/severe; The bold text within the table indicates statistical significance.
Discussion This study assessed the association of water and food insecurity on depression among people living with HIV (PLWH) in Ghana. We found that water insecurity was associated with depression among PLWH in Ghana, which is consistent with a previous study in Kenya [ 49 ]. Our findings are consistent with a recent systematic review and meta-analysis, which revealed that water insecurity was associated with depression among PLWH [ 34 ].
Consistent with another recent systematic review and meta-analysis [ 33 ], our study found that food insecurity was associated with a higher risk of depression among PLWH. This finding aligns with evidence from several studies conducted in SSA, including Togo and Côte d’Ivoire [ 44 ], Kenya [ 49 , 58 ], Ethiopia [ 59 ], Tanzania [ 60 ], Zambia [ 61 ], and South Africa [ 62 ]. Individuals experiencing food insecurity often face uncertainty, chronic stress, and emotional strain, all of which can contribute to elevated depressive symptoms.
These findings demonstrate that water insecurity and food insecurity are critical and interrelated and associated with depression among PLWH in Ghana, highlighting the need for a more comprehensive care beyond medical treatment. Incorporating screening for water and food insecurity and depression into routine HIV care would be critical in early identification of psychosocial vulnerabilities that impact the mental health of PLWH in Ghana. Also, we found that female participants were more likely to experience depression, which is consistent with findings from another study conducted in Ghana [ 63 ] as well as from Ethiopia [ 64 ], Sudan [ 65 ], India [ 66 ] and Mozambique [ 67 ].
Psychosocial, biological, and sociocultural factors may be responsible for the observed patterns of stress among women. Balancing caregiving with work responsibilities, socio-economic inequalities experienced by women can add additional, chronic stressors to their already high level of stress. For many women stress may also be impacted by hormonal changes during the menstrual cycle; pregnancy; or menopause that may cause episodes and or increased vulnerability to periods of depression and anxiety [ 67 ].
Therefore, there is a clear need for gender-sensitive mental health policies which take into account the differential impact of stress on nomen in Ghana. Recommendations for mental health interventions for women include better access to mental health services through maternal and reproductive health services; gender-sensitive drug and counseling for gender specific mental health stressors; and routine mental health screening for women. Furthermore, the study found that participants with moderate or high perceived stress had higher odds of experiencing depression, consistent with findings from China [ 4 , 68 ], and the US [ 69 ].
Previous evidence demonstrates that emotional, instrumental, and social stressors are particularly elevated among PLWH and are strongly associated with depression [ 4 ]. These findings underscore the need for early interventions aimed at reducing HIV-related stress to prevent its long-term detrimental effects on mental health and the subsequent development of depression. Limitations of the study This study has several limitations.
First, its cross-sectional design prevents causal inferences from being made about the observed associations. Second, all measures were self-reported, which may introduce social desirability bias, especially in face-to-face surveys. To minimize this, research assistants were trained to build rapport, ensure participant comfort, and allow respondents to skip questions they found sensitive.
Another limitation was the inconsistency in recall periods across the main variables. For instance, the PHQ-9 depression scale assessed depressive symptoms over the past two weeks, whereas the IWISE scale (measuring water insecurity) and the HFIAS scale (assessing food insecurity) used a four-week recall period. To mitigate potential bias arising from these differences, research assistants were trained to emphasize the importance of adhering to the specified recall period during data collection.
Despite these limitations, this study is novel and represents the first investigation of the relationship between water and food insecurity and depression among PLWH across multiple regions in Ghana. Conclusion The study found that both water and food insecurity are significantly associated with depression among PLWH. Additionally, the markedly higher likelihood of depression among women and those experiencing moderate to high perceived stress highlights the need for gender-sensitive and stress-focused mental health services within HIV care models.
Supporting information S1 Checklist Inclusivity-in-global-research-questionnaire. (DOCX)