Sociodemographic and clinical characteristics
This study included 259 (66.4%) females and 131 (33.6%) males, with only 98.9% response rate. The age of subjects ranged from 18 to 67 years, with a mean age of 36.1 and a standard deviation (SD) of 8.65 years; 77.2% have a Christian religious affiliation; and 91% live in urban residential areas. Ethnically, 61.5% were Amhara, 20.0% Oromo, 9.2% Agew, 3.1% Bertha and the rest were from other ethnic groups living in the region. About 50.0% were married, while 38.2% were either divorced or widowed at the time of data collection. Sociodemographic, economic and HIV-related clinical characteristics of our analysis sample of 390 peoples living with HIV in Benishangul Gumuz Regional State, Ethiopia, are presented in table 1.
The socioeconomic characteristics of participants were as follows: 36.4% had no formal education. Most of the participants were poor as measured by the wealth index (a proxy measure for asset possession), with 60% of them in the least relative wealth tertile, 19.5% in the middle, and only 20.5% in the highest relative wealth tertile. More than a quarter of participants were unemployed (31%). A large number of the participants (67.7%) lived on a poor mean monthly income of 1 260 Ethiopian Birr (equivalent to US$45), which is below the World Bank absolute poverty threshold of US$1.90/day [36]. The vast majority of the study participants (76%) were food insecure and (60%) had a BMI of less than 18.5 kg/m2.
All participants had been on ART follow-up for a mean duration of 2.49 years (SD = 0.77 years), while 7.7% of them had been on ART for more than 10 years. The mean CD4 count was 559 (SD = 319.6) cells/mL, with a range of 60 to 1 914 cells/mL. More than a quarter of them (33.6%) were frequently ill, with comorbidities in the previous three months that varied from pneumonia and tuberculosis to numerous opportunistic infections. Most of the participants were followed up at a public hospital (74%) and at a health centre (26%).
Health-related quality of life (HRQoL)
The PHS score ranged from 16.2 to 67.7, with a mean of 48.8 (SD = 8.9). Almost 44.6% of the study population had a PHS score of below the PROMIS population norm, below 50 T-score. The global MHS score ranged from 28.4 to 67.6, with a mean estimated at 50.8 (SD = 8.1); 41.8% of the study population had an MHS score of below the PROMIS population norm.
Table 1 Sociodemographic, economic and clinical characteristics of peoples living with HIV in Benishangul Gumuz Region, Ethiopia, 2020
Characteristics
|
Frequency (n, %)
|
SOCIODEMOGRAPHIC CHARACTERISTICS
|
|
Age (in years):
|
|
Below 25
|
33 (8.5)
|
25–35
|
182 (46.7)
|
Above 35
|
175 (44.9)
|
Gender:
|
|
Female
|
259 (66.4)
|
Male
|
131 (33.6)
|
Marital status:
|
|
Divorced/Widowed
|
149 (38.2)
|
Married
|
195 (50.0)
|
Single
|
46 (11.8)
|
Religion:
|
|
Christian
|
301 (77.2)
|
Muslim
|
89 (22.8)
|
Ethnic group:
|
|
Amhara
|
240 (61.5)
|
Oromo
|
78 (20.0)
|
Agew
|
36 (9.2)
|
Berta
|
12 (3.1)
|
Others
|
24 (6.2)
|
Residence area:
|
|
Urban
|
355 (91.0)
|
Rural
|
35 (9.0)
|
SOCIOECONOMIC CHARACTERISTICS
|
|
Education level:
|
|
Never been to school
|
142 (36.4)
|
Primary level
|
166 (42.6)
|
Secondary level
|
66 (16.9)
|
College/University level
|
16 (4.10
|
Employment status:
|
|
Unemployed
|
122 (31.3)
|
Employed
|
268 (68.7)
|
Monthly income (in Ethiopian Birr):
|
|
<1 400
|
264 (67.7)
|
1 401–2 800
|
89 (22.8)
|
>2 800
|
37 (9.5)
|
Food security status:
|
|
Food insecure
|
296 (76)
|
Food secure
|
94 (24)
|
BMI score (in kg/m2):
|
|
<18.5
|
235 (60)
|
>18.5
|
155 (40)
|
Household wealth index tertile:
|
|
1st tertile
|
234 (60)
|
2nd tertile
|
76(19.5)
|
3rd tertile
|
80 (20.5)
|
CLINICAL FEATURES
|
|
Duration of ART initiation:
|
|
Less than 12 months
|
37 (9.5)
|
1–5 years
|
155 (39.7)
|
5–10 years
|
168 (43.1)
|
>10 years
|
30 (7.7)
|
Recent CD4 count (in cells/mL):
|
|
<350
|
108 (27.7)
|
350–500
|
81 (20.8)
|
>501
|
201 (51.5)
|
Comorbidities:
|
|
Yes
|
131 (33.6)
|
No
|
259 (66.4)
|
N, frequency in number; %, percentage
Factors associated with physical HRQoL
Sociodemographic and economic inequalities, and clinical determinants associated with HRQoL of PLWHA were explored using bivariate analysis and subsequent multivariable logistic regression analyses (a stepwise backward Wald method). Table 2 presents PHS and MHS scores in bivariable linear regression analysis.
In bivariate logistic regression analyses, age, ethnic group, religious affiliation, place of residence, income, food security status, wealth index and current CD4 count have a statistical association with the PHS quality of life domain. Table 3 presents the PHS and MHS scores during multivariate linear regression analysis with selected study characteristics.
In multivariate analysis, age, ethnic group, religious affiliation, employment status, household food security status and comorbidities have a statistically significant association with global physical health summary scores. Accordingly, an age below 25 years was found to inversely associated with poor PHS scores (AOR: 0.4; 95% CI: 0.16, 0.97; p = 0.043). In terms of ethnic group, being Oromo is nearly four-fold more likely to lead to a poor PHS score than being from one of the other ethnic groups in the study (AOR: 3.7; 95% CI: 1.53, 9.25; p = 0.004). Another socioeconomic factor is lack of employment. Unemployed PLWHA were found more likely to have poor physical health (AOR: 1.73; 95% CI: 1.06, 2.82; p =0.028). In addition, food-insecure participants were twice as likely to have a poor PHS score (AOR: 1.96; 95% CI: 1.10, 3.51; p =0.022); comorbidities with HIV were among the factors that significantly associated with poor PHS score (AOR: 1.34; 95% CI: 1.01, 1.62; p = 0.038).
Table 2 PHS and MHS scores, bivariable linear regression analysis with selected study sample characteristics in Benishangul Gumuz Region, Ethiopia, 2020
Predictors
|
HRQoL
|
PHS
(PHS summary score <50)
|
|
MHS
(MHS summary score <50)
|
Bivariate model
|
|
Bivariate model
|
OR
|
95% CI
|
p- value
|
OR
|
95% CI
|
p- value
|
SOCIODEMOGRAPHIC CHARACTERISTICS
|
|
|
|
|
|
|
|
Gender:
|
|
|
|
|
|
|
Female
|
1.29
|
[0.85, 1.97]
|
0.23
|
1.38
|
[0.89, 2.12]
|
0.14
|
Male
|
1.00
|
|
|
1.00
|
|
|
Age (in years):
|
|
|
|
|
|
|
Below 25 years
|
0.64
|
[0.30, 1.35]
|
0.24
|
1.16
|
[0.55, 2.46]
|
0.70
|
25–35 years
|
0.98
|
[0.64, 1.49]
|
0.92
|
1.26
|
[0.83. 1.92]
|
0.28
|
Above 35 years
|
1.00
|
|
|
1.00
|
|
|
Marital status:
|
|
|
|
|
|
|
Married
|
1.10
|
[0.56, 2.12]
|
0.78
|
1.62
|
[0.83, 3.13]
|
0.16
|
Divorced/widowed
|
0.60
|
[0.31,1.15]
|
0.124
|
0.85
|
[0.44, 1.65]
|
0.63
|
Single
|
1.00
|
|
|
1.00
|
|
|
Ethnic group:
|
|
|
|
|
|
|
Amhara
|
1.15
|
[0.57, 2.33]
|
0.68
|
1.22
|
[0.59, 2.53]
|
0.59
|
Oromo
|
3.47
|
[1.51, 7.91]
|
0.003
|
1.29
|
[0.57, 2.93]
|
0.53
|
Agew
|
1.12
|
[0.44, 2.82]
|
0.814
|
1.97
|
[0.77, 5.08]
|
0.16
|
Others
|
1.00
|
|
|
|
1.00
|
|
|
Religious affiliation:
|
|
|
|
|
|
|
Christian
|
0.43
|
[0.26, 0.71]
|
0.001
|
0.67
|
[0.42, 1.07]
|
0.097
|
Muslim
|
1.00
|
|
|
1.00
|
|
|
Place of residence:
|
|
|
|
|
|
|
Urban
|
2.26
|
[1.10, 4.63]
|
0.026
|
1.24
|
[0.60, 2.54]
|
0.56
|
Rural
|
1.00
|
|
|
1.00
|
|
|
SOCIOECONOMIC CHARACTERISTICS
|
|
|
|
|
|
|
Education level:
|
|
|
|
|
|
|
Never been to school
|
1.15
|
[0.41, 3.23]
|
0.79
|
2.19
|
[0.67, 7.14]
|
.191
|
Primary level
|
1.34
|
[0.48, 3.74
|
0.58
|
2.47
|
[0.76, 7.98]
|
.130
|
Secondary level
|
1.27
|
[0.43, 3.81]
|
0.66
|
1.71
|
[0.49, 5.91]
|
.393
|
College/University level
|
1.00
|
|
|
1.00
|
|
|
Employment status:
|
|
|
|
|
|
|
Unemployed
|
1.95
|
[1.25, 3.04]
|
0.003
|
2.18
|
[1.41, 3.37]
|
<0.0001
|
Employed
|
1.00
|
|
|
1.00
|
|
|
Monthly household income (in Ethiopian Birr):
|
|
|
|
|
|
|
<1 400
|
1.88
|
[1.19, 2.96]
|
0.007
|
1.64
|
[1.03, 2.63]
|
0.039
|
1 400–2 400
|
1.41
|
[0.69, 2.86]
|
0.343
|
1.27
|
[0.61, 2.63]
|
0.517
|
>2 400
|
1.00
|
|
|
1.00
|
|
|
Food security status:
|
|
|
|
|
|
|
Food insecure
|
2.5
|
[1.55, 4.02]
|
<0.0001
|
7.97
|
[4.08, 12.5]
|
<0.0001
|
Food secure
|
1.00
|
|
|
1.00
|
|
|
BMI score (in kg/m2):
|
|
|
|
|
|
|
<18.5
|
1.29
|
[0.86, 1.94]
|
0.22
|
1.77
|
[1.16, 2.70]
|
0.008
|
>18.5
|
1.00
|
|
|
1.00
|
|
|
Wealth index tertile:
|
|
|
|
|
|
|
3rd tertile
|
0.57
|
[0.34, 0.96]
|
0.034
|
3.49
|
[3.19, 7.82]
|
0.007
|
2nd tertile
|
0.74
|
[0.39, 1.41]
|
0.362
|
0.74
|
[0.39, 1.37]
|
0.337
|
1st tertile
|
1.00
|
|
|
1.00
|
|
|
Duration on ART:
|
|
|
|
|
|
|
1 year or less
|
1.25
|
[0.61, 2.55]
|
0.54
|
1.25
|
[0.61, 2.54]
|
0.54
|
1–5 years
|
1.1
|
[0.71, 1.66]
|
0.69
|
1.43
|
[0.93, 2.18]
|
0.11
|
>5 years
|
1.00
|
|
|
1.00
|
|
|
Recent CD4 count (cells/mL):
|
|
|
|
|
|
|
<350
|
1.65
|
[1.02, 2.68]
|
0.042
|
2.42
|
[1.50, 3.90]
|
<0.0001
|
350–500
|
0.72
|
[0.43, 1.22]
|
0.22
|
0.58
|
[0.58, 1.71
|
0.99
|
>501
|
1.00
|
|
|
1.00
|
|
|
Comorbidities:
|
|
|
|
|
|
|
Yes
|
1.65
|
[1.88, 2.07]
|
0.002
|
1.86
|
[1.22, 2.85]
|
0.004
|
No
|
1.00
|
|
|
1.00
|
|
|
HRQoL, health-related quality of life; PHS, physical health summary; MHS, mental health summary; OR, odds ratio; CI, confidence interval
Factors associated with mental HRQoL
In bivariate logistic regression analysis, employment status, income, household food security, BMI, wealth index, CD4 cell count and comorbidities with HIV were found to have a statistically significant association with MHS scores (see table 2). During multivariate analysis, marital status, employment status, household food security status, wealth index, CD4 cell count and comorbidities were remained in the final multivariate logistic regression using backward Wald analysis. However, marital status did not show statistical significance in global MHS scores during the final steps in backward Wald multivariate analysis (see table 3).
Unemployed PLWHA were 1.45 times more likely to have poor MHS scores than those who had jobs (AOR: 2.65; 95% CI: 1.04, 2.65; p = 0.035). Food-insecure people living with HIV were around 6.5 times more likely to have poor MHS scores than their food-secure counterparts (AOR: 6.43; 95% CI: 3.22, 9.82; p < 0.0001). Participants in the third tertile of the wealth index (those with the least wealth) were almost twice as likely to have poor MHS scores than those in the second tertile of the wealth index (AOR: 1.77; 95% CI: 1.21; 3.38; p =0.036). It was also shown that those PLWHA who reported to have a current CD4 count ≥ 350 cells/mL were 1.94 times less likely to have a high MHS score than those whose CD4 count was above 350 cells/mL (AOR: 1.91; 95% CI: 1.14, 3.21; p = 0.014). Likewise, those PLWHA who were sick with HIV/AIDS-related comorbidities were found 1.52 times more likely to have poor mental HRQoL than those without the conditions (AOR: 1.52; 95% CI: 1.62, 3.68; p = 0.042).
Table 3 PHS and MHS scores, multivariable linear regression analysis with selected study sample characteristics in Benishangul Gumuz Region, Ethiopia, 2020
Variables
|
Multivariate analysis
|
AOR
|
95% CI
|
p - value
|
PHS
|
Age below 25 years
|
0.40
|
[0.16, 0.97]
|
0.043
|
Married
|
1.04
|
[0.49, 2.19]
|
0.920
|
Divorced/widowed
|
0.52
|
[0.25, 1.07]
|
0.074
|
Amhara
|
1.06
|
[0.49, 2.25]
|
0.890
|
Oromo
|
3.76
|
[1.53, 9.25]
|
0.004
|
Agew
|
1.08
|
[0.39, 2.94]
|
0.880
|
Christian
|
0.39
|
[0.23, 0.69]
|
0.001
|
Unemployed
|
1.73
|
[1.06, 2.82]
|
0.028
|
Food-insecure
|
2.31
|
[1.36, 3.94]
|
0.002
|
Comorbidities
|
1.34
|
[1.01, 1.62]
|
0.038
|
MHS
|
Married
|
1.31
|
[0.62, 2.81]
|
0.481
|
Divorced/widowed
|
0.74
|
[0.35, 1.56]
|
0.428
|
Unemployed
|
2.65
|
[1.04, 2.65]
|
0.035
|
Food-insecure
|
6.43
|
[3.22, 9.82]
|
<0.0001
|
3rd tertile wealth index
|
1.77
|
[1.21, 3.38]
|
0.036
|
2nd tertile wealth index
|
0.82
|
[0.41, 1.67]
|
0.587
|
CD4 count below 350 cells/mL
|
1.91
|
[1.14, 3.21]
|
0.014
|
CD4 count 350–500 cells/mL
|
1.04
|
[0.58, 1.85]
|
0.090
|
Comorbidities
|
1.52
|
[1.62, 3.68]
|
0.042
|
AOR, adjusted odds ratio