1. Sample Characteristics
Table 1: Sample Characteristics
Variable
|
Total Sample (N=1260)
% (n)
|
Age
|
|
14 to 15 years
|
56.98(718)
|
16 to 17 years
|
43.02(542)
|
Orphan hood Status
|
|
Orphan
|
17.06(215)
|
Non - orphan
|
82.94(1045)
|
Primary Caregiver
|
|
Biological parent
|
76.59(965)
|
Grandparent
|
11.11(140)
|
Other relative
|
12.30(155)
|
Household size (Mean, SD)
|
|
Number of people in HH
|
7.00(2.71)
|
Number of children in HH
|
3.49(2.10)
|
Family Assets (Mean, SD)
|
11.46(3.26)
|
Baseline sample characteristics are presented in Table 1 below. The average age of participants is 15.4 years, with the majority of participants (57%) between 14 and 15 years. About 83% of participants are non-orphan, and 76.6% identified a biological parent as their primary caregiver. The average household is 7 people, with three children under the age of 18 years living in the household. The average score on family asset index (including land, modes of transportation and communication, gardens, farm animals and small business) is 11.46 out of the expected 21, indicating moderate levels of asset ownership.
2. Participants’ sociodemographic and household characteristics, family and social support factors and psychological wellbeing, by age group
Table 2: Bivariate Analysis Results: Sociodemographic and household characteristics, family and social support factors, and participants’ psychological wellbeing by age groups
Variable
|
Total Sample
(N=1260)
Mean (SD)
|
14 -15 years
(n=718)
Mean (SD)
|
16- 17 years
(n=542)
Mean (SD)
|
c2/t-test
|
Orphan hood Status (%, n)
|
|
|
|
|
Orphan
|
17.06(215)
|
17.27(124)
|
16.79(91)
|
0.05
|
Non - orphan
|
82.94(1045)
|
82.72(594)
|
83.21(451)
|
|
Primary Caregiver (%, n)
|
|
|
|
|
Biological parent
|
76.59(965)
|
75.91(545)
|
77.49(420)
|
1.88
|
Grandparent
|
11.11(140)
|
10.72(77)
|
11.62(63)
|
|
Other relative
|
12.30(155)
|
13.37(96)
|
10.89(59)
|
|
Household size
|
|
|
|
|
Number of people in HH (min/max: 2-31)
|
7.00(2.71)
|
6.88(2.78)
|
7.16(6.94)
|
-1.81
|
Number of children in HH (min/max: 0-13)
|
3.49(2.10)
|
3.39(2.01)
|
3.64(3.45)
|
-2.02*
|
Family assets (min/max: 0-20)
|
11.46(3.26)
|
11.32(3.29)
|
11.64(3.21)
|
-1.71
|
Family support factors
|
|
|
|
|
Family cohesion (min/max:7-35)
|
26.58(5.69)
|
26.61(5.73)
|
26.53(5.66)
|
0.25
|
Family care and relationship (min/max: 9-30)
|
24.74(4.27)
|
25.28(4.09)
|
24.01(4.41)
|
5.27***
|
Perceived child-caregiver support (min/max: 29-81)
|
56.85(6.73)
|
57.16(6.72)
|
56.44(6.72)
|
1.85
|
Social support from multiple sources
|
|
|
|
|
Guardian (min/max: 9 -30)
|
23.66(4.19)
|
23.93(4.12)
|
23.29(4.25)
|
2.67**
|
Classmate (min/max: 7 -25)
|
17.78(3.52)
|
17.87(3.55)
|
17.67(3.47)
|
1.01
|
Teacher (min/max:10 -30)
|
22.85(4.23)
|
23.11(4.22)
|
22.49(4.21)
|
2.58**
|
Friend/peer (min/max: 26 -65)
|
51.63(7.61)
|
51.91(7.51)
|
51.27(7.74)
|
1.48
|
Psychological wellbeing
|
|
|
|
|
Tennessee Self – Concept (min/max: 44 -100)
|
80.84(11.98)
|
81.79(11.75)
|
79.58(12.17)
|
3.26***
|
Beck's Hopelessness Scale (min/max: 0 -16)
|
4.20(2.97)
|
3.96(2.90)
|
4.53(3.04)
|
-3.41***
|
Rosenberg Self Esteem Scale (min/max: 16 -40)
|
34.00(4.57)
|
34.20(4.53)
|
33.75(4.63)
|
1.62
|
*p≤.05; **p≤.01; ***p≤.001
Results from bivariate analyses for participants’ sociodemographic and household characteristics, family support, social support factors and psychological wellbeing by age groups are presented in Table 2. Participants in both age groups were similar in terms of orphanhood status and primary caregiver reports. Older adolescents were more likely to report on average, slightly more children living in the household compared to younger adolescents (3.64 versus 3.39 children). The mean difference between the two age groups was statistically significant (t = -2.02, p ≤ .05). In terms of social support factors, younger adolescents reported slightly higher scores on all measures of family and social support factors compared to older adolescents, including on family cohesion (mean = 26.61 versus 26.53), family care and relationships (mean = 25.28 versus 24.01) and perceived child-caregiver support (mean = 57.16 versus 56.44). The mean score difference on the measure of family care and relationship was statistically significant (t = 5.27, p = .001). In addition, younger adolescents were more likely to report slightly higher scores on social support from parents/guardians (t = 2.67, p ≤ .01) and from teachers (t = 2.58, p ≤ .01) compared to older adolescent. Similar results were observed on measures of psychological wellbeing. Specifically, younger adolescents reported slightly higher scores on self-concept (mean = 81.79 versus 79.58) and self-esteem (mean = 34.20 versus 33.75), and lower scores on measures of hopelessness (mean = 3.96 versus 4.53). The mean score difference between the two age groups for both self-concept (t = 3.26, p = .001) and hopelessness (t = -3.41, p = .001) were statistically significant.
3. Prevalence of depressive symptoms
Table 3: Prevalence of Depressive Symptoms Across Age groups
Depressive Symptoms
|
Total Sample (N=1260)
%(n)
|
Younger Adolescents:
14-15 years
(n=718) %(n)
|
Older Adolescents: 16-17 years
(n=542) % (n)
|
Minimal
|
22.14 (279)
|
22.70(163)
|
21.40(116)
|
Mild
|
31.83(401)
|
34.68(249)
|
28.04(152)
|
Moderate
|
29.68(374)
|
27.16(195)
|
33.03(179)
|
Severe
|
16.35(206)
|
15.46(111)
|
17.53(95)
|
As presented in Table 3, scores on the BDI scale were divided into four levels of depressive symptoms (minimal, mild, moderate and severe symptoms). In our sample, over one third (31.83%) of participants reported mild symptoms, 29.68% reported moderate symptoms and 16.35% reported severe symptoms. Within age groups, more younger adolescents than older adolescents reported minimal depressive symptoms (22.7% versus 21.4%) and mild symptoms (34.68% versus 28.04%). On the other hand, more older adolescents than younger adolescents reported moderate (33.03% versus 27.16%) and severe symptoms (17.53% versus 15.46%).
4. Predictors of depressive symptoms
Table 4: Regression on sociodemographic and household characteristics, family and social support factors, psychological wellbeing
Variable
|
Model 1:
B (95%CI)
|
Model 2:
B (95%CI)
|
Model 3:
B (95%CI)
|
Sociodemographic and Household characteristics
|
|
|
|
Age: (ref: 14 to 15 years)
|
|
|
|
16 to 17 years
|
1.15(0.02, 2.28) *
|
0.07(-0.96, 1.09)
|
-0.43( -1.41, 0.55)
|
Orphan hood Status: (ref: Orphan)
|
|
|
|
Non-orphan
|
-1.99(-3.56, -0.429) **
|
-1.32(-2.73, 0.09)
|
-0.89(-2.24, 0.45)
|
Primary Caregiver:
(ref: Biological parent)
|
|
|
|
Grandparent
|
0.59(-1.23, 2.42)
|
0.27(-1.36, 1.91)
|
0.48(-1.05, 2.02)
|
Other relative
|
-0.21(-2.01, 1.59)
|
-0.41(-2.02, 1.21)
|
-0.39(-1.90, 1.13)
|
Household size
|
|
|
|
Number of people in HH
|
0.29(-0.04, 0.62)
|
0.14(-0.16, 0.43)
|
0.09(-0.19, 0.38)
|
Number of children in HH
|
0.01(-0.41, 0.44)
|
0.08(-0.29, 0.46)
|
0.01(-0.35, 0.38)
|
Family Assets
|
-0.34(-0.51, -0.15) ***
|
-0.14(-0.30, 0.02)
|
0.07(-0.08, 0.23)
|
Family and Social Support Factors
|
|
|
|
Family cohesion
|
|
-0.0003(-0.10, 0.09)
|
0.03(-0.07, 0.12)
|
Family care and relationship
|
|
-0.38(-0.51, -0.24) ***
|
-0.24(-0.37, -0.11) ***
|
Perceived Child and Caregiver support
|
|
-0.12(-0.19, -0.03) **
|
-0.04(-0.12, 0.04)
|
Social support from multiple sources
|
|
|
|
Guardian
|
|
-0.31(-0.46, -0.17) ***
|
-0.16(-0.29, -0.016) *
|
Classmate
|
|
-0.14(-.30, 0.03)
|
-0.02(-0.18, 0.14)
|
Teacher
|
|
-0.13(-0.28, 0.01)
|
0.08(-0.06, 0.22)
|
Friend/peer
|
|
-0.27(-0.35, -0.19) ***
|
-0.12(-0.19, -0.04) **
|
Psychological wellbeing
|
|
|
|
Tennessee Self – Concept
|
|
|
-0.28(-0.33, -0.22) ***
|
Beck's Hopelessness Scale
|
|
|
0.56(0.35, 0.77) ***
|
Rosenberg Self Esteem Scale
|
|
|
-0.21(-0.33, -0.09) ***
|
The F-value
|
4.38***
|
25.67***
|
35.37***
|
R2
|
0.024
|
0.224
|
0.356
|
Adjusted R2(df)
|
0.02(7)
|
0.22(14)
|
0.35(17)
|
Change in R2
|
|
0.200***
|
0.132***
|
N
|
1260
|
1260
|
1106
|
*p≤.05; **p≤.01; ***p≤.001
Results from hierarchical regression analyses are presented in Table 4. Predictors were entered into the model as a block of predictors. In model 1, we control for participants’ sociodemographic and household characteristics (age, orphanhood type, primary caregivers, household size and family assets). In model 2 we add family and social support factors (family cohesion, family and care relationship, perceived child-caregiver support and social support from multiple sources). In model 3 we add measures of psychological wellbeing (self-concept, self-esteem and hopelessness). Age, orphanhood status and family assets were associated with depressive symptoms. Specifically, being an older adolescent girl (16-17 years) was associated with higher levels of depressive symptoms (B=1.15, 95% CI= 0.02, 2.28, p ≤.05). On the other hand, having both parents alive, i.e. non-orphaned adolescent (B=-1.99, 95% CI= -3.56, -0.429, p ≤.01) and higher levels of family assets (B= -0.34, 95% CI= -0.51, -0.15, p =.001) were both associated with lower levels of depressive symptoms
When we added family and social support factors in model 2, age, orphanhood status and family assets became non-significant. Family care and relationship (B = -0.38, 95% CI= -0.51, -0.24, p =.001), perceived child-caregiver support (B = -0.12, 95% CI= -0.19, -0.03, p ≤.01), social support from a parent/caregiver (B= -0.31, 95% CI= -0.46, -0.17, p =.001), and social support from a friend/peer (B = -0.27, 95% CI= -0.35, -0.19, p =.001) were all associated with lower levels of depressive symptoms among adolescent girls.
Finally, controlling for measures of psychological wellbeing (model 3), family care and relationship (B = -0.24, 95% CI= -0.37, -0.11, p =.001), social support from the parent/caregiver (B = -0.16, 95% CI= -0.29, -0.016, p ≤.05), social support from the peer/friend (B = -0.12, 95% CI= -0.19, -0.04, p ≤.01) remained significant predictors of lower depressive symptoms. In addition, while self-concept (B = -0.28, 95% CI= -0.33, -0.22, p =.001) and self-esteem (B = -0.21, 95% CI: -0.33, -0.09, p =.001) were both associated with lower levels of depressive symptoms, feelings of hopelessness were associated with higher levels of depressive symptoms (B= 0.56, 95% CI= 0.35, 0.77, p =.001).
The model containing sociodemographic and household characteristics accounted for 2.4% (R2 =.024) of the variance in depressive symptoms. When we added family and social support factors (model 2), we are able to explain 22.4% (R2 = .224) of the variance. The 20-percentage change between model 1 and model 2 was statistically significant (p =.001). Adding measures of psychological wellbeing (model 3) allowed us to explain 35.6% of the variance in depressive symptoms (R2 = .356). The 13.2 percentage change between model 2 and model 3 was statistically significant (p =.001).