Measures

For parental depression, child behavioral outcomes and demographic/family functioning measures (available in English and Luganda), all have been validated with Ugandan population [26–28]. For parenting measures, we selected measures that have been used with diverse cultural groups from parenting literature [29–32]. The Luganda version of parenting assessments were translated based on the recommended method suggested in the literature (i.e., applying translation and back-translation, and using a team review approach to resolve any discrepancies between the versions and to determine whether the translated material is appropriate and meaningful for English and Luganda speakers) [33, 34]. We carefully examined the psychometric properties of each scale to ensure its reliability and validity in the present study sample, as describe below.

The Patient Health Questionnaire (PHQ-9; 10 items; α = .83 using our Ugandan sample) [35, 36], a brief depression screening measure, was used for this study. PHQ-9 has been widely used in many countries and validated with Ugandan adults [26]. Parents rated 9 symptom items over the last 2 weeks on a 4-point scale (0 = not at all; 3 = nearly every day; sample item: “little interest or pleasure in doing things” “thoughts that you would be better off dead or of hurting yourself in some way”). The scale also includes an overall functioning rating that evaluates the level of functional impairment (0 = not at all difficult to function, 3 = extremely difficult to function). A total score was created for nine symptom items. PHQ-9 has been validated previously based on Ugandan samples [26, 27], using the Mini-International Neuropsychiatric Interview (MINI; a widely used short structured diagnostic interview for DSM-IV and ICD-10 psychiatric disorders) [37]. The clinical cut-off score of 10 has been suggested for the Ugandan population (with sensitivity of 0.91 and specificity 0.81). Individuals with a score of 10 or above would suggest a high likelihood of having a depressive disorder [26]. The 10 cut-off is compatible to the developers’ version, which defines 1–4: minimal depression, 5–9: mild depression, 10–14: moderate depression, 15–19: moderately severe depression, and 20–27: severe depression. For the purpose of this study, we defined a score of 10 or above as “depressed” and a score below 10 (1–9) as “non-depressed.” The dichotomized scale (or dummy variable) was used for analyses.

Four measures were used to assess four areas of parenting. Constructs that had been identified as significant predictors for child development in the literature were targeted. The Parent–Child Relationship Scale [38] (12 items) measures Conflicted Parent–Child Relationship (e.g., child and I always seem to be struggling with each other) on a 5-point scale (1 = definitely does not apply; 5 = definitely applies). The Parenting Practices Interview (PPI) [39] assessed Harsh and Inconsistent Discipline (11 items; e.g., “how often do you slap or hit your child when he/she misbehaves,” “how often does your child manage to get around the rules you set”) on a 4-point scale (0 = never, 3 = Often). Both Parent–Child Relationship and Harsh and Inconsistent Discipline scales have been used with diverse ethnic populations in studies conducted in HICs, and showed adequate reliability in this Ugandan study sample (α = .72 and .66, respectively). Parent Involvement in Education was assessed using 2 items. Parents were asked to rate how often they help their child with school-type activities (e.g., reading or discussing a story together, working on a project together) on a 7-point scale (1 = never, 4 = a few times per month, 7 = everyday). They were also asked to rate the number of total hours they help their child with any education related activities (e.g., spend time talking about school activities, doing homework, reading together) in the last two school days (1 ≤ 0.5 h, 4 = 1.5–2 h, 7 = 3 or more hours). The two items were correlated (r = .23, p < .001); therefore, we combined the items into one scale. To consider information bias (e.g., potential under-report of harsh discipline use by parents), we also included the Attitude About Corporal Punishment scale (3 items; α = .78), which is a measure we developed for this study. This measure assesses parents’ attitude toward corporal punishment (e.g., in order to bring up a child properly, you need to physically punish your child) on a 5-point scale (1 = strongly disagree, 5 = strongly agree). This attitude scale considers a proxy measure for parental use of corporal punishment (r = .14, p < .05).

Four scales were used to assess child health and development. The Strengths and Difficulties Questionnaire (20 items) [40, 41], is a brief screening measure for child mental health problems including ratings of emotional symptoms, conduct problems, hyperactivity, and peer problems (e.g., “Often fights with other children or bullies them,” “Often unhappy, down-hearted or tearful”). The measure has been widely used for research purposes in the United States and other high and LMICs [42–46]. The parent version was used in this study. Parents responded in terms of how true each behavior was for their child on a 3-point scale (0 = not true, 2 = certainly true). A total difficulty sum score, ranging from 0–40, was created. A score of 0–13 is considered normal, a score of 14–16 is considered borderline, and a score of 17–40 is considered abnormal. In this Ugandan sample, internal consistency is adequate (α = .63), and about 33 % of children were rated to have abnormal levels of problem behaviors (relative to 6–7 % for the US 4–10 year olds using the same measure) [47]. The Social Competence Scale (12 items; e.g., “shares things with others,” “copes well with failure”; α = .80 for Ugandan sample) [48, 49], evaluates children’s positive social behaviors, including emotion regulation, prosocial behaviors, and communication skills. Parents were asked to rate how well the statements describe their child on a 5-point scale (0 = not at all to 4 = very well). Child Health was measured with two global items. Parents were asked to rate their overall perception of their child’s health on a 4-point scale (1 = poor health; 4 = very good or excellent health) and their perception about their child’s tendency to get physical illness on a 3-point scale (1 = certainly true, 3 = not true). Both items are statistically significantly related (r = .52; α = .68); therefore, we created a sum score. A higher score indicates good health. The Child School Functioning scale (3 items; α = .52), developed for this study, measures children’s difficulty in school (i.e., lacks motivation/enjoyment in going to school, complains about school, and has trouble with school grades). Parents were asked to rate these behaviors on a 3-point scale (1 = not true, 3 = certainly true), and a sum score was then created; a higher score indicates poor school functioning.

Additional test–retest reliability (for a subsample of 42 families) and concurrent validity (assessed with correlations) among key study constructs were examined. Test-test reliability (assessed 6 month apart using Pearson correlations) show adequate reliability for the parenting measures (rs for 4 measures ranged from .36 to .59) and child functioning measures (rs for 4 measures ranged from .36 to .50) Inter-correlations within domains (i.e., parenting, child outcomes) and across domains were also correlated in expected directions (see Table 2), suggesting support of concurrent and construct validity of study measures for use in Uganda. For example, harsh/inconsistent parenting, value of corporal punishment, and conflicted parent–child relationship were related to poor child physical health, school functioning and higher problem behavior. Better child health was related to fewer problem behaviors and higher social competence. Higher parental attitude about corporal punishment and harsh/inconsistent discipline were associated with higher conflicted parent–child relationship.

Correlations among study variables

P–C = Parent Child. Corporal punishment = attitude about corporal punishment.

To consider demographic predictors of parental depression and potential confounders, parent education, employment status, household size, religion, food insecurity, parental health, and social support were included. Categories for parental education, employment status, and religion (including the reference categories) are specified in Table 1. Religious groups were considered because of the diverse religious background in Uganda, and their potential impact on psychological well-being in both adults and children. Food insecurity (3 items; α = .85 for the study sample) was assessed using the Household Hunger Scale [50], which evaluates family food resources and hunger status over the past 4 weeks on a yes/no dichotomized scale (e.g., no food to eat because of lack of resources to purchase food; any household member going to sleep at night hungry, any household member going a whole day and night without eating anything at all because there was not enough food in the home). If any food resource or hunger issue was reported (or the sum score equal or above 1), the family would be grouped into the food insecure group. Parental health was assessed based on parent perception of overall health and quality of life (2 items; α = .84 for the study sample) on a 5-point scale (1 = poor, 5 = excellent). Social Support (4 items; α = .85 for the study sample) was evaluated using the Multidimensional Scale of Perceived Social Support [51, 52]. The scale evaluates perceived support for comfort, sharing emotion, and needing help on a 5-point scale (1 = strongly disagree, 5 = strongly agree).