Background
The Joint United Nations Programme on AIDS estimates that 88% of the global 3.3 million children and adolescents living with HIV (CA-HIV) reside in Sub-Saharan Africa [
1]. Most of these CA-HIV who were perinatally infected with HIV are at risk of developing emotional and behavioural problems [
2]. Although many countries in Sub-Saharan Africa including Uganda [
3] have adopted the WHO policy recommendation [
4] that calls for the integration of mental health services into HIV care, most HIV clinics on the continent have yet to implement them [
5].
In Uganda, prevalence of HIV in the general population has stabilised around 7.4% [
6], with highest rates in the Central Region (10.6%) and the lowest in Mid-Eastern Region (4.1%). Prevalence is higher among women (8.3%) than men (6.1%); 3.7% of young women and men between 15 and 24 years of age are HIV-positive [
3]. There is fairly compelling evidence that HIV contributes to neurocognitive impairment and possibly neuropsychiatric illnesses [
7‐
10]. CA-HIV experience relatively high rates of psychopathology [
11‐
15] that could arise from the direct and indirect effects of HIV including the psychosocial problems of family disruptions; poor social support; and a high burden of negative life events, stigma, poverty and chronic ill health [
16] and the burden of long-term treatment with antiretroviral therapy (ART) [
17].
Previous studies have found that ADHD is one of the most common psychiatric disorders among CA-HIV (Scharko, as cited in Mellins and Malee, 2013) [
2]. However, little is known about ADHD among CA-HIV in Sub-Saharan Africa. Most of what is known comes from two studies [
15,
18] conducted in Kenya and South Africa. The Kenyan study reported a prevalence of 12.2% as based on both parent and youth report [
18], whereas the South African study found rates of 88% and 17% based on parent and teacher ratings, respectively [
15]. The rates of ADHD reported in the Kenyan and South African studies differed significantly possibly due to differences in the assessment instruments used. Whereas the Kenyan study used the MINI International Neuropsychiatric Interview for children and adolescents (MINI-KID) [
18], a
DSM IV based psychiatric assessment tool that undertakes both a symptom count and assessment of functional impairment, the South African study used the Swanson, Nolan and Pelham rating scale [
15], a
DSM IV-based psychiatric assessment tool that only assesses symptom count. To date, no study has established the prevalence of ADHD among CA-HIV in Uganda.
Although it is widely recognized that ADHD is highly heritable [
19], a wide range of variables are associated with symptom severity in CA-HIV [
2], and these include impaired social and academic functioning [
2,
17,
20‐
24], which can negatively affect quality of life. However, relatively little is known about child, family, and illness characteristics associated with severity of ADHD among CA-HIV in East Africa or their implications for functional outcomes. The primary objectives of the present study were to (a) document the prevalence of ADHD among a large sample of CA-HIV attending rural and urban HIV clinics in Uganda, (b) characterize the relation of ADHD severity with a wide range of commonly studied clinical correlates, and (c) examine the relation of ADHD with important indices of academic and social functioning. We were particularly interested in ADHD because it is associated with emotional and behavioural dysregulation thus increasing the risk of impaired decision making, poor impulse control, risky sexual behaviour, and aggression among CA-HIV [
2].
Discussion
This study aimed to determine the prevalence of ADHD among CA-HIV in Uganda and associations of ADHD with commonly studied mental health clinical correlates and social, academic and clinical functioning. The prevalence of caregiver-rated ADHD was 6%, which is similar to caregiver-reported CASI ADHD rates for non-HIV, community-based samples in the United States [
40,
41]. When we adopted a more clinically oriented criterion for any ADHD based on both symptom count cut-off score and impairment cut-off score, the prevalence of any ADHD was 2.2%, which underscores the importance of considering criteria for defining ADHD when comparing results across studies. With regard to CA-HIV, a study conducted in the United States that used CASI [
17,
35] symptom count cut-off scores found that 16% met criteria for ADHD according to caregiver ratings [
36]. These findings seem to suggest that the rate of ADHD in the United States study is much higher than that in Uganda, at least among HIV populations. However, Gadow et al. [
17] also found the prevalence of ADHD among uninfected youth in the United States from HIV-affected families was similar to CA-HIV (12.6%), and rates for both groups (infected and affected) were higher than comparable rates for population-based samples [
40,
41]. The authors noted that one plausible explanation for these differences between infected/affected youth and population-based samples in the United States is the much higher percentage of minority youth among the former and the well-established relation of minority status with mental health concerns, clinical correlates, and less than optimal health care. To this we would add that socio-cultural differences in symptom grading among caregivers in more industrialised Western countries as compared to the more agrarian Ugandan environment (where the caregivers are expected to have a tolerance for psychiatric symptomatology as compared to their Western counterparts) may also contribute to disparate rates of ADHD.
In Uganda the most common ADHD presentation was ADHD-I, followed by ADHD-HI with ADHD-C being the least prevalent. This relative distribution of ADHD presentations was similar for children and adolescents with the exception that rates of ADHD-HI and ADHD-C were identical among adolescents. In the United States study conducted by Gadow and colleagues [
17], among their younger cohort (6–12 years old) both ADHD-I and ADHD-HI were tied in first place. Among their older cohort (13–18 years old), ADHD-I was more common than ADHD-HI and ADHD-C, rates for which were identical. Conversely, the South African study by Zeegers and colleagues [
15] found that caregiver ratings indicated ADHD-HI and ADHD-C were the most and least common, respectively, but teacher assessments of the same sample indicated ADHD-C was the most frequent presentation with ADHD-I and ADHD-HI both coming in second.
In our Uganda study, several clinical correlates were associated with the different ADHD presentations. Increasing age was associated with increased risk for ADHD-I, which is consistent with the findings from review of literature from nine studies conducted in Africa [
42] and is likely explained by the increasing demands school curricula as youth grow older. We observed an increase in the risk of ADHD-I with increasing socio-economic status whereas previous studies of non-HIV samples conducted in the West have reported just the opposite [
35,
37]. A possible explanation for our Uganda finding is that more affluent families are better able to pay school fees, and the school is arguably the primary setting in which ADHD-I symptoms are most problematic. In addition, higher socio-economic status is associated with caregiver educational attainment and greater sensitivity to their child’s academic success. Consistent with this interpretation, there was a marginally significant association between increasing caregiver educational status and rates of ADHD-I.
Increasing caregiver psychological distress was associated with higher rates of all ADHD presentations. Similar results have previously been reported by others both among an HIV [
2] and a non-HIV [
21] population in the West. Possible explanations suggested by others include heritability of ADHD (or any other mental health problems), stressful family and social environments associated with ADHD (or any other mental health problems), and the erosion of parenting capacities which often accompany a parent with HIV or mental illness [
2,
21,
43]. In support of the latter, we found that deteriorating quality of the child-caregiver relationship (communication style and emotional reaction between the child and caregiver) was associated with all three ADHD presentations, which is consistent with the findings of others for both HIV [
43] and non-HIV [
44] samples. Suffering physical abuse (i.e., having ever been beaten) was associated with a fivefold increased risk of ADHD: HI in our adolescent-age sample, which is in accord with the extant literature [
44]. It has been suggested that ADHD symptoms may illicit feelings of hostility in the caregiver causing them to use aversive behaviour management strategies with CA-HIV [
44].
Additional variables associated with ADHD were orphanhood and CD4 counts. Orphanhood, which not surprisingly was highest among those who reported loss of both parents, was marginally associated with ADHD-C. Previous studies among HIV populations have reported the association between orphanhood and youths’ mental health [
45,
46]. Decreasing CD4 counts were marginally associated with increase in the odds of ADHD-I, and prior studies undertaken in the United States report similar findings [
2,
10,
35,
36,
43,
47,
48].
ADHD can have a substantial impact on school functioning and is associated with poor exam performance, grade retention, and failure to graduate from secondary school [
49]. Cognitive deficits that may interfere with academic performance have commonly been reported among CA-HIV [
22,
35,
50]. In our Ugandan sample, we found that any ADHD presentation and ADHD-I specifically were associated with poor academic performance. In addition, having any ADHD presentation was associated with problems at school, which is in accord with prior research involving CA-HIV [
2,
35].
Early onset of sexual intercourse (only assessed among adolescents) was marginally associated (
p = 0.07) with a two- to threefold increase of having any ADHD or ADHD-I (aORs of 2.93 and 3.43, respectively). Studies conducted elsewhere have reported a positive association between ADHD and risky sexual behaviour [
21,
23,
35], which may have important implications for disease transmission. Although others have reported that among CA-HIV, co-occurring ADHD is associated with poorer adherence to HIV treatment, number of hospital visits, and hospital admissions [
43,
50], this was not the case in our study.
This study has several strengths to include an exceptionally large sample of CA-HIV, comprehensive assessment battery, and cross-cultural orientation, there are also limitations. Because our analyses are cross-sectional, we cannot comment on the casual directions, but these will be addressed in future publications about the longitudinal component of the CHAKA study. Since ADHD symptoms are influenced by environmental variables and therefore different informants can disagree about symptom severity [
51], our prevalence rates for ADHD in Uganda should be considered conservative estimates as we did not obtain CASI-5 ratings from our sample’s school teachers. Because we do not have a comparable sample of seronegative youth from the same geographic areas and environment, it is not possible to know whether relations between ADHD and putative mental health factors and functional outcomes are influenced by HIV status. This does not, however, detract from the clinical implications of our findings for CA-HIV. By design, the present study focused on CA-HIV living in Uganda, and owing to considerable cultural variation in East Africa, our results may not generalize to other countries in the region.
Conclusions
In summary, approximately 6% of CA-HIV living in Uganda met DSM-5 symptom count criteria for ADHD. Clinical ccorrelates of ADHD were reported for all domains (general socio-demographic, caregiver, psychosocial environmental and HIV illness). ADHD among CA-HIV was associated with poorer academic performance, school disciplinary problems, and earlier age of onset of sexual intercourse, all of which may have important implications for clinical management, quality of life, long-term outcome and possible disease transmission. Moving forward, there is a definite need to integrate mental health services into routine HIV care to include the development of cost-effective assessment and treatment strategies that have high probability of success in challenging intervention settings.
Acknowledgments
The authors wish to thank the managers of the five study sites (Lubowa Joint Clinical Research Centre, Nsambya homecare department Children’s HIV Care clinic; Nsambya hospital, the Children’s clinic at The AIDS Support Organisation; TASO Masaka, Uganda Cares/Masaka Regional Referral Hospital and Kitovu Mobile AIDS Organisation, Masaka) for permitting the study to be conducted at their specialised HIV/AIDS clinics. The authors extend appreciation to the Medical Research Council, Uganda (MRC, Uganda) for funding and facilitating the study. Special gratitude is extended to the staff working at the five specialised HIV/AIDS clinics where the study was conducted. Appreciation is extended to the diligent work of research assistants. Appreciation is extended to the participants for their time and trust.