Ethiopian DHS Finds Functional Difficulty Concentrated in Affective Symptoms and School Absence
A nationally representative survey in Ethiopia has placed functional difficulty in children and adolescents aged 5 to 17 years squarely on the public health agenda. Using the 2024-2025 Demographic and Health Survey and the UNICEF/Washington Group Child Functioning Module, investigators estimated that 9.5% of children had difficulty in at least one functional domain or reported daily anxiety or depression. The burden was not evenly distributed. Emotional symptoms dominated the case mix, while non-affective limitations such as sensory, physical, or cognitive difficulty were much less common.
Daily depression was reported in 6.8% of children and daily anxiety in 6.6%. Among those identified as having any functional difficulty, 70.3% had only affective difficulty. In practical terms, the survey did not find a broad pattern of severe physical disability across the population; it found a much larger signal of internal distress. That pattern matters for schools, families, and primary care teams, because children with emotional symptoms can remain invisible until concentration, attendance, or daily functioning begins to slip.
Mental Distress Dominated the Functional Profile
The study included 11,122 children, with weighted analysis based on 10,726 children and 4,739 mothers. Functional difficulty was defined conservatively as “a lot of difficulty” or “cannot do at all” in any of 11 domains, or daily anxiety or depression. By that standard, difficulty in at least one non-affective domain affected 2.8% of children, and each individual non-affective domain ranged from 0.5% to 1.4%.
The narrow spread across non-affective domains suggests a relatively low prevalence of severe sensory or physical limitation in the national sample, but the affective signal was large enough to reshape the overall prevalence estimate. Daily emotional symptoms have different service implications from mobility impairment or vision loss. They are less likely to trigger referral through routine disability pathways and more likely to pass through schools unnoticed unless teachers, caregivers, or community health workers ask directly about mood, worry, sleep, and participation.
Prevalence also varied sharply by region, ranging from 1.5% in Southwest Regional State, Ethiopia, to 16.5% in Oromia Regional State. The survey does not explain why the regional spread was so wide, but such differences often reflect a mix of access to services, social stressors, household structure, school participation, and local reporting patterns. When a countrywide estimate spans that kind of range, national averages can obscure pockets of much higher need.
School Attendance Emerged as the Strongest Modifiable Signal
In adjusted multilevel models, children not attending school had higher odds of non-affective difficulty, with an adjusted odds ratio of 2.82 and a 95% confidence interval of 2.01 to 3.96. Girls had lower odds of non-affective difficulty than boys, with an adjusted odds ratio of 0.63 and a 95% confidence interval of 0.48 to 0.84. The sex association was specific to non-affective difficulty and did not shift the overall finding that school absence marked a higher-risk group.
School non-attendance can operate in both directions. Children with unrecognized impairment may leave school because they cannot keep up with the physical, cognitive, or social demands of the classroom. At the same time, children already out of school may lose the structure, observation, and support that often bring functional problems to attention. Either pathway leaves the same bedside and policy consequence: a child outside the education system is less likely to be screened, documented, or linked to care.
One result stood out with particular force. Outcomes clustered strongly within families and clusters, with sibling correlation at 67.7% and cluster intraclass correlation at 20.9%. Functional difficulty was not randomly scattered across the population; it was concentrated in households and local settings. Such clustering is consistent with shared social exposures, caregiving context, environmental disadvantage, or interviewer-related measurement effects. It also means that family-level and community-level interventions may outperform approaches aimed only at individual children.
Interviewer Effects Changed the Prevalence Estimate
The analysis also exposed a striking measurement issue. Six of 163 interviewers, representing 7.9% of the weighted sample, accounted for 58.2% of weighted cases of daily anxiety or depression. In the same clusters, those interviewers recorded daily symptoms for 56.4% of children, while other interviewers recorded 2.1%. Excluding 11 statistically outlying interviewers reduced the prevalence of any difficulty to 4.6% and daily anxiety or depression to 3.3%.
That kind of spread raises immediate questions about how sensitive caregivers were to symptom wording, how interviews were conducted, and whether some enumerators elicited distress more readily than others. The associations with school non-attendance and sex remained unchanged after removing outlying interviewers, which suggests the main risk patterns were stable even as absolute prevalence shifted. Still, the size of the interviewer effect shows how a household survey can be shaped by the person asking the questions as much as by the child or caregiver answering them.
What the Survey Means for Child Health Services
The most useful reading of these data is not that Ethiopian children are broadly disabled, but that functional difficulty in this age group is often emotional, frequently concentrated in out-of-school children, and unevenly detected across interviewers and clusters. Pediatric and school health programs need a framework that recognizes anxiety and depression as functional problems, not only as psychiatric labels. In a setting where most affected children had affective rather than non-affective difficulty, screening programs limited to visible disability would miss much of the burden.
Resource allocation should follow the shape of the problem shown here. School re-entry pathways, adolescent mental health screening, caregiver education, and community-based identification may all be more relevant than a narrow focus on physical disability alone. The survey’s cross-sectional design limits causal interpretation, and the prevalence estimates remain dependent on how the functioning module was administered. Even so, the national data provide a clear signal: children outside school and children with emotional symptoms are carrying a disproportionate share of functional difficulty, and service planning will need to account for both.