From Childhood Abduction to Adulthood: What 20 Years Reveal About Women in Northern Uganda

Between the mid‑1980s and the mid‑2000s, northern Uganda was gripped by devastating conflict as the Lord’s Resistance Army (LRA) abducted tens of thousands of children, many of them girls who were forced into domestic labor and coerced marriages. While the immediate horrors caused by these abductions have been documented, less is known about how the formerly abducted are doing today. In a new study, we ask: what does life look like for the northern Ugandan women who were taken as children and are now adults raising families of their own?

Drawing on original survey data and incentivized behavioral experiments with more than 500 women in northern Uganda, we trace the enduring consequences of childhood abduction nearly 20 years after the conflict ended. What we find is both sobering and policy-relevant: trauma from conflict has lasting psychological and social effects that may be missed by standard economic measures. Addressing these effects requires multidimensional interventions that combine economic support with accessible, culturally appropriate mental health care and efforts to rebuild social ties.

A rare opportunity for insight into the effects of war

Studying the long‑run impacts of violence and trauma experienced in childhood is notoriously difficult. Who would these women have become had they not been abducted? We build on an insight from earlier work by Jeannie Annan and Chris Blattman which showed that LRA abductions were largely indiscriminate: raids occurred at night, across villages, and targeted whoever happened to be present. Because abductions happened largely at random, we can compare formerly abducted women with never‑abducted women from the same villages.

Our sample includes 541 women aged 18–54, nearly half of whom report having been abducted as children or adolescents. The data illustrate two critical features of these experiences: first, most women were abducted very young (often before age 13). This implies that the vast majority of abductions took place during childhood–years when everything from educational attainment and human capital accumulation to emotional regulation and stress responses are being formed. Second, the abduction lasted on average slightly more than a year and a half, with 45 percent of abductees kept for less than a year, and 55 percent at least one year.

Figure 1. Age when abducted and time in abduction

Image

 

Image

 

Economic outcomes: Less change than expected

Perhaps surprisingly, the economic impacts of abduction are quite muted: we find little difference between abducted and non‑abducted women in terms of current income, food insecurity, or water scarcity.

This does not mean abduction was economically inconsequential. Formerly abducted women are significantly less likely to have completed primary school, a gap that may partly reflect years spent in captivity during school age. Yet in a region where nearly everyone relies on subsistence agriculture and paid employment is rare, lost schooling does not translate neatly into lower measured income.

Family life and social support: Some evidence of impacts

More differences emerge in family structure and social connections. Formerly abducted women have more biological children on average and report lower levels of social support, as measured by a validated social support scale. Social support matters, not only for well‑being in its own right, but because it buffers stress and depression. Its erosion may therefore be one channel through which early trauma continues to shape mental health decades later.

Mental health: Scars that do not fade

But our most striking results concern mental health. Using the Edinburgh Postnatal Depression Scale and the Perceived Stress Scale—both validated in low‑income and conflict‑affected settings—we show that psychological distress is pervasive across the entire sample. Even among women who were never abducted, rates of likely depression and severe stress are alarmingly high–a stark reminder of the indirect effects of conflict in a region that was devastated by the LRA insurgency, as well of the effects of poverty on mental health.

But formerly abducted women fare significantly worse.

Our results (Figure 2 below) reveal that, two decades after returning home, formerly abducted women are about 25 percent more likely to meet the threshold for likely depression and 40 percent more likely to experience severe stress than their non‑abducted peers. These gaps persist even after correcting for multiple hypothesis testing and are largest among women abducted at younger ages.

Figure 2. EPSD and PSS-10

Image

 

Image

 

Note: Panel (a) and Panel (b) display the Kernel densities of the two measures of mental health: the Edinburgh Postnatal Depression Scale (EPDS) and the Cohen stress scale (PSS-10), respectively. The vertical dashed lines indicate the threshold levels used to identify high likelihood of clinical depression for the EPDS index and severe stress for the Cohen index.

How women respond to stress

A novelty of our study is the analysis of how women respond to stress. Drawing on psychological theories of fight‑or‑flight and tend‑and‑befriend responses, we construct indices capturing four behavioral reactions to threat and find two key patterns, presented in Figure 3. First, tend‑and‑befriend responses—seeking closeness, caring for others, and strengthening social ties—are more common than testosterone-driven fight‑or‑flight reactions among women overall. Second, childhood abduction amplifies both types of responses. Formerly abducted women score higher on fight, flight, and befriend indices, suggesting heightened physiological and emotional reactivity to stress.

This combination is telling. Rather than produce emotional numbing, abduction appears to leave women more reactive—simultaneously vigilant and socially oriented—long after the original threat has passed.

Figure 3. Stress response: Tend, befriend, fight, flight

Image

Note: We report the average score for each of the four stress responses. Each index sums the answers to 4 questions eliciting how often the respondent reacts to stressful situations in a given way, on a 5-point Likert scale (1 = Never to 5 = Always). Therefore, each index ranges from a minimum of 4 to a maximum of 20 points.

Grit, competition, and prosociality

We also use incentivized behavioral games to measure non‑cognitive traits—a relatively rare feature when studying previously abducted populations.

Results show that formerly abducted women display greater grit, persisting more often on difficult tasks with higher potential rewards. This finding resonates with theories of post‑traumatic growth and resilience: surviving extreme adversity can strengthen perseverance.

At the same time, the evidence suggests lower prosociality and higher competitiveness among abducted women, though these effects are more modest. Notably, we find no significant difference in risk preferences, adding nuance to a literature that has produced mixed findings on violence and risk‑taking.

Putting it all together

Together, these results suggest that trauma is not simply destructive. It reshapes people’s capacities, sometimes in adaptive ways, sometimes in ways that carry lasting social costs.

These persistent effects also underscore a central message: psychological trauma from childhood victimization does not simply “heal with time.” In contexts of chronic poverty and limited access to mental health care, early trauma can shape emotional well‑being throughout life.

Our results also highlight that in settings where opportunities are universally scarce, even large shocks to human capital may not be captured by the standard economic measures of well-being.

Why this evidence matters

These findings deliver a clear policy message: recovery from conflict cannot be measured solely in terms of income or employment. Even when women appear economically “reintegrated,” the psychological and social costs of childhood victimization may remain severe. And this is not just a story from the past: child abductions are still tragically commonplace in conflicts around the world today.

The coexistence of heightened grit and elevated depressive symptoms is especially instructive. Resilience should not be mistaken for well‑being. Programs that rely on women’s perseverance while ignoring mental health may risk deepening invisible scars.

Two decades after the LRA left northern Uganda, its legacy of abduction lives on in the daily emotional experiences of the girls–now women– that they abducted. Addressing that legacy requires policy consideration of the impacts on long-term mental health, not just economic wellbeing. Our team has done more research in this context and will be reporting back on potential interventions that work–watch this space.