Improving Mental Health of Mothers Living under Extreme Conditions: Evidence from Afghanistan

How to cite

Dinarte, Lelys; Asad, Saher; Reynoso, Ana; Shah, Manisha. 2026. "Improving Mental Health of Mothers Living under Extreme Conditions: Evidence from Afghanistan." AEA RCT Registry. AEARCTR-0017378. https://doi.org/10.1257/rct.17378-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Lelys Dinarte* , Saher Asad , Ana Reynoso , Manisha Shah

* Primary investigator

Registry ID
AEARCTR-0017378
DOI
https://doi.org/10.1257/rct.17378-1.0
First registered
Last updated
Status
On going
Trial dates
2025-08-01 – 2026-05-31
Countries
Afghanistan ()
JEL codes
I14, I15, I31, J16, J13
Registry record
http://www.socialscienceregistry.org/trials/17378

Abstract

This study provides experimental evidence on how to improve the mental health of mothers living under extreme conditions at scale. The intervention targets mothers of children aged 2–6 years in Afghanistan, a context characterized by conflict, severe constraints on women’s mobility and social interaction, and very limited access to formal mental health services.

The study addresses two main research questions. First, we examine whether the content of support matters for improving maternal mental health. We compare interventions that aim to improve mothers’ wellbeing indirectly, by strengthening household functioning through a positive discipline parenting program delivered by phone (via live calls or IVR), with interventions that aim to improve mothers’ wellbeing directly, by delivering mental health support through SMS messages adapted from the Thinking Healthy Program (THP), a cognitive-behavioral–based intervention simplified and contextualized for low-resource and low-literacy settings.

Second, we test whether the mode of delivery influences effectiveness. We study two mechanisms. The first is an isolation channel, comparing interventions delivered through live phone calls with a human voice versus automated IVR, to assess whether reducing social isolation improves maternal mental health. The second is a trust channel, varying the identity of the messenger delivering phone-based content—a professional profile versus a profile resembling a trusted community member—reflecting evidence that mothers in Afghanistan often place greater trust in peers or older women from their communities than in formal professionals.

The evaluation uses a randomized controlled trial design, varying the content and delivery modality. In total, 4,203 caregivers across six provinces are randomly assigned to different intervention arms or a pure control group. Together, the study sheds light on what types of content and delivery mechanisms are most effective for scaling maternal mental health and parenting support in settings with extremely low individual-level capacity and strong social constraints.

Intervention

The intervention consists of a remote program delivered to mothers of children aged 2–6 years through mobile phones. The program is designed to improve maternal wellbeing in a low-resource, low-literacy, and high-constraint setting. All intervention content was adapted to the Afghan context, translated into Dari and Pashto, and pretested for clarity and cultural appropriateness. The program includes two components:

Maternal wellbeing component (direct channel)
This component consists of a maternal mental health support intervention. Mothers receive short SMS messages focused on emotional wellbeing, stress management, and cognitive coping strategies. The messages are adapted from the Thinking Healthy Program (THP), an evidence-based intervention grounded in cognitive behavioral therapy and simplified for low-resource and low-literacy settings. The SMS content reinforces themes such as recognizing unhelpful thoughts, problem-solving, self-kindness, and seeking social support when possible.

Parenting support component (indirect channel)
This component focuses on positive parenting and nonviolent discipline. Mothers receive weekly calls lasting approximately 15 minutes, delivered either through live phone calls or via interactive voice response (IVR). The parenting content emphasizes: Warm and responsive caregiver–child interactions, play and positive engagement, emotional regulation for both caregivers and children, clear communication and managing challenging child behaviors. The parenting component is designed to improve mothers’ ability to manage daily caregiving demands and child behavior, with the objective of indirectly reducing maternal stress and psychological distress.

The delivery of the intervention varies along two dimensions. First, the delivery modality varies between live phone calls with a human voice and automated IVR delivery, allowing the study to assess whether interpersonal contact reduces social isolation and improves outcomes. Second, the identity of the messenger varies between a professional profile (trained staff delivering content in a professional tone) and a community-referent profile (individuals whose voice and presentation resemble a trusted community member). Both messengers deliver identical scripted content and receive standardized training.

Experimental design

The study includes a total sample of 4,203 mothers, randomly assigned at the individual caregiver level. Randomization occurs in two stages to address two distinct research questions: (i) the effectiveness of different delivery modes and messenger identities, and (ii) the effectiveness of different content bundles.

Stage 1: Delivery mode and information source identity
To study whether the mode of delivery and the identity of the information provider affect outcomes, participants are first randomly assigned to one of the following five groups:
• Phone-based calls delivered by a professional (≈ 750 mothers)
• Phone-based calls delivered by a community referent (≈ 750 mothers)
• IVR-based calls delivered by a professional (≈ 750 mothers)
• IVR-based calls delivered by a community referent (≈ 750 mothers)
• Control group (≈ 1,200 mothers), receiving no phone-based parenting content
The first four groups receive the same 10-week parenting curriculum, differing only in delivery modality (live call vs. IVR) and messenger identity (professional vs. community referent).

Stage 2: Content assignment (Parenting only vs. parenting + mental health SMS)
To study whether adding direct maternal mental health support improves outcomes, a second randomization assigns participants to receive SMS-based mental health messages in addition to their initial assignment described in stage 1.
Among the four parenting groups described above, half of participants are randomly assigned to receive parenting support only, while the other half receive parenting support plus SMS-based maternal mental health messages adapted from the Thinking Healthy Program (THP). Within the original control group, half of participants are randomly assigned to receive SMS-based mental health messages only, while the remaining half continue as a pure control group.

This yields the following distribution across content groups:
• Parenting program only: ≈ 1,400 mothers
• Parenting program + SMS mental health support: ≈ 1,400 mothers
• SMS mental health support only: ≈ 600 mothers
• Pure control group: ≈ 600 mothers

Data Collection: Baseline survey with mothers will be conducted prior to the intervention. Endline survey with mothers will be conducted after the program has been completed. Throughout the intervention, the team will also collect process data on take-up and compliance, such as call completion rates and SMS reception. In addition, a shorter endline survey will be collected from a secondary caregiver in the household. Finally, a short survey of the implementing staff (professionals and community referents) will capture their experiences delivering the intervention. All data collection instruments will be translated into Dari and Pashto and pretested for linguistic and cultural appropriateness. A longer-term follow-up survey may be conducted one year after the intervention, contingent on field conditions and funding.

Design details

Randomization method
Randomization done in office using a code designed by the research team.
Randomization unit
Individual (mother)
Clusters
4,203 Afghan mothers
Observations
4,203 Afghan mothers
Arms
The distribution of the sample sizes was described before and summarized as follows: For the delivery mode research question, participants are allocated across four parenting intervention arms and a control group—varying by delivery modality (live phone call versus IVR) and messenger identity (professional versus community referent)—as follows: • Phone based-call delivered by a professional (~750 mothers) • Phone based-call delivered by a referent (~750 mothers) • IVR-based call delivered by a professional (~750 mothers) • IVR-based call delivered by a referent (~750 mothers) • Control Group (~1,200): Mothers will not receive any information. For the content research question, a second randomization assigns participants to receive SMS-based maternal mental health messages in addition to their initial assignment, as follows: • 1,400 were assigned to parenting program only (1/2 of the first 4 groups described above) • 1,400 were assigned to both (1/2 of the first 4 groups described above) • 600 were assigned to SMS messages only (1/2 of the control group) • 600 remained as pure control group.
Primary outcomes
Maternal mental health
Secondary outcomes
Compliance with phone calls and SMS messages, perceptions about phone calls and SMS messages, knowledge index. Intra-household interactions/decision making indicators, parental self-efficacy/self-confidence, female empowerment indicators, positive discipline indicators, child behavior, parent-child interactions.
Sponsors
Name: The World Bank Location: United States Url: https://www.worldbank.org/ext/en/home
Partners
Name: UNICEF Partner_type: ngo Url: https://www.unicef.org/afghanistan/

Cited by

No citation data available for this registration yet.

Keywords