**Background:** Childhood obesity is a global public health concern, with adiposity in childhood predicting poorer health trajectories and NCD risk, particularly in LMICs. Pre-pregnancy obesity is predictive of obesity in children, and promoting healthy behaviours in young women of child-bearing age can improve their own health and increase the probability of healthier pregnancies and offspring. However, preconception health remains understudied and under-resourced in LMICs. The Healthy Life Trajectories Initiative (HeLTI) is an international consortium testing the hypothesis that an integrated complex intervention from preconception through early childhood will reduce childhood adiposity and NCD risk. HeLTI South Africa's Bukhali trial is being conducted in Soweto, a densely populated, low-income, urban setting. Epidemiological data show that 67% of young women (18–28 years) are overweight or obese during pregnancy, and if a girl was obese at 5 years, she had a 42 times greater risk of being an obese adult. Additional risks include unhealthy diet, high sedentary behaviour, physical inactivity, and high prevalence of antenatal depression (27%) and anxiety (15%). Formative qualitative work highlighted socioeconomic challenges, normalisation of obesity, safety concerns, complex family dynamics, and a 'preconception knowledge gap'. Pilot data (~1600 women aged 18–25) showed 27% classified as socially vulnerable (increasing to 44% for women with one child and 64% for those with more than one child), 33% food insecure, 20% at risk of food insecurity, and 44% overweight/obese. Less than half reported leisure-time physical activity, a third reported poor quality sleep, 19% had depression, 15% had anxiety, and 24% were at risk for harmful alcohol use.
**Methods:** The Bukhali intervention was developed using the UK MRC Guidelines for Complex Interventions and is theoretically grounded in behaviour change principles. It is delivered individually by 'Health Helpers' (CHWs) who are women aged 23–43 years (mean age 31.9 ± 5.21). The intervention includes health literacy materials, health screening (obesity, anaemia, depression, diabetes, hypertension), feedback and referral, behaviour change support through individual sessions (in-person or telephonic), free HIV and pregnancy testing, a free CV printing service, and multi-micronutrient supplementation. Intervention materials were developed for four life-course stages: preconception (Bukhali), pregnancy (Bukhali Baby), infancy (birth–2 years), and early childhood (Bukhali Mntwana, 2–5 years). Content experts and young women test groups were involved. A 'traffic light' system (green/orange/red) guides risk screening and referral. The control arm receives standard of care 'plus' delivered telephonically, with 6-monthly in-person visits, free HIV and pregnancy testing, and CV printing, covering non-health topics. The process evaluation follows UK MRC guidance, focusing on context, implementation, and mechanisms of impact using mixed methods. Fidelity monitoring uses the NIH Behaviour Change Consortium Treatment Fidelity Framework, though context-specific challenges were identified for intervention receipt and enactment in this LMIC setting.
**Key Results:** Several adaptations were made: preconception sessions shifted from community-based peer groups to individual sessions (largely telephonic) after the pilot phase; in-person home visits were not feasible; Healthy Conversation Skills required contextual adaptations including simplifying goal-setting, adapting for multilingual settings and low literacy, and incorporating a trauma-informed perspective. In mid-2021, a registered dietician joined to provide additional nutritional support for overweight/obese participants. In 2021, based on intervention team experience and process evaluation findings, specific attention was directed towards trauma. Discussion sessions with the intervention team revealed that trauma was understood in experiential and emotional terms, mental health stigma was acknowledged and linked to culture, trauma was often normalised and minimised, and participants felt pressure to be 'strong'. Health Helpers reported feeling sad, 'bad', and helpless when unable to change participants' traumatic circumstances, and noted that participants sometimes triggered their own trauma. Health Helpers were seen as a trusted source of support, often being the first to hear of traumatic events. Local referral systems were perceived as not working, and phone consultations were not trusted. An onsite counsellor was recommended, with the condition that the counsellor be able to relate to participants and be trustworthy. In March 2022, steps were taken to provide additional mental health support for Health Helpers incorporating trauma-informed care principles, and onsite counselling provision is being investigated.
**Clinical Implications:** The Bukhali trial demonstrates that in LMIC settings, mental health and trauma may be more salient priorities than physical health for young women, and that CHW-led interventions must address these broader support needs. The pragmatic trial design and ongoing process evaluation provide critical insights for future implementation within overburdened public health systems. The findings underscore that contextual realities may only emerge as a trial progresses, impacting recruitment, retention, and behaviour change. The application of implementation science frameworks like CFIR can facilitate learning across similar LMIC settings. The ongoing documentation and publication of the intervention development and process evaluation provide short-term learning benefits, and the focus on efficacy (phase 2) provides a foundation for future effectiveness (phase 3) evaluation within the public health system. Limitations include the inability to fully capture intervention complexity and the difficulty of continually comparing outcome and implementation evaluations due to the long-term nature of the trial and blinding.