Background Globally, excess body weight causes 4 million deaths annually. The number of people affected by humanitarian crises stands at a record high level with 1 in 95 people in the world being forcibly displaced. These epidemics overlap.

Addressing obesity is a post-acute phase activity in the management of non-communicable diseases in humanitarian settings. Information is needed to inform guidelines, timing and targets of interventions. The objective of this review was to explore the prevalence of overweight and obesity in populations directly affected by humanitarian crises; the cascade of care in these populations and patient knowledge and attitudes to overweight and obesity.

Methods Literature searches were carried out in five databases. Grey literature was identified by searching online repositories of non-governmental organisations. The population of interest was non-pregnant, civilian adults (aged 18 years or older) who had experience of humanitarian crises (armed conflict, complex emergencies and natural disasters). All study types published from January 1^st, 2011, onwards were eligible. Screening and data extraction were carried out in duplicate. Risk of bias (ROB) was assessed by two authors across internal, external and overall domains. A narrative synthesis approach was taken with predefined subgroups (age, gender, displacement, exposure and geography). Greater interpretative weight was given to reports with lower internal ROB.

Results 20,376 non-duplicate search results were identified and screened. Forty-three reports were excluded at full text review due to inaccessible data reporting. Fifty-six reports from forty-five studies were included. Included studies were heterogenous in terms of population and exposures. Global distribution was skewed with few studies from Africa, South America or South Asia. Prevalence estimates varied widely across the studies and by subgroups. Estimates of overweight and obesity combined ranged from 6.4% to 82.8%. Increasing adiposity was seen over time, in older adults and in women. The effect of displacement was difficult to tease out. Only six studies were at low risk of bias overall. Body mass index was the predominant measure used. Outcomes other than prevalence were infrequently considered. No qualitative studies fitting our criteria were identified.

Discussion Overweight and obesity varied in crisis affected populations but were rarely absent. Improved reporting of existing data could circumvent the need for more primary research. Worsening obesity may be prevented by acting earlier in long-term crises and targeting higher risk groups. The use of waist circumference would provide useful additional information for risk factor management. Qualitative studies are needed to inform service development.