?Background
The consequences of armed conflict extend beyond direct casualties and can impact health systems and public health initiatives. Efforts to meet the United Nations’ Sustainable Development Goal (SDG) Target of a one-third reduction in premature mortality for non-communicable diseases (NCDs), including cancers, by 2030, may be hindered by armed conflict.

Objectives
This study aimed to assess trends in the burden of cancer in countries according to levels of conflict intensity, and to quantify any association between armed conflict and premature mortality due to cancer.

Methods
We analysed trends in the age-standardised rates (ASR) of premature mortality from cancer, across countries, stratified by conflict intensity. We utilised a longitudinal panel dataset that integrated country-level cause-of-death data extracted from the World Health Organization mortality database, armed conflict data from the Uppsala Conflict Data Program, and socio-economic covariates from relevant global data sources. Applying linear mixed-effects models (LMM), we calculated the adjusted mean ASRs, adjusted annual rates of change in ASR and temporal effects of conflict on premature cancer mortality, with countries stratified by conflict intensity.  To quantify the burden, an LMM-based regression counterfactual estimation method was employed to estimate the excess premature mortality of cancer due to armed conflict in that period.

Results
We analysed a total of 2395 country-year observations from 88 countries with high- or medium-quality mortality data, for which armed conflict status was available for 1989-2018. Although the ASR for premature mortality due to cancer declined in most countries in the past three decades, the reduction was slower in conflict-afflicted countries. The average annual rate of change in ASR for premature cancer mortality was -1.08% in non-conflict countries, compared with -0.96% in minor conflict-afflicted settings and -0.78% in war-affected countries. This general pattern was also observed for specific cancer types, particularly breast, cervical, colorectal, leukemia, bronchial and tracheal cancer. After adjustment for socioeconomic covariates, the adjusted mean ASR for premature cancer mortality was higher in conflict settings (139 per 100,000 population in war-affected countries and 138 in minor conflict-affected countries) than in non-conflict settings (130 per 100,000). In the LMM, the decline in adjusted ASR for premature cancer mortality was also significantly slower over time in countries exposed to armed conflict, with an approximately 0.7 lesser ASR decline per year in minor- and war-affected countries (p <0.001). The lagged effect of conflict on premature mortality from cancer persisted up to a decade after conflict. We estimate that 1.3 million excess premature deaths from cancer have occurred due to armed conflict over the 30-year period.

Conclusions
Armed conflict has likely had an effect on premature cancer mortality, delaying progress towards the SDG targets. Our study highlights the need for global action to address cancer disparities in conflict-afflicted settings, whilst acknowledging substantial challenges in those settings.

Adjusted Annual Rate of Change of Age-standardized rate of premature mortality due to cancer among different conflict intensity (1989-2018)