The research team, led by Guillermo Barahona-Fuentes and Gerson Ferrari of Universidad de Santiago de Chile alongside Leandro F.M. Rezende of Universidade Federal de São Paulo, set out to address a persistent blind spot in global cancer epidemiology. Worldwide, approximately 37.8 percent of new cancer cases are attributable to modifiable risk factors, but the vast majority of that evidence originates from high-income countries in North America and Europe. Latin America, which records roughly 1.4 million new cancer cases and 670,000 cancer deaths each year, has historically been under-represented in international syntheses. Global burden studies often rely on extrapolations that may not capture the region’s specific exposure patterns, its dual epidemiologic burden of infection-linked malignancies such as HPV-driven cervical cancer and Helicobacter pylori-associated gastric cancer, and lifestyle-driven tumors of the breast, colorectum, and lung.

The methodology behind the new analysis was deliberately conservative. The researchers screened 984 records, whittling them down through title and abstract review and full-text assessment to 47 studies, 46 of which used case-control designs. Recognizing that pooling estimates across wildly different exposure definitions, cancer sites, and contrast categories can produce misleading numbers, the team restricted confirmatory quantitative meta-analysis to just two site-specific combinations where independent studies shared a homogeneous exposure construct. Everything else was summarized through a structured direction-of-effect synthesis, known as SWiM, in which each estimate is classified as harmful, protective, or null, and the consistency of direction is tested with exact binomial probabilities. This decision reflects a growing methodological awareness in evidence synthesis: a single pooled odds ratio generated from non-comparable constructs can obscure more than it reveals.

The two confirmatory pools yielded the study’s most precise quantitative findings. A higher score on the dietary inflammatory index, a measure of how strongly a person’s diet promotes systemic inflammation, was associated with increased prostate cancer risk in a meta-analysis of two independent samples, one from Argentina and one from Mexico, with a random-effects odds ratio of 1.38 (95 percent confidence interval 1.11–1.73) and moderate heterogeneity of 36 percent. Obesity, defined as a body mass index of 30 or higher, was associated with a sharply elevated risk of endometrial cancer, with a random-effects odds ratio of 3.81 (95 percent confidence interval 1.08–13.45), though the authors stress that this estimate is fragile because it rests partly on a 66-participant pilot study from Mexico. Mechanistically, the obesity-endometrial link is biologically plausible: excess adiposity promotes chronic systemic inflammation, hyperinsulinemia and insulin resistance, hormonal dysregulation, and perturbations of the gut microbiome, all of which reshape the tumor microenvironment.

The direction-of-effect synthesis across behavioral domains revealed patterns that were consistent even where pooling was impossible. Unhealthy dietary exposures, ranging from Western dietary patterns and pro-inflammatory diets to high chili pepper and capsaicin consumption and diets dominated by animal-origin and ultra-processed foods, were associated with increased cancer risk in 15 of 16 estimates, with a median odds ratio of 1.60 and a binomial probability of 0.0005. Healthy dietary exposures, by contrast, were protective in every single contributing estimate: nine out of nine, spanning phytoestrogens, folate, vitamin A, cinnamic acid polyphenols, vitamin D, prudent dietary patterns, mate and tea intake, and vegetarian diets, with a median odds ratio of 0.41 and a binomial p value of 0.0039. The consistency of this protective signal across six different cancer sites is among the most striking results of the analysis.

Tobacco smoking told a similarly coherent story, but in the opposite direction. Six of seven estimates indicated increased cancer risk, with a median odds ratio of 4.00 and a binomial p value of 0.0312. The strongest associations appeared for lung cancer, where ever-smokers faced roughly a fourfold increase in risk compared with never-smokers in a Mexico City study, and for esophageal cancer in a multinational IARC study spanning Argentina, Brazil, Paraguay, and Uruguay, where smoking carried an odds ratio of 3.24. Bladder cancer risk in Argentina was more than four times higher among ever-smokers. The single null result concerned cervical cancer, which after adjustment for other factors showed no independent smoking effect. Notably, the researchers corrected earlier syntheses that had treated male and female strata of the same multicenter study as independent, a subtle but important methodological refinement.

Alcohol consumption was consistently associated with elevated risk, particularly for cancers of the upper aerodigestive tract and esophagus, consistent with the classification of ethanol and its metabolite acetaldehyde as Group 1 carcinogens. In the multinational IARC study, ever-drinkers faced an odds ratio of 4.62 for oral cavity and upper aerodigestive tract cancers. Gastric cancer risk was more than doubled among the heaviest wine consumers in a Mexican study. Associations for breast and colorectal cancer were weaker and non-significant, and the authors emphasize that with only five estimates in the alcohol domain, these findings are exploratory. Physical activity showed a suggestive protective direction in all three available estimates, with a median odds ratio of 0.61, but the evidence base remains too thin for robust conclusions, a gap the researchers flag as a priority for future cohort research.

Underlying the headline findings is a stark data-quality problem that the study confronts head-on. The geographic concentration is extreme: 24 of the 47 studies were conducted in Mexico and 10 in Brazil, with Argentina contributing four and Uruguay, Chile, and Peru one each. Risk-of-bias assessment using the QUIPS tool classified 40 percent of studies as low risk, 53 percent as moderate, and 6 percent as high. Exposure measurement was a frequent concern, with two-thirds of studies rated moderate risk, largely because dietary intake was typically assessed with food-frequency questionnaires without documented validation. The authors also corrected several published estimates during their audit: a previously reported extreme gastric cancer odds ratio of 17.11 for the highest chili pepper consumption category was replaced with the study’s standardized any-versus-none contrast of 5.49, and a total fat estimate for endometrial cancer was corrected from 4.00 to a non-significant 1.19 after verification against the original publication.

The regional context amplifies the public-health significance of these associations. In several Latin American countries, more than 15 percent of adults smoke, often alongside occupational exposures and nutritional deficiencies that can potentiate carcinogen activation. Obesity affects more than 25 percent of adults regionally, and in Mexico, Chile, and Brazil more than 60 percent of adults carry excess body weight, a figure that has climbed steadily since the 1990s in parallel with a dietary transition replacing fresh and minimally processed foods with ultra-processed products. More than 40 percent of Latin American adults fail to meet minimum physical activity recommendations, and global per-capita alcohol consumption is projected to rise from 6.5 to 7.6 liters of pure alcohol per adult by 2030. Each of these trends pushes in the direction of greater cancer incidence, and the region’s lower five-year survival rates relative to high-income countries mean that prevention carries disproportionate value.

The authors argue that their findings support scaling up proven regional interventions, including Chile’s front-of-package warning labels, Mexico’s taxation of sugar-sweetened beverages, and Brazil’s tobacco control programs. They also lay out a minimum measurement agenda for future Latin American studies: reporting alcohol in grams per day with standardized beverage conversions, capturing binge versus regular drinking patterns, using lifetime abstainers rather than former drinkers as the reference group to avoid sick-quitter bias, and adopting a harmonized core set of confounders including age, sex, smoking status, and socioeconomic position. Building national cohorts, consolidating interoperable cancer registries and survey linkages, and committing to pre-registration and shared analysis code are described as essential steps toward reproducible, policy-relevant evidence.

While the authors caution that the predominantly case-control design of the included studies limits causal inference, and that findings from Mexico and Brazil may not generalize across a region marked by deep socioeconomic and health-system heterogeneity, the direction of the evidence is clear and aligns with global data. Tobacco, poor diet, alcohol, obesity, and inactivity are each modifiable, and the new synthesis demonstrates that the well-established links between these exposures and cancer observed in wealthy nations hold, in sometimes distinct patterns, across Latin America as well. For a region where cancer increasingly competes with infectious disease for health-system resources, the study provides what its authors describe as decision-relevant evidence: a quantified, region-specific case that prevention, from tobacco regulation to healthy food environments, is among the most powerful cancer-fighting tools Latin America possesses.