Processed meat

Caution

3 studies · 1 recommendation

Last updated: February 25, 2026

Processed meat – Breast Cancer
Caution3 studies

Processed meat consumption linked to modestly increased breast cancer risk

Three studies encompassing over 1.47 million participants consistently associate processed meat intake with elevated breast cancer risk. A dose-response meta-analysis of 46 prospective studies (1.15 million participants) found a 7% higher risk for highest versus lowest intake (RR 1.07, 95% CI 1.01–1.14) and a 9% increase per 50 g/day serving (RR 1.09, 95% CI 1.02–1.17), with low heterogeneity and no publication bias. The EPIC cohort (319,826 women, 7,119 cases, median 8.8-year follow-up) reported a hazard ratio of 1.10 (95% CI 1.00–1.20) for the highest quintile of processed meat consumption. A smaller case-control study (n = 39) identified significant associations between processed meat products and breast cancer status (p < 0.05). The effect size is modest but consistent across study designs, supporting cautious limitation of processed meat intake.

Evidence

Authors: Ho, JCM, Huang, J, Li, X, Wu, J, Zeng, R, Zhang, J, Zheng, Y

Published: January 1, 2016

Meta-analysis of 46 prospective studies including over 1.15 million participants. For processed meat, the summary relative risk (RR) comparing highest versus lowest intake was 1.07 (95% CI 1.01–1.14, I² = 34.6%). Dose-response analysis showed RR of 1.09 per 50 g/day serving increase (95% CI 1.02–1.17, I² = 11.8%). Low heterogeneity across studies strengthens the finding. No publication bias was detected by Begg's test or Egger's test.

Authors: Fahmi, Irawati

Published: May 7, 2013

A case-control study at RSUD Dr. Moewardi compared fat consumption patterns between 13 breast cancer patients and 26 matched controls using Food Frequency Questionnaires and 24-hour dietary recall. Kolmogorov-Smirnov testing identified meat sausage and canned sardines among foods with statistically significant associations with breast cancer status (p < 0.05). Overall MUFA intake differed significantly between cases and controls (p = 0.017), as did PUFA intake (p = 0.024), though total fat intake difference did not reach significance (p = 0.103). The small sample size (n = 39) and accidental sampling method limit generalizability.

Authors: Androniki Naska, Anja Olsen, Anne Tjønneland, Annika Steffen, Antonia Trichopoulou, Armstrong, Balsari, Bingham, Boeing, Bohlscheid-Thomas, Carla H van Gils, Carlos Alberto Gonzalez Svatetz, Carmen Navarro, Cerhan, Cho, Claudia Agnoli, Cochran, Cross, Cui, Cummings, Dagrun Engeset, Dominique S Michaud, Duncan, Egeberg, Eiliv Lund, Elio Riboli, Elisabet Wirfält, Elizabeth Spencer, EPIC Group of Spain, Eva Ardanaz, Ferrari, Franceschi, Franco Berrino, Françoise Clavel-Chapelon, Freudenheim, Friedenreich, Fränzel JB van Duijnhoven, Fung, Fung, Fung, Giovanna Masala, Gonzalez, Goodman, Graham Byrnes, Grambsch, Gray, Guri Skeie, Göran Hallmans, H Bas Bueno-de-Mesquita, Haftenberger, Heiner Boeing, Hermann, Hirohata, Hirose, Hjartaker, Holmes, Isabelle Romieu, Iscovich, Jakob Linseisen, Jonas Manjer, Kaaks, Kaaks, Kabat, Kay-Tee Khaw, Key, Keys, Kim Overvad, Lauber, Laudina Rodriguez, Lichtenstein, Linos, Linos, Maria-José Sánchez, Marianne Uhre Jakobsen, Marie-Christine Boutron-Ruault, Missmer, Mokbel, Morales Suarez-Varela, Ocke, Paolo Vineis, Per Lenner, Petra HM Peeters, Pietro Ferrari, Pilar Amiano, Pisani, Prieto-Ramos, Riboli, Riboli, Rohrmann, Ronco, Rosario Tumino, Sabina Rinaldi, Sabina Sieri, Sabine Rohrmann, Salvatore Panico, Sara Grioni, Shannon, Sheila Bingham, Shin, Sieri, Slimani, Slimani, Tajima, Taylor, Teresa Norat, Thompson, Timothy J Key, Touillaud, Tretli, Valeria Pala, van der Hel, Vassiliki Benetou, Vittorio Krogh, Willett, Wiseman, Women’s Health Initiative, World Cancer Research Fund/American Institute for Cancer Research, Zhang, Zheng

Published: January 1, 2009

In the EPIC cohort of 319,826 women followed for a median of 8.8 years, 7,119 breast cancer cases were identified. Women in the highest quintile of processed meat consumption had a hazard ratio of 1.10 (95% CI: 1.00–1.20) compared to the lowest quintile, with a borderline trend (P for trend = 0.07). The association was observed in multivariate Cox proportional hazard models adjusting for relevant confounders.