BACKGROUND: The exact link among dietary behavior, body constitution, and risk of breast cancer is still ambiguous, potentially influenced by the fact that breast cancer is a multitude of diseases with different bases for transformation and consequently etiology. METHODS: In a cohort of 17,035 women enrolled in the Malmo Diet and Cancer population study, 346 emerging breast cancers were subcategorized by conventional pathology parameters such as tumor type, grade, and proliferation as well as by expression of key suppressor genes and oncogenes involved in the cell cycle control. Tumors were studied on tissue microarrays. Subcategories were then related to diet history information and objective body measurements determined several years before the cancer diagnosis. RESULTS: Lobular breast cancers were, in comparison to ductal cancers, linked to a higher alcohol intake. Smaller hip size and lower BMI were associated with low-grade tumors, whereas total energy, total fat, and energy-adjusted fat intake were inversely associated with tumor proliferation. Similar findings were observed for cyclin D1 overexpression, whereas cyclin E overexpression was associated with a higher energy-adjusted fat intake. Surprisingly, there was no association between estrogen receptors and diet or body measurements. CONCLUSIONS: Dietary behaviors and body constitution were clearly linked to the development of specific types of breast cancer defined by conventional pathology parameters or key cell cycle regulators. In general, high intakes of energy, fat, and polyunsaturated fatty acid but lower BMI were linked to the development of low-malignant breast cancer, and cyclin D1 and E seemed to be key targets in the process.