Association between Dietary Energy Density and Risk of Breast, Endometrial, Ovarian, and Colorectal Cancer among Canadian Women

Breast, endometrial, ovarian, and colorectal cancers are among the leading cancers affecting Canadian women (1). The mechanism that underpins the development of these cancers is not completely understood, but epidemiologic and experimental studies suggest that relatively common metabolic disorders such as obesity, diabetes, and the metabolic syndrome are potential risk factors for these cancers (2). These metabolic disorders result partly from poor-quality diets characterized by high intake of nutrient-deficient foods which are high in fats, refined sugars, and alcohol (3, 4). In this regard, dietary energy density (DED), a new indicator of diet quality, has been positively associated with risk of these cancer-associated metabolic disorders (3, 4). Therefore, it is possible that DED may influence carcinogenesis, but very few studies have assessed the association between DED and cancer risk. One recent study found that DED is positively associated with risk of breast cancer (5), but to our knowledge, no studies have been conducted to examine its association with endometrial, ovarian, and colorectal cancers. Nevertheless, given the association of the aforementioned metabolic disorders with risk of these cancers, it is of interest to also determine whether DED influences their risk.

Therefore, we investigated the association between DED and risk of breast, endometrial, ovarian, and colorectal cancers among women in the Canadian Study of Diet, Lifestyle, and Health (CSDLH).

The CSDLH, described in detail elsewhere (6), is a prospective cohort study that includes a total of 73,909 participants (34,291 males, 39,618 females) who were recruited between 1992 and 1998. At enrolment, each participant completed a self-administered quantitative food frequency questionnaire that collected information on usual dietary intake of 166 food items over the previous year, and also completed a self-administered lifestyle questionnaire that ascertained information on sociodemographic, anthropometric, and lifestyle factors. For our analyses, the study comprised an age-stratified subcohort of 3,120 women, and 922 and 269 incident invasive breast and colorectal cancer cases, respectively. For endometrial cancer, we excluded women with a history of hysterectomy prior to enrolment in the study, resulting in a subcohort of 3,092 women and 188 incident cases, while for ovarian cancer, we excluded women with a history of bilateral oophorectomy leaving a subcohort of 2,826 women and 104 incident ovarian cancer cases.

Total DED, the exposure of interest, was calculated by dividing the total energy content [kilocalories (kcal)] of reported food intake by the total food weight (gm). All beverages were excluded from the calculation, as the participants' water intake was unknown. The endpoints of interest were incident breast, endometrial, ovarian, and colorectal cancers. Incident cancer cases were ascertained via record linkage to the Canadian Cancer Registry and to the Ontario Cancer Registry.

Cox proportional hazards regression was used to estimate HRs and 95% confidence intervals for the association of DED with risk of reproductive cancers (breast, endometrial, and ovarian) and colorectal cancer, adjusting for education, smoking status, physical activity, alcohol intake, and BMI in all models. Models for reproductive cancers were additionally adjusted for age at menarche, parity, breastfeeding, menopausal status, HRT use, oral contraceptive use, and family history of breast cancer in a first-degree relative, while colorectal cancer was additionally adjusted for red/processed meat intake, dairy intake, folate, and folate. The statistical analyses were conducted using Stata 14.1 (StataCorp).

Median DED in the study population at baseline was similar for both subcohort members and cases (1.3 kcal/g; Table 1). After adjusting for potential confounders, we found no associations between DED and risk of breast, endometrial, ovarian, or colorectal cancer among women (Table 2).

Despite the relatively strong associations that have been observed between DED and cancer-associated metabolic disorders (3, 4), in this study, there was no evidence to support independent associations between DED and risk of breast, endometrial, ovarian, and colorectal cancers with or without adjustment for BMI. Contrary to our results, one large prospective study reported a positive association between DED and risk of breast cancer (n = 2,509 breast cancer cases; ref. 5). To our knowledge, the associations between DED and endometrial, ovarian, and colorectal cancers have not been previously studied. Our null results for endometrial and ovarian cancer may be due in part to the relatively small number of cases. Errors in the DED estimates resulting from recall and reporting biases may also have contributed to our findings. In addition, the food coding database may not have fully accounted for water loss during cooking, which may impact the energy density of a food (7). Nonetheless, an association between DED and risk of breast, endometrial, ovarian and colorectal cancers should not be ruled out, as DED is strongly correlated with metabolic disorders such as obesity and metabolic syndrome which may generate chronic inflammation, oxidative stress, and other processes that predispose to carcinogenesis (2). Therefore, further investigations in larger cohort studies are warranted, as the results of such research may have important implications for the primary prevention of these cancers.

No potential conflicts of interest were disclosed.

Conception and design: R. Arthur, T.E. Rohan

Development of methodology: R. Arthur, T.E. Rohan

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): V. Kirsh, T.E. Rohan

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): R. Arthur, T.E. Rohan

Writing, review, and/or revision of the manuscript: R. Arthur, V. Kirsh, T.E. Rohan

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): R. Arthur, V. Kirsh, T.E. Rohan

Study supervision: T.E. Rohan

Drs. Arthur and Rohan are supported by a grant from the Breast Cancer Research Foundation (BCRF-16-137).