Connecting the dots between diet, adiposity, and cancer risk

From the J. Lampe Group and P. Lampe Lab, Division of Public Health Sciences

An estimated 40% of all cancer cases in the United States could be prevented through modifiable lifestyle-based risk factors. Excess body weight and poor diet, characterized by low fruit, vegetable, and dietary fiber intake, are included in this list of lifestyle factors. Given this, sorting out how diet and excess adiposity contribute to cancer risk and whether interactions between these factors exist could have a substantial public health impact.

The glycemic response to foods, i.e. post-consumption change in blood glucose level, can be described by the glycemic index and glycemic load (GL). Diets that are low in fruits, vegetables, and whole grains generally have a high GL while those that are rich in these foods have a low GL. The consumption of diets that elicit a high glycemic response have been linked to risk for certain types of cancer. New advances in understanding this diet-cancer connection were recently published in the journal Cancer Prevention Research by researchers in the Division of Public Health Sciences. The work was led by postdoctoral fellow Dr. Carly Garrison and investigated the impact of a low GL diet rich in whole grains (WG) compared to a high GL diet rich in refined grains (RG) on cancer-related pathways and whether the impact varied by level of body fat mass.

The authors analyzed plasma samples previously collected from a randomized crossover diet study that included eighty healthy adults. “A unique aspect of this work is the use of a controlled feeding study to evaluate effects of two dietary patterns on cancer-related pathways,” said Dr. Johanna Lampe, senior author of the study and Fred Hutch member. All subjects consumed both diets for 28 days (with a wash-out phase between the diets) and all foods were provided to the subjects during the two diet periods. In addition to the rigorous study design, innovative methodology previously developed at Fred Hutch allowed the authors to readily gain insight into changes in dozens of relevant biological pathways. Dr. Garrison explained, “This study was possible due to the use of the proteomics arrays developed in the P. Lampe lab. The arrays provide the ability to study thousands of proteins with a small amount of biological sample in a cost-effective manner.”    

Graphical representation of differential protein expression in plasma samples in response to WG (whole grain) and RG (refined grain) diets.
Differential protein expression in plasma samples in response to WG (whole grain) and RG (refined grain) diets. There were forty-eight proteins that were differentially expressed in common between the low and high fat mass (FM) groups. Proteins with higher expression in the WG diet are shown in red, those with higher expression in the RG diet are shown in blue. Image from Dr. Carly Garrison

First, differences in plasma protein expression by diet (WG vs RG) were assessed. One hundred and seventy-two proteins, of 2,072 included on the array, were found to be differentially expressed in response to diet when results from all eighty subjects were included. When participants were then categorized based on high or low fat mass (FM), a greater number of differentially expressed proteins were identified. There were 266 proteins in the low FM group and 221 in the high FM group. When the lists of proteins were compared between these two groups, 48 proteins were found to be in common (see Figure). Half of these proteins changed in the same direction independent of FM level while the other half changed in the opposite diet between the FM groups. For example, eleven proteins were higher after the WG diet in the low FM group but higher after the RG diet in the high FM group (see Figure).

The authors then conducted pathway analyses to gain broader insight into the dietary responses. With all subjects together in the initial analysis, 69 pathways were identified as overrepresented. Consistent with the individual protein analyses, many more pathways were identified when the subjects were stratified by FM, with 16 pathways in common. Post-translational protein modification pathways were overrepresented in both FM groups. However, within the high FM group, the DNA repair, cell cycle, and DNA replication primary pathways were overrepresented for proteins that differed in response to the WG and RG diets. These results are noteworthy, as alterations of these pathways are required for the development of cancer. Additionally, immune system and signaling by WNT pathways were overrepresented in the high FM group. Taken together, the results suggest that responses of proteins and cancer initiation related pathways to different dietary patterns are dependent on adiposity or adiposity-associated factors.

“Although we know that metabolic differences by percent body fat may affect risk of several cancers, the effects of diets in these different populations and how this affects recommendations for cancer prevention are less well understood. We see substantial differences in cancer-related pathway responses to diet between the overweight/obese and the normal weight participants,” said Dr. Lampe. While this work clearly demonstrates an interaction between adiposity and diet on cancer-related pathways, there is more work to do. “More research is needed to identify the biological effects of diets in populations differing in race, age, and weight so that we may better provide dietary recommendations for cancer prevention,” added Dr. Garrison.


This work was supported by the National Institutes of Health.

Fred Hutch/UW Cancer Consortium members Drs. Johanna Lampe, Paul Lampe, Daniel Raftery, Mario Kratz, Timothy Randolph, and Marian Neuhouser contributed to this research.

Garrison CB, Zhang Y, Navarro SL, Randolph TW, Hullar MAJ, Kratz M, Neuhouser ML, Raftery D, Lampe PD, Lampe JW. 2019. Proteomic analysis of plasma reveals fat mass influences cancer-related pathways in healthy humans fed controlled diets differing in glycemic load. Cancer Prevention Research. doi: 10.1158/1940-6207.CAPR-19-0175.