After the fallout of the lipid hypothesis and creation of the Food Pyramid and Standard American Diet (SAD), anti-fat sentiment picked up speed and began spreading virally throughout the medical field. Dietary fat was reluctantly, but quickly, gaining a reputation as being responsible for nearly every major malady, from clogged arteries to acne.1 Articles like this sprang up regularly, implicating the thick, gooey substance as the grease that collected within our arteries, incarcerating blood and other substances in its presence, leading to arterial congestion and eventual complete obstruction.2 The detrimental health effects of an action or substance are rarely viewed in isolation, and it was only a matter of time before the link between dietary fat and breast cancer took the spotlight.
The evidence was accumulating and fat had no alibi. The majority of studies were epidemiologic, or in other words, they took a bird’s-eye view of a group of subjects, asked them what they ate at random time periods and then performed a series of complicated statistical calculations to analyze whether a relationship extended between fat and breast cancer. These studies are expensive and complicated – participants were often asked about their dietary habits days or even years after the fact. Dozens of other issues abound with this technique, as discussed before, and we are often left with confusion over conclusions.
I won’t belabor the point, as I dedicated several chapters of Misguided Medicine to the topic, but eventually fat, cholesterol, and particularly, saturated fat were chastised and we were strongly advised to avoid all fat or replace foods heavy in saturated fats with vegetable oils. Traditional diets were abandoned, and small Italian grandmothers like the one that helped raise me, threw out their cooking lard and began steaming their vegetables. The taste of steamed Brussels sprouts was so repulsive, that they eventually abandoned many vegetables, replacing them with “heart healthy” grains. It is not surprising that years later, studies have surfaced providing us with the exact opposite recommendations,3 and based on the long history of fat as a common source of nourishment within our diet,4,5 it is also not surprising to find that we may have had the cross-hairs on the wrong “villain.”
Dietary Fat and Breast Cancer – A Brief History
Before we get into the exact mechanism of how fat and breast cancer may be connected, let’s take a stroll through the background of the “link” between fat and breast cancer. The human link of fat and breast cancer was described in one of the largest population studies by Armstrong and Doll – the same renowned Doll that first described the link between tobacco and lung cancer. In 1975, they analyzed many worldwide environmental and dietary factors and the links between different types of cancer.6 The data associated different foods with cancer, like fish consumption in Asia with stomach cancer, and coffee consumption with kidney cancer. Among these myriad associations was the connection between fat consumption and breast cancer.
The take-home message was that different behaviors appeared to be associated with an increased or decreased risk of cancer. The emphasis on “associated” by the authors was an important point to note, and they cautioned against making any conclusions from their data. As they described in the article, the findings may “reflect some other variable correlated with economic development” and “the quality of the cancer incidence data is significantly affected by economic factors, particularly as controlling for any of the food consumption variables can reduce the correlation.” For example, coffee is associated with kidney cancer in their data but coffee does not necessarily cause kidney cancer; people at higher risk of kidney cancer may simply drink more coffee and people who drink more coffee are usually from more developed societies that engage in higher amounts of unhealthy behaviors. Armstrong and Doll also cautioned that the association between breast cancer and obesity makes further calculations difficult. However, the association stood, dietary fat may be a cause of breast cancer, and further studies were warranted. However, the caution from Armstrong and Doll of the danger of controlling for food consumption may be the lasting lesson of their important work.
Decades later, one of the earliest large population-based studies assessing the link between fat and breast cancer was published in 1987 by Dr. Walter Willett. Willett is the Fredrick John Stare Professor of Epidemiology and Nutrition and chair of the department of nutrition at Harvard School of Public Health. He also holds the title of professor of medicine at Harvard Medical School. In other words, he was and is a very important player in shaping the dietary landscape within the U.S.
Willett and his team from Harvard attempted to expand upon Doll’s findings by assessing the diets of nearly 90,000 nurses aged 34-59 with no history of cancer. They used a dietary questionnaire to estimate the amount of total fat, saturated fat, linoleic acid (polyunsaturated fat), and cholesterol. Within this population, they also studied a sub-sample of 173 women who consumed either the most fat (44% of their diet) or the least fat (32% of their diet). Women were followed for four years, at which point 601 were diagnosed with breast cancer.
Willett and his coauthors, after adjusting for calories, then compared the highest fat eaters with the lowest fat eaters to assess any relationships between fat and breast cancer. They found none. Or as they described:
“These data are based on a limited period of follow-up and do not exclude a possible influence of fat intake before adulthood or at levels lower than 30 percent of calories. They suggest, however, that a moderate reduction in fat intake by adult women is unlikely to result in a substantial reduction in incidence of breast cancer.”
The description leaves the reader to believe that their analysis was less of a question of whether dietary fat causes or protects us from breast cancer, but rather how much fat causes breast cancer. The bias – which we assume was largely influenced from the initial work of Armstrong and Doll – is clearly detectable within their description and merely foreshadows the condemnation of fat during this time period. Their study was negative, yet their conclusion was that perhaps the women who ate the least amount of fat did not go low enough. In other words, a moderate reduction of fat is unlikely to reduce breast cancer risk, but what about a large reduction? Also, perhaps they simply did not follow the group of women long enough to gain significant insight into the interaction between fat and breast cancer. Regardless of the undertones of bias, these were all reasonable questions that required further examination. An additional theory, perhaps unthinkable at the time, was that we were focusing our interrogation on the wrong suspect.
Along these lines, a dramatically different picture is painted when we read beyond their conclusions and question the negative results. The risks of breast cancer based on each type of fat within the diet were as follows:
- Total Fat – 18% reduced risk of breast cancer (confidence limits 0.64 and 1.05)
- Saturated Fat – 16% reduced risk of breast cancer (confidence limits 0.66 and 1.08)
- Linoleic Acid – 12% reduced risk of breast cancer (confidence limits 0.69 and 1.12)
- Cholesterol – 9% reduced risk of breast cancer (confidence limits 0.70 and 1.18)
The confidence limits listed on the right indicated whether the relationship is statistically significant or not. If it crosses 1 – which it did in all cohorts assessed – it is no longer significant. As we go from total fat to saturated fat, then linoleic acid, and finally cholesterol, the confidence interval further crosses 1 and becomes less significant. But the issue here is obvious: the more fat these women consumed, the lower risk they had of being diagnosed with breast cancer. Yes, it was not significant (though rather close for total and saturated fat), but the conclusion could have easily been:
“These data are based on a limited period of follow-up and do not exclude a possible influence of fat intake to decrease the risk of breast cancer at or above levels of 44%. They suggest, however, that a moderate increase in fat intake by adult women is unlikely to result in a substantial reduction in incidence of breast cancer.”
Willett and his group did not find the diamond in the rough that they were searching for, and based on Doll, Armstrong, and others’ findings, such bias is understandable as they were, in theory, trying to disprove that dietary fat caused breast cancer. However, they were not alone in their initial findings. A smaller, lesser-known study published two years before found no association between dietary fat and breast cancer.7 While Willett and his group’s work may have hinted at the exact opposite of what they were searching for, their work stood as a launching to further the science on the connection between diet and cancer. Others followed their lead. Less than a decade later, six additional similar studies emerged assessing dietary fat and breast cancer. Willett’s group combined all six to perform a massive analysis including over 337,000 women and 4980 cases of breast cancer. The results were similar and no association was found between dietary fat intake and the risk of breast cancer.8
While Willett and his team’s initial biases could be forgiven, as they were based on both animal and epidemiologic data implicating fat as a cause of breast cancer, the bias began to get the better of them. He attempted to answer those questions listed in the conclusion of the original study by following the group of nurses longer, now for a total of eight years. As suggested in their initial publication, perhaps we simply needed longer follow up or more extreme lowering of fat. Willett got both in his update, which he published in 1992.14 This time his findings were no different, further cementing a lack of a significant interaction of fat on cancer incidence. He assessed fiber consumption in this group of women, which also revealed no protective effect. Another negative study in the books, yet, much like the first study, peculiar conclusions were provided:
“These data provide evidence against both an adverse influence of fat intake and a protective effect of fiber consumption by middle-aged women on breast cancer incidence over 8 years. Nevertheless, the positive association between intake of animal fat and risk of colon cancer observed in many studies provides ample reason to limit this source of energy.”
In other words, though we have now provided two negative studies linking fat and breast cancer, other data links fat and colon cancer, so we should limit its consumption. Many voiced concerns with this conclusion, as the undertones of bias were becoming more apparent. The fact that this comment, (1) addressed colon cancer and was completely tangential to the study at hand, (2) made it through the peer review process as a primary conclusion of the study, and (3) was published by the popular Journal of the American Medical Association (JAMA), merely illustrates just how the far the anti-fat sentiment had spread throughout medicine and society.
Seven years later, Willett and his group published a final analysis of this group of women, now at 14 years of follow up.15 They had finally accumulated data from women eating less than 20% dietary fat (we should remember that this was after decades of aggressive low-fat dietary advice). The data still showed no benefit in reducing breast cancer risk with substantial fat avoidance. They found similar results as the first study, and higher fat consumption again trended to a lower risk of breast cancer.
However, this time around, they assessed fat versus carbohydrate consumption and found that increasing fat consumption and reducing carbohydrate consumption by 5% lowered the risk of breast cancer by 4%. This time around, these findings were also statistically significant. Their data not only showed a significant reduction of breast cancer with the increased consumption of fat, but also by replacing carbohydrates – the poster child of dietary recommendations over the past 30 years – with this fat. They also found that there was a statistically significant increase in breast cancer with dietary fat reduction during their calculation of trends within their data set. In keeping with bias of previous studies, their conclusions stated:
“We found no evidence that lower intake of total fat or specific major types of fat was associated with a decreased risk of breast cancer.”
Since this publication, multiple studies have shown fat to increase, decrease, and remain unrelated to breast cancer. This is unsurprising, as a compilation of nearly all food studies shows that all foods can cause or prevent cancer.16
Despite the fact that study after study revealed either no effect or a beneficial effect of fat consumption and breast cancer risk, anti-fat recommendations permeated throughout the cancer world and continue to plague it, consuming countless research dollars, misdirecting appropriate advice, and leaving patients as the victims. Doll, who was knighted in 1971 and died in 2005 at age 92, may have left us with some important advice at the end of his decades’ old analysis of environmental influences on cancer incidence. While it may have been modesty over his findings, it was an important lesson nonetheless:
“Given the many weaknesses of this method in terms of the quality of the data, allowances for latent periods and the uncertainty as to whether the most relevant environmental variables have even been included in the correlation matrices, it is clear that these and other correlations should be taken only as suggestions for further research and not as evidence of causation or as bases for preventive action.”
It is clear these correlations should be taken only as suggestions for further research. After multiple negative attempts to link dietary fat and breast cancer, studies remain negative. To Willett’s credit, he has been vocal about the lack of data showing that dietary fat may cause breast cancer. He has even commented that a low-fat, high-carbohydrate diet may increase the risk of breast cancer for a portion of women.17 At the 2012 National Cancer Institute Annual Advances in Cancer Prevention Lecture, Willett went as far as stating:
“There was never any strong evidence for this idea, but it was repeated so often that it became dogma in the 1980s and 1990s”
It certainly became dogma, but what happens when these suggestions are consistently proven wrong, but still become the basis of our dietary recommendations?
Dietary Fat and Breast Cancer – References:
- Davidovici BB. The role of diet in acne: facts and controversies. Clin Dermatol. 2010;28(1):12-16. doi:10.1016/j.clindermatol.2009.03.010.
- Hu FB, Stampfer MJ, Manson JE, et al. Dietary Fat Intake and the Risk of Coronary Heart Disease in Women. N Engl J Med. 1997;337(21):1491-1499. doi:10.1056/NEJM199711203372102.
- Chowdhury R, Warnakula S, Kunutsor S, et al. Association of Dietary, Circulating, and Supplement Fatty Acids With Coronary Risk. Ann Intern Med. 2014;160(6):398-406. doi:10.7326/M13-1788.
- Larsen CS. Animal Source Foods and Human Health during Evolution. J Nutr. 2003;133(11):3893S–3897S. http://jn.nutrition.org/content/133/11/3893S.abstract.
- Fine EJ, Champ CE, Feinman RD, Márquez S, Klement RJ. An Evolutionary and Mechanistic Perspective on Dietary Carbohydrate Restriction in Cancer Prevention. J Evol Heal. 2016;1(1). doi:10.15310/2334-3591.1036.
- Armstrong B, Doll R. Environmental factors and cancer incidence and mortality in different countries, with special reference to dietary practices. Int J Cancer. 1975;15(4):617-631. doi:10.1002/ijc.2910150411.
- Graham S, Marshall J, Mettlin C, et al. Diet in the Epidemiology of breast Cancer. Am J Epidemiol. 1982;116(1):68-75. doi:10.1093/oxfordjournals.aje.a113403.
- Hunter DJ, Spiegelman D, Adami H-O, et al. Cohort Studies of Fat Intake and the Risk of Breast Cancer — A Pooled Analysis. N Engl J Med. 1996;334(6):356-361. doi:doi:10.1056/NEJM199602083340603.
- Schaefer EJ, Augustin JL, Schaefer MM, et al. Lack of efficacy of a food-frequency questionnaire in assessing dietary macronutrient intakes in subjects consuming diets of known composition. Am J Clin Nutr. 2000;71(3):746-751. http://www.ncbi.nlm.nih.gov/pubmed/10702168.
- Pan A, Sun Q, Bernstein AM, et al. Red meat consumption and mortality: results from 2 prospective cohort studies. Arch Intern Med. 2012;172(7):555-563. doi:10.1001/archinternmed.2011.2287.
- Kipnis V, Midthune D, Freedman L, et al. Bias in dietary-report instruments and its implications for nutritional epidemiology. Public Health Nutr. 2002;5(Supplement 6a):915-923. doi:doi:10.1079/PHN2002383.
- Kipnis V, Subar AF, Midthune D, et al. Structure of Dietary Measurement Error: Results of the OPEN Biomarker Study. Am J Epidemiol. 2003;158(1):14-21. doi:10.1093/aje/kwg091.
- Kristal AR, Peters U, Potter JD. Is It Time to Abandon the Food Frequency Questionnaire? Cancer Epidemiol Biomarkers Prev. 2005;14(12):2826-2828. doi:10.1158/1055-9965.epi-12-ed1.
- Willett WC, Hunter DJ, Stampfer MJ, et al. Dietary Fat and Fiber in Relation to Risk of Breast Cancer. JAMA. 1992;268(15):2037. doi:10.1001/jama.1992.03490150089030.
- Holmes MD, Hunter DJ, Colditz GA, et al. Association of Dietary Intake of Fat and Fatty Acids With Risk of Breast Cancer. JAMA. 1999;281(10):914. doi:10.1001/jama.281.10.914.
- Schoenfeld JD, Ioannidis JPA. Is everything we eat associated with cancer? A systematic cookbook review. Am J Clin Nutr. 2013;97(1):127-134. doi:10.3945/ajcn.112.047142.
- Willett WC. Diet and breast cancer. J Intern Med. 2001;249(5):395-411. doi:10.1046/j.1365-2796.2001.00822.x.
© 2017 CDR Health and Nutrition, LLC. All Rights Reserved.