“We learn little from victory, much from defeat.”
~ Ancient Japanese Proverb
The relationship between food and health is complex, to say the least. Over the decades during which our current narrative and recommendations have evolved, there have been many connections and assumptions based on the prevailing logic and understanding of the moment. This is a proper course in the development of hypotheses and the construction of theory.
But during that time, there have also been many outliers and contrary findings, holes in the fabric that construct our momentary reality. This, too, is a proper course in the development of hypotheses and construction of theory. As Karl Popper observed, “In so far as a scientific statement speaks about reality, it must be falsifiable; and in so far as it is not falsifiable, it does not speak about reality.”
This week, I share two studies that directly address this aspect of falsifiability. Both, quite simply, are negative findings with respect to the Mediterranean diet (one study utilizes the MIND diet, which is a close cousin of the Mediterranean diet).
 The Mind Diet is defined as “a hybrid of the Mediterranean diet and the DASH (Dietary Approaches to Stop Hypertension) diet, with modifications to include foods that have been putatively associated with a decreased risk of dementia.”
The MIND Study:
- A two-site, randomized, controlled study involving 604 subjects (301 using the MIND diet and 303 in the control diet group) over a three-year period.
- The adults were older, overweight, and had a documented “suboptimal” diet prior to study enrollment.
- The MIND diet and the control diet employed mild caloric restriction
- The primary endpoint was the change from baseline in a global cognition score and four cognitive domain scores, all of which were derived from a 12-test battery.
- The MIND diet is a paragon of current healthy eating recommendations; it focuses on whole grains and green leafy vegetables, olive oil as the primary fat, and it emphasizes nuts, fish, and poultry as protein sources.
- The trial was based on epidemiological data, as highlighted on the Harvard website, where “Researchers found a 53% lower rate of Alzheimer’s disease for those with the highest MIND [diet] scores. Even those participants who had moderate MIND scores showed a 35% lower rate compared with those with the lowest MIND scores. The results didn’t change even after adjusting for factors associated with dementia.”
- When the MIND diet hypothesis was actually rigorously tested, the researchers found that “Among cognitively unimpaired participants with a family history of dementia, changes in cognition and brain MRI outcomes from baseline to year 3 did not differ significantly between those who followed the MIND diet and those who followed the control diet with mild caloric restriction.”
- The MIND diet with mild caloric restriction is no better in preventing age-related cognitive impairment than a control diet with mild caloric restriction (it does not refute that the MIND diet may positively impact age-related cognitive impairment; an important distinction).
- The epidemiologic evidence must be appreciated for what it is: hypothesis-generating data, not a conclusion of fact.
- As these findings reinforce, correlation does not equate to causation.
The DIANA-5 Study:
- The study enrolled 2,132 women surgically treated for stage 1-3 invasive breast cancer within the previous five years from enrollment 2008 to 2012.
- There were 1,542 women randomized (590 who were enrolled were excluded), with 769 in the intervention group and 773 women in the control group.
- The study was a multi-institutional (11 Italian sites), randomized, controlled dietary intervention trial.
- The control group received general guideline-derived recommendations for cancer prevention.
- The intervention group received the same as well as support “through kitchen courses, monthly community meals, and physical activity recommendations.”
- The trial was based on previous data demonstrating decreased bioavailable sex hormones, insulin-like growth factor-1 (IGF-1), and body weight utilizing a Mediterranean diet with increased phytoestrogen intake (macro-Mediterranean diet) in postmenopausal women and women previously diagnosed with breast cancer (DIANA-1 & 2 Trials).
- The current AICR/WCRF Decalogue suggests that breast cancer survivors should achieve a dietary pattern high in vegetables, fruits, whole grains, and legumes, low in saturated fat and alcoholic beverages, and should increase the consumption of soy products.
- The primary outcome was breast cancer recurrence.
- The intervention group significantly improved their consumption of recommended foods (P<0.001) and decreased their consumption of discouraged foods (P<0.001).
- The intervention group significantly improved their Dietary Index score compared to the control group, with an average difference of 3.1.
- Breast cancer recurrence occurred in 95 women in the intervention group (12.4%) and 98 women (12.7%) in the control group.
- The researchers concluded that the specific dietary intervention used “did not elicit any difference between the IG [intervention group] and the CG [control group]… either because changing the diet and improving physical activity recommendations has no effect or because the difference achieved by the intervention is too small to elicit any difference in disease-free survival.”
- The macro Mediterranean diet is no better in preventing breast cancer recurrence in women with stage 1-3 invasive breast cancer than general guideline-derived recommendations for cancer prevention.
- Different dietary interventions may have yielded different results.
- Even “logical” extrapolation of previous trial data needs rigorous hypothesis testing.
The data is the data. The real challenge is how we receive these and other “negative” trial results and what we do at that moment. Do we move behind the curtain, entranced by the light of discovery that beckons us on? Or do we simply choose to ignore whatever is behind the curtain? Or perhaps worst of all, do we fall victim to our own biases of the way we want things to be? Do we put on our white hat and spin our own tall tales, reworking and reimagining findings that are confounding and contradictory to the currently accepted or preferred storyline?
White hat bias is “a type of bias that occurs when researchers, scientists, or authors present information in a way that portrays a particular intervention, practice, or technology as more beneficial, effective, or positive than the available evidence supports. This bias often arises unintentionally due to the researchers’ positive intentions, such as a desire to promote public health or advocate for a particular cause.”
There are several defining characteristics of white hat bias, including:
- Overemphasis on Positive Outcomes:
Researchers may overemphasize the positive aspects of an intervention while downplaying or neglecting potential drawbacks or limitations.
- Selective Reporting:
There might be a tendency to selectively report data or focus on specific outcomes that support the favorable view of the intervention, ignoring less favorable or negative results.
- Downplaying Risks or Uncertainties:
White hat bias may involve downplaying the risks, uncertainties, or potential harms associated with a particular intervention, making it seem more straightforward or beneficial than it is.
- Inadequate Disclosure:
Researchers may not adequately disclose conflicts of interest or affiliations that could influence their perspective, leading to a potential bias in the presentation of information.
- Advocacy Language:
The use of advocacy language, enthusiastic language, or strong endorsements in research articles or reports can contribute to white hat bias by creating a positive narrative.
It is essential for researchers to be aware of the potential for bias and to strive for objectivity in presenting their information. However, it is equally incumbent for those reading and reporting on the results to acknowledge limitations and present a balanced view of the results to ensure the credibility of scientific communication. It is crucial to ensure policymakers and the public receive accurate and unbiased information for informed decision-making.
Applying the philosophy of Geoffrey Rose, an eminent British epidemiologist, that if a medical hypothesis has a chance of being true and thus saving lives, it should be treated as if it is, can lead to undesirable outcomes. Such assumptions, particularly when they manifest for a period of time, can solidify into dogma.
In our current age of soundbites and social media hyperbole, it takes little for people to invest themselves in the validity of such statements, and that includes researchers and other decision-makers. Once entrenched, it becomes increasingly difficult to dislodge what becomes “common wisdom.”
When the new physics of quantum mechanics appeared over a century ago with its perplexing and, at times, contradictory findings, there was significant initial resistance to its validity. So much so that it led the German physicist Max Planck to observe that “A new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents die, and a new generation grows up that is familiar with it.”
Or, as we conveniently paraphrase, science advances one funeral at a time.
As we continue to unravel the complex connections between food and health, let us be willing to embrace the positive and the negative and enjoy the journey wherever it may lead.
Barnes, Lisa L.; Dhana, Klodian; Liu, Xiaoran; Carey, Vincent J.; Ventrelle, Jennifer; Johnson, Kathleen; Hollings, Chiquia S.; Bishop, Louise; Laranjo, Nancy; Stubbs, Benjamin J.; Reilly, Xavier; Agarwal, Puja. (2023) Trial of the MIND Diet for Prevention of Cognitive Decline in Older Persons. New England Journal of Medicine; 389:602-611. https://DOI:10.1056/NEJMoa2302368
Franco Berrino, Anna Villarini, Giuliana Gargano, Vittorio Krogh, Sara Grioni, Manuela Bellegotti, Elisabetta Venturelli, Milena Raimondi, Adele Traina, Maurizio Zarcone, Rosalba Amodio, Maria Piera Mano, Harriet Johansson, Salvatore Panico, Maria Santucci de Magistris, Maggiorino Barbero, Cristina Gavazza, Angelica Mercandino, Elena Consolaro, Rocco Galasso, Luciana Del Riccio, Maria Chiara Bassi, Milena Simeoni, Paolo Premoli, Patrizia Pasanisi, Bernardo Bonanni, Eleonora Bruno. (2023) The Effect of Diet on Breast Cancer Recurrence: The DIANA-5 Randomized Trial. Clin Cancer Res; https://doi.org/10.1158/1078-0432.CCR-23-1615