Are the New England Journal of Medicine and Harvard T.C. Chan School of Public Health Dishonest or Just Incompetent?

You probably know where I come down on this. A new study spotlights this issue in spades.

Recall:

  1. The New England Journal of Medicine (NEJM) recently published a study from the Harvard T.C. Chan School of Public Health (HSPH) claiming to prove that air pollution kills.
  2. I asked to the NEJM to retract the study based on scientific misconduct by the study authors and others.
  3. Jim Enstrom asked the NEJM to allow a special article in response.

So the NEJM published today a new study from HSPS researchers claiming that quality of diet affects premature mortality. The media release is below.

Points to ponder:

  1. The new NEJM/HSPH study identifies quality of diet as a risk factor for premature death.
  2. The NEJM/HSPH study on PM2.5 did not adjust its (dubious) statistical associations for PM2.5 and mortality for any confounding risk factors.
  3. The new study’s reported association between diet and premature death is more than large enough to significantly affect the (dubious) results of the PM2.5 study.
  4. At the very least, the NEJM editors knew at the time of publication of the PM2.5 study that the new diet study was in their publication pipeline. Their publications came, after all, only two weeks apart.
  5. Although both HSPH studies have the same health endpoint (premature death), the NEJM editors have ignored the mutual impacts of each other on their results.
  6. And are HSPH researchers/administrators really that ignorant of what their colleagues are publishing?

Is this dishonesty or just mere incompetence? We report. You decide.

###

Improving diet quality over time linked with reduced risk of premature death
HARVARD T.H. CHAN SCHOOL OF PUBLIC HEALTH

Key takeaways:

  • Improving the quality of one’s diet over time, even with modest changes, is associated with lower total and cardiovascular mortality
  • Study supports a focus on overall healthy eating patterns, rather than on individual nutrients

Boston, MA – People who improve the quality of their diets over time, eating more whole grains, vegetables, fruits, nuts, and fish and less red and processed meats and sugary beverages, may significantly reduce their risk of premature death, according to a new study from Harvard T.H. Chan School of Public Health. It is the first study to show that improving diet quality over at least a dozen years is associated with lower total and cardiovascular mortality, and underscores the importance of maintaining healthy eating patterns over the long term.

The study will be published in the July 13, 2017 issue of the New England Journal of Medicine.

“Overall, our findings underscore the benefits of healthy eating patterns including the Mediterranean diet and the DASH diet. Our study indicates that even modest improvements in diet quality could meaningfully influence mortality risk and conversely, worsening diet quality may increase the risk,” said lead author Mercedes Sotos-Prieto, who worked on the study while a postdoctoral fellow in the Harvard Chan School Department of Nutrition and who is currently an assistant professor of nutrition at Ohio University.

Sotos-Prieto and colleagues analyzed the association between changes in diet quality among nearly 74,000 adults over a 12-year period (1986-1998) and their risk of dying over the subsequent 12 years (1998-2010). Data came from two long-term studies, the Nurses’ Health Study and the Health Professionals’ Follow-up Study, in which participants answered questions about their diets every four years and about their lifestyle and health every two years.

The researchers assessed people’s diet quality by using three different scoring methods: the 2010 Alternate Healthy Eating Index, the Alternate Mediterranean Diet score, and the Dietary Approaches to Stop Hypertension (DASH) diet score. Each of these methods assigns scores to various types of food or nutrients; less healthy foods or nutrients have lower scores and healthier foods or nutrients have higher ones.

The study found that improved diet quality over a 12-year period was associated with reduced risk of death in the subsequent 12 years, no matter which score was used. Food groups that contributed most to an improvement in diet quality were whole grains, fruits, vegetables, and fish or n-3 fatty acids.

A 20-percentile increase in diet-quality scores–the kind of increase that can be achieved by swapping out just one serving of red or processed meat for one daily serving of nuts or legumes–was linked with an 8%-17% reduction in the risk of death, depending on the diet score. In contrast, worsening diet quality was associated with a 6%-12% increase in the risk.

Among those who maintained higher rather than lower scores according to any of the three healthy diet patterns for 12 years, there was a 9%-14% reduction in mortality from any cause. Among those who had relatively unhealthy diets at the beginning of the study but whose diet scores improved the most, the risk of death in subsequent years was also significantly reduced.

“Our results highlight the long-term health benefits of improving diet quality with an emphasis on overall dietary patterns rather than on individual foods or nutrients. A healthy eating pattern can be adopted according to individuals’ food and cultural preferences and health conditions. There is no one-size-fits-all diet,” said Frank Hu, professor and chair of the Harvard Chan School Department of Nutrition and senior author of the study.

###

Other Harvard Chan authors included Shilpa Bhupathiraju, Josiemer Mattei, Teresa Fung, Yanping Li, Walter Willett, and Eric Rimm.

The study was supported by grants (HL034594, HL088521, HL35464, HL60712, P01 CA055075, P01 CA87969, UM1 CA167552, and UM1 CA186107) from the National Institutes of Health. Sotos-Prieto was supported by a postdoctoral fellowship from the Fundación Alfonso Martin Escudero.

“Association of Changes in Diet Quality with Total and Cause-Specific Mortality,” Mercedes Sotos-Prieto, Shilpa N. Bhupathiraju, Josiemer Mattei, Teresa T. Fung, Yanping Li, An Pan, Walter C. Willett, Eric B. Rimm, and Frank B. Hu, NEJM, July 13, 2017, doi: 10.1056/NEJMoa1613502

9 thoughts on “Are the New England Journal of Medicine and Harvard T.C. Chan School of Public Health Dishonest or Just Incompetent?”

  1. “Healthy” is a value judgement. If you describe a diet as “healthy “the research is an infomercial. Everybody, including the Harvard blog, says that reporting relative risk is misleading. This is not good science. You can decide the motivation.

  2. “Among those who maintained higher rather than lower scores according to any of the three healthy diet patterns for 12 years, there was a 9%-14% reduction in mortality from any cause. Among those who had relatively unhealthy diets at the beginning of the study but whose diet scores improved the most, the risk of death in subsequent years was also significantly reduced.”
    BASED ON WHAT?

  3. We can only hope, and be thankful in anticipation, that the general public is forever exposed only to such life-shortening factors as campfire fumes and hot dogs…
    Should the majority ever have their attention focused on trivial childish hypersubjective junkscience ‘research’ on such subjects as ‘air quality’, ‘global temperature’, ‘substandard diet’, ‘pedagogy’ etc., mean life-expectancy will surely plummet although the death-rate will, mercifully, never drop below 100.0000%

  4. I guess “what-we-say” is more important than ” these-are-the-facts”. How about, this is the null hypothesis, this is our experiment/test design, these are our test samples and subsets and their sizes, how we selected them, characteristics of each set and subset and of individuals within each, and so on. The information should allow anyone and any group to follow all steps and assumptions and be able cross check everything and look for repeatability. I wonder how many “scientists” versus scientists do that?

  5. The basic problem here is simply that “premature mortality” is not the same as “mortality”. Mortality is individual and has “cause(s) of death”, which is always a proximate link to one or more events or disease. Premature mortality is a population attribute, aggregated for a cause – as reported or as concocted – and is in turn attributed (pardon the irony) to “risk factors”.

    Read the Global Burden of Disease reports. They apply the power of algorithms and computing to make up causes of death, then distribute the blame to risk factors. There is NO claim that “risk factors” are causes.

    Attributability is not causality, and attributable is not avoidable. EPA and public health ideologues abuse language in order to create hysteria. That no PM2.5 is “safe”, while the question is, what is safe enough for risk of a particular disease profile, given that there are many sources of PM2.5?

  6. Perhaps a “healthy diet” is defined by what an “unhealthy diet” is:

    – Arsenic
    – Cyanide
    – High concentration HCl
    – Plutonium

Leave a Reply

Your email address will not be published.