A very large prospective study assessed dietary patterns and healthy aging
“Our findings provide evidence to support that adherence to healthy dietary patterns represents a potential strategy for healthy aging, patterns that particularly are richer in fruits, vegetables, whole grains, unsaturated fats, nuts and legumes, that include some dairy products, and that are lower in trans fats, sodium, and red and processed meats.”
In the journal*Nature Medicine*this week there was an important open-access publication about a large combined cohort of over 105,000 health professionals prospectively followed for 30 years.Only 9.3% reached the age of 70 years with “healthy aging” —without 11 major chronic diseases and no impairment of cognitive or physical function or mental health. In this edition of Ground Truths, I’ll review the report’s findings, limitations, and put it into context with other studies that have assessed the association of diet and freedom from major age-related diseases.
The New Study
There have been hundreds of reports from the Nurses’ (all female) and Health Professionals (all male) studies over the years from these 2 very large cohorts that began their assessment in 1986. The data for the participants relies on mailed questionnaires for lifestyle and medical status completed every 2 years with a high rate (90%) of follow-up. While self-reported data can be regarded as a clear limitation, and used in most nutritional observational studies, the authors have previously published on the reproducibility and validity of such data.
Specifically for the dietary data, the participants reported their intake of 130 items at baseline (in 1986) and then every 4 years, from which eight dietary pattern scores were computed (and adherence by quintiles) along with the calculated Alternative Healthy Eating Index (AHEI) based on 11 foods: fruits, vegetables, nuts and legumes, alcohol, red and processed meat, whole grains, sodium, trans fats and long-chain n-3 (omega-3) fatty acids.
The 8 Diet Patterns Assessed for Adherence
aMED (Alternative Mediterranean Index)— adherence to the Mediterranean diet, which is rich in plant-based foods, healthy fats, and lean proteins
DASH (Dietary Approaches to Stop Hypertension)—reduced sodium and increased fruits, vegetables, whole grains, and lean protein
MIND (Mediterranean—DASH Intervention for Neurodegenerative Delay)—increased whole grains, green leafy vegetables, berries, nuts, beans, fish, and poultry
hPDI (Healthful plant-based diet)—maximized fruits, vegetables, whole grains, legumes, and nuts
PHDI (Planetary Health Diet Index)—Increased fruits, vegetables, whole grains, nuts, and legumes, minimal red and processed meats and eggs
EDIP (Empirically inflammatory dietary pattern)— inflammatory index of processed meat, red meat, fish, vegetables, refined grains, drinks, and tomatoes
EDIH (Empirical dietary index for hyperinsulinemia)— score for foods to predict insulin secretion
(UPF) Ultra-processed food consumption
The Healthy Aging Outcome
The freedom from 11 major diseases were the following: cancer (except for non-melanoma skin cancers), diabetes, heart attack, coronary artery disease, congestive heart failure, stroke, kidney failure, chronic obstructive pulmonary disease, Parkinson disease, multiple sclerosis and amyotrophic lateral sclerosis.
The Confounders and Adjustment
There are an untold number of covariates that could influence the outcome. To attempt to counter this issue, there was statistical multivariate regression for known factors such as: age, sex, ancestry, BMI, socioeconomic status, physical activity energy (caloric) intake, smoking, use of multivitamins, living alone, family history of dementia, heart attack, cancer diabetes, history of depression, postmenopausal status and hormone use, alcohol intake, and more.
Results
Who were the 9% of participants who reached age 70 and 75 years with healthy aging? The AHEI index had the strongest association with healthy aging, more than doubling the odds of reaching age 70 (and for age 75, odds ratio of 2.24). In the highest to lowest quintile adherence plot below you can see the dietary patterns were associated with higher odds of healthy aging. The AHEI led the pack but there was considerable overlap (as seen by the horizontal bars representing the 95% confidence intervals).
Specific food intake with a “heat map” (darker the color, more the association) odds-ratio (green-increased, red-decreased) for healthy aging were presented below, also after multivariable adjustment. So the association for the worst food for healthy aging were trans fats with an odds ration of 0.52 (about half as likely to achieve healthy aging). In contrast, the best food associations were fruit, monounsaturated fats, whole grains and vegetables, with close to doubling the odds of healthy aging.
Other key findings included:
Higher intake of UPFs was associated with lower odds of healthy aging
The associations for diet and healthy aging were stronger in women, smokers, BMI > 25, low physical activity, and lower socioeconomic status (each with statistically significant subgroup interactions). These subgroups are generally indicative of higher risk and a stronger link.
Limitations
While very large with respect to number of participants and outstanding for 30 years of follow-up, there are certainly many issues with interpreting the results. That includes theywere all health professionals, thedata were acquired through self-reported questionnaires, all theextensive statistical adjustment is still incomplete and leaves room for confounding factors that were not included. Morevover, it can yield spurious results as an outgrowth of the*multitude of comparisons.*That was further complicated by analyzing and reporting for other outcomes like intact cognitive function, intact physical function, intact mental health, freedom from chronic dresses, and survival to 70 years age. I view that as a mistake by these researchers, adding other outcomes beyond the primary one of healthy aging based on an observational study, further introducing spurious findings from tons of comparisons that could have been avoided.
When I posted the study results on social media this week, many reacted to the notion that fast foods and fried foods were good for healthy aging and right up there with fish and seafood. That’s a wrong interpretation since, as you can see in the Figure above, for healthy aging, these foods were in the neutral zone, showing little association one way of another, no less the concern about multiple comparisons. Yes, it defies the expectation that fish is good for you and fried or fast foods are bad. But the signal from this work was weak for either of these types of food. The heat map is best interpreted when the green or red colors are dark, with the examples I gave with the Figure.
In my mind, a major limitation is the lack of a prospective hypothesis for these cohorts at the outset. These large cohort studies were not undertaken with a prespecified data analysis for healthy aging. Medical research is best when there is a specific prospective hypothesis being tested, not when a large data set is being extrapolated for a question thought up along the way. It’s about an experiment directed to answer a question, rather than relying on a form of data-dredging that is a big drawback of observational studies.
Nevertheless, while we’d love to have randomized trials of such diets in large number of participants and decades of follow-upwith objective assessment (recently the gut microbiome composition was suggested to provide this potential), the simple fact is that we are not going to have such data to work from. With all the warts of such studies as the new one reviewed here, it’s likely the best we can draw from to understand links between diet and healthy aging. And remember a link is not a cause and effect relationship. That’s possible, but not established.
Context
The new report on a strong association on healthy foods, mainly a Mediterranean-typediet, is fully consistent with, and a replication/extension of many others independent studies. Some are reviewed in the Discussion section of the paper. In fact, in 2013, these same researchers published apaper based on 10,670 women in the Nurses’ study with 15-year follow-up and similar results.
InThe Atlantic magazine this week [among other big Signal things ;-)] there was anarticle(←gift link) on the loss of appeal of plant-based eating. The lede is “Making America healthy again, it seems, starts with a double cheeseburger and fries. Earlier this month, Health Secretary Robert F. Kennedy Jr. visited a Steak ’n Shake in Florida and shared a meal with Fox News’s Sean Hannity.. The setting was no accident.”
In it, Yasmin Tayag reviews the reasons for and evidence for the decline of diets emphasizing plant-based eating. This is unfortunate since it goes against the extensive data in hand.
In my new book SUPER AGERS, I review the evidence linked to healthy aging with well over 1,800 citations from the literature. The biggest chapter is on lifestyle + factors (the + indicates factors well beyond diet, sleep and exercise) and within it is a section calledGood Food where I review the data from diet studies. Here is a brief excerptof that section:
***********
Many studies have confirmed the beneficial effects of “healthy eating.” Three large prospective studies with cumulatively more than 165,000 women and 43,000 men, followed up for thirty-two years, were consistent in showing a 15 to 20 percent lower cardiovascular risk linked to eating well. A lower risk of all-cause mortality was found in two of these prospective studies, with a dose relationship seen with cardiovascular, cancer, and neurodegenerative disease. For all these ailments, the higher the dose of healthy foods, the lower the risks. That’s good, but what exactly does healthy eating mean?
We know it includes fruits and vegetables; legumes, such as lentils and garbanzo beans; whole grains; nuts and seeds; healthy fats, such as olive oil and avocados; and fatty fish, such as salmon and tuna, rich in omega-3 fatty acids. Consumption of olive oil, per se, has been tied to about a 20 percent lower all-cause mortality, along with reduction of cardiovascular, cancer,and neurodegenerative-related deaths. Among more than ninety-two thousand participants followed for twenty-eight years, there was a significant lower risk of dementia: 28 percent less risk if a person consumed 7 grams (half a tablespoon) a day or more of olive oil compared with rare or no consumption.
In aggregate, these foods provide good sources of dietary fiber, something we all tend to need more of. The recommendation for about 30 grams of fiber per day, adjusted by age, gender, and body size, is derived from a meta-analysis of 185 studies that showed people who consumed a high fiber diet had 31 percent less heart disease and 16 percent less type 2 diabetes or colon cancer. Fiber is the polar opposite of ultra-processed foods—it slows digestion, lessens glucose spikes, and helps reduce cholesterol. What is referred to as the Western diet, rich in ultra-processed foods, is pro-inflammatory and linked with metabolic dysfunction, cardiovascular disease, diabetes, and colon cancer. A randomized trial of a high fiber, “microbiome enhanced” diet compared to the Western diet with the same percentage of calories (isocaloric) for macronutrients led to less gut absorption of over one hundred calories per day (with some people over four hundred calories per day). Another way to put it: pooping calories instead of absorbing them!
**********
There are many other publications I reviewed elsewhere in the book linked to reduction of the 3 major age-related diseases (cancer, cardiovascular, and neurodegenerative), and include some randomized trials. Suffice it to say all of the available data from peer-reviewed publications support the findings of the present study. With the critique I’ve laid out above, it’s so important to see back up from the whole body of evidence. So I’d conclude that the new study is strongly reinforced by many others, with a common thread of imitations, but in totality very much concordant on the diet associated with healthy aging.
A quick poll:
Loading...
*************
As you can imagine, I’m excited to get my new book out. I've worked with my publisher, Simon & Schuster, to share an exclusive offer just for Ground Truths subscribers. Through May 5, you can pre-order a hardcover copy of the book atBookshop.org and get 15% off with codeSUPERAGERS15. It’s notable that Bookshop.org supports independent bookstores across the country. Of course there’s alsoAmazon which currently offers a discount on the book, but it’s certainly not helping these bookstores.
In this post, I’ve only scratched the surface about the content of the book. Here’s the back cover to give you an idea of what some people had to say about it.
Thanks for reading and subscribing to Ground Truths.
If you found this interesting please share it!
That makes the work involved in putting these together especially worthwhile.
All content on Ground Truths—its newsletters, analyses, and podcasts, are free, open-access.
Paid subscriptions are voluntary and all proceeds from them go to supportScripps Research. They do allow for posting comments and questions, which I do my best to respond to. Please don't hesitate to post comments and give me feedback. Many thanks to those who have contributed—they have greatly helped fundour summer internship programs for the past two years