Does red meat increase dementia risks?

A recent study published in the journal Neurology by Yuhan and colleagues, suggested that consuming (processed) red meat increases the risk of dementia in US adults (1). The findings come from researchers led by a team at the Harvard TH Chan School of Public Health in Boston, based on records of 133,771 people in two US health databases. The participants were followed for up to 43 years, with dietary patterns charted against the probabllity of developing dementia.

Based on the 11,173 people in the study who did develop dementia over the research period, averaging a quarter of a serving or more of processed red meats every day (about half a slice of bacon) meant a 14% higher relative risk (more about relative risk later) compared to those who only ate a small amount of red meat, and that replacing red meat with other foods, such as poultry, fish, eggs, low-fat dairy, unuts and legumes, can reduce dementia risks.

As cognitive assessments were included in the health databases, the researchers looked in these associations too, and observed that for every extra serving of processed red meat averaged per day, roughly equating to one additional hot dog per day, it appeared to accelerate cognitive aging by about 1.6 years.

Unprocessed red meat, such as beef, lamb, or pork that hadn’t been preserved, was not associated with dementia risk. However, there was a link between processed and unprocessed meat and self-reported subjective cognitive decline.

Book one one of my programmes

NT for complex diseases

BRAINCOG optimise mind

Weight loss & detox

Wellwoman / man

Men’s body & mind

Better skin, look younger

Important nuances from the study by Yuhan and colleagues:

1. There was a tendency to lump different types of meats together in the food frequency questionnaires (FFQ). For example, in the unprocessed meat category, sandwiches and mixed dishes, such as lasagna, were ‘lumped’ with individual cuts, such as steak and lamb. Additionally, self-reported FFQs can also under or over-estimate foods consumed which can considerably skew outcomes.
2. There was tremendous healthy user bias. Example, quote “…participants who consumed more red meat had a higher prevalence of current smoking, hypertension, and diabetes and lower levels of dietary quality, education level, socioeconomic status, physical activity…”
3. “Unprocessed red meat intake was not significantly associated with any measure of objective cognitive function.” This is particularly notable in the context of the healthy user bias since the mechanisms they describe for ‘why’ red meat may harm the brain are mostly not specific to processed red meat and may be attributable to other dietary factors and underlying health conditions.
4. This is further notable because when you start to unpack the attempt at biological plausibility, the arguments implicitly implode. For example, the discussion reads, “diets rich in saturated fats can reduce brain growth factors…” but if you look at the study, the diet is literally called a “high fat sucrose diet.” thus fails to consider the potential effects of daily high (free) sugar intake.
5. The health databases captures self-reported cognitive assessments, which is subject to huge variances as people may under or over-represent their own cognitive health without an evaluation by a qualified medical professional, therefore can also skew study outcomes and potentially reduce the veracity of the findings.

Food questionnaires
Nutritional epidemiological studies typically ask people to keep a food diary or complete food frequency questionnaires (FFQ) about their intake over the past 24 hours, a week, or even several months. Biostatisticians have long warned that people can misremember or be reluctant to cop to what they consume. Although using FFQ has been shown to be a reliable method to assess dietary intakes in a population (2), a recent research article (3) found that many studies of nutritional epidemiology that try to link dietary exposures to disease outcomes are founded on really dodgy data, and that surveys that link dietary patterns to human health outcomes may be irredeemably biased, a finding that has also been echoed by other researchers (4).

Why reporting relative risk can be misleading
Studies often report relative risk (RR) when describing the association between consuming a food (e.g., processed red meat) and a health outcome (e.g., dementia). While relative risk provides a ratio of the likelihood of an event happening between two groups (e.g., meat-eaters vs. non-meat-eaters), it can sometimes exaggerate the perceived importance of the findings because it does not account for the baseline absolute risk (AR) of the outcome in the population.

Absolute Risk vs Relative Risk
RR compares the risk in one group to another (e.g., a 20% increased risk of dementia), whilst AR provides the actual likelihood of the event occurring in the population (e.g., the baseline risk of dementia is 1%). Relative risk can make small changes in absolute risk seem much larger and more alarming (5).

Imagine a study finds that people who consume processed red meat have a 20% higher RR of developing dementia compared to those who don’t eat processed red meat. So if the baseline risk (absolute risk) of dementia is 1% for the general population, a 20% increase means that the risk for processed meat-eaters rises from 1% to 1.2%.This RR increase is only 0.2 percentage points, which may not be clinically significant for most people, despite the seemingly alarming “20% higher risk” headline.

Reporting the 20% relative increase without absolute risk is typically favoured by news channels, as it can make the association seem more serious than it is, especially if there are vested interests or lobbying groups pushing to reduce red meat consumption for specific agendas. Thus, for an individual, the actual risk difference might not justify a major dietary change, especially if processed meat is consumed in moderation and other health habits are good.

Conclusions
RR can exaggerate perceived harm or benefit, especially for low-probability outcomes such as dementia. AR contextualises findings within real-world probabilities, helping individuals and clinicians make informed decisions. When evaluating diet-disease studies, it’s therefore important to consider both relative and absolute risks, along with broader lifestyle factors and the overall quality of the evidence presented. In summary, in my expert opinion, consuming small amounts of processed red meats once in a while especially in conjunction with a healthy lifestyle, regular physical activity, and a daily intake of fruits and vegetables, moderate consumption of processed meats is unlikely to cause significant health concerns in the big scheme of things.