Tracking more than 5,000 middle-aged British civil servants for a decade, researchers found that people with higher blood levels of the inflammatory signaling molecule interleukin-6 (IL-6) in midlife showed measurably faster decline in reasoning and a higher chance of meaningful drop on a global cognitive test years later. C-reactive protein (CRP), the inflammation marker most doctors actually order, showed no such predictive power once the analysis accounted for other health factors. The combined effect of high IL-6 was equivalent to roughly 3.9 extra years of cognitive aging for reasoning ability. The effects are real and statistically robust, but individually small, and the study can show association only, not causation.
For years the dominant story of dementia has been about amyloid, the sticky protein that clogs the aging brain. This study, drawn from the long-running Whitehall II cohort of British government workers, adds weight to a competing and complementary idea: that chronic, low-grade inflammation smoldering in the body during middle age is quietly linked to how the brain ages later.
The researchers, led by Archana Singh-Manoux at the French national health institute INSERM and University College London, did something most previous studies had not. Instead of measuring inflammation once, they measured two markers, IL-6 and CRP, twice, five years apart, and averaged them. Single inflammation readings bounce around with every cold, stubbed toe, or bad night. Averaging two readings taken years apart gives a cleaner picture of a person’s long-term inflammatory tone, and it is a genuine methodological upgrade over the field’s earlier work.
They then followed the participants’ thinking skills, memory, reasoning, and verbal fluency, across three testing waves spanning ten years, starting when the volunteers were on average 56 years old.
The headline finding splits the two markers apart. People in the highest third of IL-6 declined faster in reasoning than those in the lowest third, and were about 1.8 times more likely to show a significant drop on the Mini-Mental State Examination, a standard cognitive screen. CRP, by contrast, predicted nothing once the researchers accounted for obesity, smoking, cardiovascular risk, and chronic disease. This matters because CRP is the cheap, common test; IL-6 sits one step upstream in the biology and appears to carry the real signal.
The big idea is twofold. First, the window of relevance is midlife, not old age. Because these people were in their late forties to sixties and mostly free of preclinical dementia, the link is less likely to be the brain disease causing the inflammation and more plausibly inflammation traveling alongside, or contributing to, cognitive aging. Second, the marker matters. Watching CRP may miss what watching IL-6 catches.
The honest caveat, which the authors themselves stress, is that this is an association. It cannot prove that lowering IL-6 protects the brain, and the per-person effect is modest. But it sharpens a question worth billions in dementia research: is chronic inflammation a cause we can target, or just a passenger we can measure?
Actionable Insights
The single most useful takeaway is that if you or your clinician want a blood marker that tracks with long-term brain aging, IL-6 appears more informative than the standard CRP test, at least in midlife. In this study CRP predicted essentially nothing about cognitive decline after accounting for other health factors, while IL-6 did.
How big is the benefit of being a low-inflammation person rather than a high-inflammation one? Think of it in terms of brain years. Compared with the lowest IL-6 group, the highest IL-6 group performed as though their brains were about 2.0 years older for memory, 2.2 years older for reasoning, and 3.0 years older for verbal fluency at the starting line, and then declined faster on top of that. Combining the starting gap and the ten-year decline, high IL-6 corresponded to roughly 3.9 years of extra cognitive aging for reasoning. In plain effect-size terms these are small differences (standardized differences of about 0.08 to 0.13, where 0.2 is considered a small effect), so no single person should expect a dramatic change, but across a population the signal is consistent.
The practical lever is that IL-6 is not fixed by fate. In this same dataset, high IL-6 traveled with obesity, smoking, diabetes, and cardiovascular risk, all modifiable. The reasonable, low-regret interpretation is that the lifestyle work that lowers systemic inflammation, maintaining healthy weight, not smoking, exercising, and controlling metabolic disease, aligns with the direction this data points. The study does not prove those steps protect cognition, but they carry little downside and the correlational arrow points the right way. [Confidence: Medium]
Context and Source
Open access paper (2014 paper): Interleukin-6 and C-reactive protein as predictors of cognitive decline in late midlife.
Authors and institutions: Archana Singh-Manoux and colleagues, spanning INSERM (Centre for Research in Epidemiology and Population Health, Villejuif, France), the Department of Epidemiology and Public Health at University College London (UK), and the University of Versailles St-Quentin (France). The underlying cohort is the Whitehall II study of British civil servants.
Country: Data collected in the United Kingdom; lead analysis based in France and the UK.
Journal: Neurology (official journal of the American Academy of Neurology)
Impact evaluation: The impact score of this journal is approximately 8.5 to 8.9 (2024 to 2025 Journal Impact Factor; it was near 8.3 at the time of this article’s 2014 publication), evaluated against a typical high-end range of 0 to 60+ for top general science and flagship clinical journals, therefore this is a High impact journal.