Chronic inflammation (inflammaging)

With age, inflammation stops switching off and settles into a low, body-wide hum with no infection left to fight. That state is real, cheap to measure and a strong predictor of heart disease and death, and it responds to ordinary habits. What is oversold is the idea that pushing the number down with a pill will buy you a longer life: no drug has been shown to do that in healthy people.

Lower it the unglamorous way: a Mediterranean-style diet with plenty of fibre, regular strength and aerobic exercise with proper recovery, losing excess belly fat. Getting insomnia treated may help too; it lowered CRP in one trial. A routine hs-CRP blood test is a cheap read on where you stand; treat a high result as a reason to look for its cause, not as a number to chase. Fish-oil capsules and anti-inflammatory drugs are not worth taking for this. If you have coronary heart disease, low-dose colchicine is a prescription option to discuss with your cardiologist.

Inflammation that never switches off

Acute inflammation is one of the body's most useful tools: a short, self-limiting response that destroys germs, clears debris, starts repair and then stops. Inflammaging is what happens when that programme never fully shuts down. It is sterile, meaning germ-free: the alarms are set off by molecules released from the body's own stressed and dying cells, which the immune system mistakes for an infection, and they keep a master switch of the inflammatory response, a protein called NF-κB, turned on.[1] Because the signals travel in the blood, scattered local damage becomes a body-wide inflammatory tone.[2]

It is the eleventh of the twelve hallmarks of aging, and one of the best supported: that inflammation rises with age in industrialised countries, and that it predicts disease and death, is well established in large human studies rather than mice. It may not be inevitable, though. When researchers applied an inflammaging measure built from 19 immune signals in an Italian group to Amazonian forager-farmers and an Indigenous Malaysian population, the signals clustered differently, showed little or no rise with age and were not linked to age-related disease.[3]

The goal is not to abolish inflammation. The acute, resolving kind is essential; what you want back is its off-switch.

Where it comes from

Several age-related failures feed the same inflammatory fuel into the blood. Each is well described; how much each contributes is not known.

  • Zombie cells. Senescent cells, damaged cells that stop dividing but refuse to die, keep releasing a mix of inflammatory signals, and recent work shows they help drive the chronic inflammation of ageing.[4]
  • Leaked mitochondrial DNA. Damaged mitochondria, the cell's power plants, spill their own DNA into the cell, where sensors built to detect bacteria read it as an infection and fire.
  • An ageing immune system. Blood-cell production skews toward pro-inflammatory types, the thymus makes far fewer fresh immune cells, and worn-out immune cells accumulate. This immunosenescence both fuels inflammaging and is worsened by it.[5]
  • Belly fat. Fat stored around the organs is an active inflammatory organ: as it grows, its resident immune cells release signals such as interleukin-6 (IL-6) and tumour necrosis factor alpha (TNF-α) into the blood.
  • A leakier gut, in some people. In animal studies the gut wall becomes leakier with age and lets bacterial fragments into the blood. In people the picture is less clear: direct tests found the barrier works about as well in healthy 65–75-year-olds as in young adults, and only older people with irritable bowel syndrome leaked more, in the small intestine rather than the colon.[6] See dysbiosis.

How it turns into heart disease, and possibly dementia

The link to the heart is a well-worked-out mechanism that fits the population data, though no trial has tested it end to end. In a healthy artery, insulin tells the lining to make nitric oxide, which relaxes the vessel and calms inflammation. When the lining becomes insulin resistant, a state that typically comes with high blood sugar, excess blood fats and inflammation, the nitric-oxide signal weakens while the signal that tightens vessels keeps working or even speeds up. Arteries drift toward stiffness, higher blood pressure and inflamed walls where plaque builds.[7][8]

The brain link is a working hypothesis, not an established chain. The proposed sequence is that inflammation loosens the blood-brain barrier, the brain's own immune cells stay switched on, and brain cells become insulin resistant and build up more of the amyloid and tau changes of Alzheimer's disease. Much of it comes from animal studies that do not all agree.[9] In people the support is mostly observational: long-running inflammatory conditions such as obesity, type 2 diabetes and rheumatoid arthritis are linked to faster cognitive decline, though not every study agrees.[10] That shared root is one reason the same habits tend to protect heart, metabolism and brain together; see dementia prevention.

Measuring it: a strong warning light, not a target

The everyday markers are cheap blood tests, and they predict hard outcomes. High-sensitivity C-reactive protein (hs-CRP) is the standard one: among 113,491 white Danish adults, the fifth with the highest CRP had about 70% higher risk of dying over follow-up than the lowest fifth.[11] Pooling 14 groups with 59,396 people, those in the top quarter for IL-6 had roughly 45% to 110% higher risk across nine heart and mortality outcomes than the bottom quarter.[12] The neutrophil-to-lymphocyte ratio, which comes free with any standard blood count, is higher in older age groups and, even after accounting for other risk factors, tracks the risk of death.[13]

A simple inflammation score holds its own against the DNA-methylation clocks. In 3,113 US adults with an average age of 68, a score built from seven blood markers of inflammation (CRP, IL-6, the neutrophil-to-lymphocyte ratio and four other research-lab markers) predicted four-year death slightly better on its own than any of the six DNA-methylation ageing clocks it was compared with; only age itself did better. Once the usual risk factors were added, the two performed about the same.[14] In another study of 970 adults, an inflammatory clock was more weakly linked to physical and mental capacity than the epigenetic clocks were, so the two kinds of measure look complementary rather than rivals.[15]

What a high reading cannot tell you is whether pushing the number down helps. CRP itself looks like a bystander rather than a cause: in genetic analyses, people carrying variants that raise their CRP did not die earlier.[16] And a drug that lowered CRP after heart attacks did not reduce further events (see the drug section below). Read a high value as a warning about what is driving it.

What lowers it: diet and weight

Whole eating patterns have better evidence than single supplements, and much of the benefit tracks weight loss and overall diet quality.

The Mediterranean diet has the most trial evidence: pooling 33 randomised trials with about 3,500 adults, it lowered high-sensitivity CRP and IL-6 against control diets, though not every marker moved.[17] An earlier pooling of 22 trials found clear drops only in IL-6 and IL-1β, with CRP not falling reliably.[18] The DASH diet has far fewer trials, but in those it lowered high-sensitivity CRP by about 1 mg/L against a usual diet and did neither better nor worse than other healthy diets.[19] The fair reading is that a whole-food, plant-heavy pattern helps and the Mediterranean version is simply the best tested; see dietary patterns and ultra-processed food.

Fibre and fish oil. Gut bacteria ferment fibre into butyrate, which calms immune signalling and fuels the gut lining.[20] In trials in adults with overweight or obesity, about 8 g a day more fibre lowered CRP slightly, by roughly 0.4 mg/L, an effect that only just reached statistical significance.[21] The marine omega-3 fats are raw material for signalling molecules that actively switch inflammation off, but a fish-oil capsule does little to the markers of healthy people: in the VITAL trial, 1 g a day for four years did not change CRP, IL-6 or TNF-α in about 1,000 adults averaging 65.[22] Drops in CRP show up mainly in people who already have heart disease, high blood pressure or metabolic syndrome, at up to about 1.2 g of EPA plus DHA a day, and not in people who are simply overweight.[23] Dosing is covered under omega-3.

Calorie restriction has the strongest single trial. In the two-year CALERIE study, healthy adults of normal or slightly high weight who ate about 12% less lost about 10% of their body weight, and their CRP fell by more than a third; quality of life did not suffer. There were trade-offs: active thyroid hormone dropped, bone density at the spine and hip fell slightly more than in controls, and a few participants had to pause for anaemia.[24]

Fasting schedules add little on top of the calories. In a 12-month trial in 90 adults with obesity, time-restricted eating and ordinary calorie cutting each took off only 4–5% of body weight and neither lowered CRP, IL-6 or TNF-α against a control group, though the trial was probably too small to tell.[25] Pooled studies agree: time-restricted eating did not lower CRP or IL-6,[26] and in people with overweight, calorie restriction lowered CRP and IL-6 while intermittent fasting moved only TNF-α, from far fewer trials.[27] One review of alternate-day fasting found CRP fell only in trials where people lost more than about 6% of their weight.[28] The energy deficit and the fat loss appear to do most of the work, not the clock.

Exercise: the IL-6 paradox

Exercise lowers inflammation, even though working muscle releases a burst of IL-6, usually one of inflammaging's villains. Muscle-made IL-6 acts through a different route from the one that drives IL-6's harmful effects in disease,[29] and when healthy volunteers were given IL-6 at the level seen in hard exercise, two anti-inflammatory signals rose and TNF-α did not.[30]

Trials back the long-term effect. Pooling 49 randomised trials in people averaging 65 or older, exercise training produced a small-to-moderate drop in CRP and smaller drops in TNF-α and IL-6, with the clearest gains in people who already had a chronic disease.[31] A 2026 ranking across 55 trials in healthy middle-aged and older adults put strength training first for CRP and aerobic training first for IL-6 and TNF-α, which argues for doing some of each.[32] See zone 2, VO₂ max and resistance training.

The dose has a ceiling. After a long, exhausting endurance effort, inflammatory signals stay raised for about a day and immune defences can dip for anywhere from 3 to 72 hours; repeated hard bouts without recovery can keep inflammation raised.[33] Recovery is part of the prescription.

Stress and sleep: plausible, modestly supported

Stress. Under long-running stress, immune cells can stop responding to cortisol's calming signal. In healthy volunteers deliberately exposed to a cold virus, people facing a long-running stressful situation showed this resistance, and those with it were more likely to catch the cold and made more inflammatory signals once infected.[34] Stress-reduction training may lower inflammation a little, but the evidence is soft. Two large poolings, of 45 and 48 randomised trials, found that meditation or mindfulness programmes lowered CRP, by a small amount in the larger one.[35][36] But an early review found a clear CRP drop in just one of the six trials that measured it,[37] and a preregistered pooling of 32 trials found an effect on inflammation and stress markers that only just missed statistical significance, warning that weak studies and selective publishing may have biased the results.[38] See mindfulness meditation and stress.

Sleep. Pooled data from more than 50,000 people show that those with disturbed sleep, and those who sleep very long hours, have slightly higher CRP and IL-6, while short-term sleep loss in the lab does not reliably raise them.[39] Treating the problem may help: in a randomised trial in 123 older adults with insomnia, four months of cognitive behavioural therapy for insomnia (CBT-I) lowered CRP against a sleep-education class, and the drop was still there 16 months later.[40]

Drugs: some help in heart patients, none proven for healthy people

The one place drugs that calm inflammation have cut hard outcomes is cardiology, in people who already have heart disease. None of the longevity drugs below is established for healthy adults; they are covered in depth under geroprotectors.

In heart patients. In a trial of about 10,000 people who had had a heart attack and still had a raised CRP (2 mg/L or more), canakinumab, an antibody that blocks one inflammatory signal, cut repeat heart attacks, strokes and cardiovascular deaths by about 15% at the 150 mg dose, but fatal infections were more common and overall deaths did not fall.[41] Low-dose colchicine, an old and cheap gout drug taken at 0.5 mg a day, cut events from 9.6% to 6.8% in people with stable coronary disease,[42] but its largest trial, started right after a heart attack, found no benefit even though CRP fell.[43] Pooled, it cut major cardiovascular events by about a quarter with no clear effect on deaths from heart disease.[44] GLP-1 drugs also lower inflammation: in adults with heart disease and overweight, weekly semaglutide cut CRP by about 38% over two years, beginning before most of the weight came off, and part of its 20% cut in cardiovascular events may run through that, though the estimates are too imprecise to say how much.[45][46] Blocking IL-6 directly is still being tested in people with atherosclerosis and kidney disease,[47] and a trial in a common form of heart failure was stopped early for lack of effect.[48] None of these trials enrolled healthy people or tested whether the drugs slow ageing.

Metformin activates the cell's low-fuel sensor and damps inflammation in the lab.[49] The TAME trial, designed to test it against age-related disease in non-diabetic adults aged 65 to 79,[50] has never been fully funded.[51] As of late 2026 it is not registered on ClinicalTrials.gov and has enrolled nobody.

Rapamycin damps the inflammatory secretions of senescent cells in cell and mouse studies.[52] People using it for longevity usually take low weekly doses, aiming to avoid the immune suppression of daily transplant doses. Its best human test, a 48-week randomised trial in 114 healthy adults, missed its primary goal: visceral fat did not move. Stomach upset was more common on the drug, the encouraging lean-mass result came from just 8 women, and the trial used compounded drug delivering only about a third as much as generic sirolimus.[53]

Senolytics, drugs that clear senescent cells, carry real cautions: in a small uncontrolled pilot of the popular dasatinib-plus-quercetin combination, several older-style epigenetic clocks read older, not younger, after six months.[54] The human evidence is laid out under cellular senescence.

Further reading

  • Karpuzoglu E, et al. Inflammaging: triggers, molecular mechanisms, immunological consequences, sex differences, and cutaneous manifestations. Front Immunol 2025.[55]
  • Müller L, et al. Immunosenescence and inflammaging: Mechanisms and modulation through diet and lifestyle. Front Immunol 2025.[56]
  • Meier HCS, et al. Systemic inflammation and biological aging in the Health and Retirement Study. GeroScience 2023 — n=3,113; on its own the seven-marker score slightly out-predicted the six clocks it was compared with for four-year mortality, a near tie once other risk factors were added.[57]
  • Ravussin E, et al. A 2-Year Randomized Controlled Trial of Human Caloric Restriction: Feasibility and Effects on Predictors of Health Span and Longevity. J Gerontol A Biol Sci Med Sci 2015 — CALERIE Phase 2.[58]
  • Lin S, et al. Time-Restricted Eating Versus Daily Calorie Restriction: Effects on Inflammatory Markers over 12 Months in Adults with Obesity. Nutrients 2025.[59]
  • Shi X, Hu L, et al. Exercise workload: a key determinant of immune health — a narrative review. Front Immunol 2025.[60]
  • Black DS, Slavich GM. Mindfulness meditation and the immune system: a systematic review of randomized controlled trials. Ann N Y Acad Sci 2016.[61]
  • Moel M, Harinath G, Lee V, et al. Influence of rapamycin on safety and healthspan metrics after one year: PEARL trial results. Aging (Albany NY) 2025 — n=114; primary outcome null.[62]
  • Barzilai N, Crandall JP, Kritchevsky SB, Espeland MA. Metformin as a Tool to Target Aging. Cell Metab 2016 — the TAME rationale.[63]

— § —