An Older-Looking Brain Tracks Worse Chronic Pain – But Placebo Still Works
- In 84 people with temporomandibular disorder (TMD) and 84 age- and sex-matched healthy controls, two independent machine-learning models – Gaussian process regression on cortical thickness and a convolutional neural network on raw T1 volumes – both estimated that the TMD brains were "older" than their chronological age.
- A larger brain-age gap (estimated brain age minus chronological age) was associated with greater pain severity, and statistically mediated the link between chronological age and pain severity – older age predicted worse pain partly because it predicted an older-looking brain.
- The deep-learning model further linked older brain age to greater pain interference and higher odds of high-impact chronic pain, controlling for sex and race.
- Critically, neither estimated brain age nor the brain-age gap altered the magnitude of the placebo response evoked by verbal suggestion plus classical conditioning – analgesic placebo effects were preserved across the brain-aging spectrum.
Chronic pain has long been described as a condition that "ages" the brain, but most prior work stopped at correlation: people with persistent pain show patterns of cortical thinning and atrophy that resemble older brains. This study, led by a University of Maryland group, advances the argument in two ways. First, it uses brain age as a quantitative biomarker rather than a metaphor, deriving a single per-person number from structural MRI and asking whether that number carries clinical signal. Second, it tests not only whether an older brain predicts worse pain, but whether it predicts a worse response to a core endogenous pain-modulation mechanism – the placebo effect.
The mediation result is the conceptual core. Chronological age and pain severity are correlated, but the analysis shows that estimated brain age sits causally between them: the statistical path from age to pain runs substantially through the brain-age gap. This reframes "older patients hurt more" as "patients whose brains have aged more hurt more," and the two are not identical. Two people of the same calendar age can have very different brain-age gaps, and it is the gap – not the birthday – that tracks pain. For a clinician, this is the difference between an unmodifiable demographic and a candidate biomarker that could, in principle, be moved.
The convergence of two architecturally distinct models strengthens the finding. Gaussian process regression on cortical-thickness features and an end-to-end convolutional network on whole volumes encode brain aging through largely different feature spaces; that both flag the same TMD-versus-control difference, and both link older brains to worse pain, makes the result harder to dismiss as an artifact of one pipeline. The CNN additionally captured the more clinically loaded outcomes – interference and high-impact pain – suggesting that finer-grained volumetric information carries signal beyond cortical thinning alone.
The placebo result is the genuinely surprising part. One might expect that an aged, structurally compromised brain would mount a weaker endogenous analgesic response, since placebo analgesia depends on prefrontal and descending modulatory circuitry vulnerable to aging. Instead, placebo magnitude was independent of brain age. The descending and expectancy-driven machinery that produces placebo analgesia appears robust to the structural drift that worsens baseline pain. Practically, this decouples two things often assumed to decline together – and it argues against writing off older or structurally "aged" patients as poor responders to expectancy-based and conditioning-based interventions.
This is a cross-sectional design, so the brain-age-to-pain link is associative, not proven causal in time. But as a biomarker proposition it is clean: a reproducible structural index that tracks pain burden while leaving a key therapeutic lever intact.
Why Brain Age, Not Calendar Age
The clinical appeal of a brain-age gap is that it is potentially actionable. Sleep, vascular health, physical activity, and pain itself have all been linked to brain-aging trajectories. If the gap, rather than chronological age, is what tracks pain, then interventions that slow or reverse brain aging become legitimate targets for chronic-pain care – a hypothesis this study motivates without yet testing.
What It Means for Expectancy-Based Care
Because placebo magnitude held steady across brain age, conditioning- and suggestion-based components of treatment – therapeutic framing, ritual, structured expectancy – should not be presumed less effective in patients with older-looking brains. For practitioners, the takeaway is to keep deploying expectancy-driven elements regardless of a patient's structural brain age, while treating a high brain-age gap as a marker of likely greater pain burden rather than of treatment resistance.
Two people of the same age can have very different brains – and it is the brain's age, not the birthday, that tracks how much chronic pain hurts.
Cross-sectional design precludes causal claims about temporal order; the sample was a single chronic-pain phenotype (temporomandibular disorder) drawn from one cohort, limiting generalizability; brain-age estimates depend on the training data and model architecture; and the placebo procedure was experimental rather than a clinical treatment trial.