Cognitive Decline by Age: What's Normal and What's Not

Raymond Cattell's distinction between fluid and crystallised intelligence, developed in the 1940s and 1950s and refined with John Horn, remains one of the most useful frameworks for understanding cognitive change across the lifespan. Fluid intelligence encompasses the abilities most people associate with 'being smart in real time': novel problem-solving, working memory, processing speed, and the capacity to reason about unfamiliar material without relying on prior knowledge. Fluid intelligence peaks in the early-to-mid 20s and then declines steadily throughout adulthood.

Crystallised intelligence, by contrast, encompasses the accumulated product of all that fluid reasoning applied over a lifetime: vocabulary, general knowledge, domain expertise, and the ability to apply learned procedures automatically. Crystallised intelligence grows well into the 60s and 70s before plateauing. The result is that most people experience their raw cognitive speed weakening while their practical wisdom deepens — a trade of speed for pattern recognition that evolution has apparently found adaptive.

Processing speed — the rate at which the brain completes basic cognitive operations — shows measurable decline from the late 20s, though it is typically not clinically significant until the 60s. Working memory capacity is relatively stable until around 50 and then undergoes more notable decline. Episodic memory (memory for specific personal events) begins declining in the 30s but slowly. Semantic memory (general factual knowledge and language) is largely preserved until the 70s. Prospective memory — remembering to do something at a future time, like take medication or call someone back — is particularly vulnerable to age-related change and is often the first functional memory complaint in healthy older adults.

Within attention, the picture is nuanced: selective attention (filtering relevant from irrelevant stimuli) holds up relatively well into old age. Divided attention — managing two tasks simultaneously — declines more substantially and earlier, consistent with reduced processing speed and working memory capacity. This means older adults can often match younger adults on single-focused tasks but struggle more when tasks require parallel management of multiple streams.

Several abilities either hold steady or genuinely improve with age. Vocabulary and verbal knowledge continue to grow into the 60s — this is the crystallised intelligence effect at work. Emotional regulation improves substantially with age: older adults are better at managing negative emotions, show reduced amygdala reactivity to negative stimuli, and bias their attention toward positive information in a phenomenon called the positivity effect, described by Laura Carstensen's socioemotional selectivity theory. This is not cognitive decline — it is a form of emotional optimisation.

Domain expertise continues accumulating throughout working life, allowing experienced professionals to compensate for processing speed declines through richer pattern libraries that make individual decisions less effortful. Research on expertise consistently shows that experienced physicians, chess players, and engineers make fewer errors despite slower raw processing speeds, because their accumulated pattern libraries front-load the answer before deliberate analysis is needed. Education level also acts as a protective factor: not because educated people decline less, but because higher education builds cognitive reserve — redundant neural pathways — that buffers the cognitive impact of age-related brain changes.

Physical inactivity is the strongest modifiable risk factor for accelerated cognitive decline. Sedentary lifestyle is associated with reduced hippocampal volume and accelerated grey matter loss. Cardiovascular disease and hypertension reduce cerebral blood flow, directly impairing the nutrient and oxygen supply that neurons require for efficient function. Type 2 diabetes damages cerebral blood vessels and is associated with roughly doubled risk of dementia. These are not abstract epidemiological associations — the mechanisms are well-characterised.

Chronic sleep deprivation accelerates the accumulation of amyloid-beta plaques in the brain — the same plaques implicated in Alzheimer's disease — because the glymphatic system that clears these waste products is primarily active during deep sleep. Social isolation is also an underappreciated risk factor: longitudinal studies consistently find that socially engaged older adults maintain cognitive function better than isolated peers, even after controlling for depression and general health. The mechanism likely involves the cognitive demands of social interaction sustaining neural circuit activity.

Normal age-related change involves taking longer to retrieve information, occasionally forgetting names before they come back, or needing more time on unfamiliar tasks. These reflect reduced access speed — the information is still stored, just harder to retrieve quickly. The analogy is a search engine with a slower index: the data is there; retrieval takes longer. This is annoying but not clinically significant.

The more concerning pattern is loss of stored information rather than just slowed retrieval: forgetting recent conversations that definitely happened, not just struggling to recall the exact words; getting lost in familiar environments that should be automatic; personality changes; inability to follow familiar multi-step instructions. Another red flag is loss of insight — normal aging is accompanied by awareness of the changes; a person who doesn't notice significant lapses that are obvious to others is showing a pattern more consistent with pathological change. If functional independence is affected — managing finances, medications, or navigation — that warrants clinical evaluation, not just reassurance.