What the data actually shows
A consistent finding from epidemiological research is that the prevalence of anxiety and depressive disorders does not climb steadily with age — in many studies it is actually lower among older adults than among younger and midlife groups. The pattern is not universal across every condition or country, and measurement in older populations is genuinely hard (symptoms can overlap with physical illness, and some surveys underreach the oldest and most isolated). But the simple story of 'mental health gets worse as you age' is not what the broad data shows.
Alongside this sits work on emotional experience. Laura Carstensen's research on socioemotional selectivity and the 'positivity effect' finds that as people perceive their time horizons as shorter, they tend to prioritise emotionally meaningful goals and relationships, attend more to positive information, and report more stable, well-regulated emotional lives. Older adults in these studies often experience fewer negative emotions and recover from them faster than younger adults.
Zoom out to overall life satisfaction and a related pattern appears. A large body of work, much of it associated with David Blanchflower, describes a roughly U-shaped curve across adulthood: life satisfaction tends to dip in midlife and rise again afterward, with many people reporting higher wellbeing in their 60s and beyond than in their 40s. The exact shape is debated and varies by measure and country, but the upturn in later life is a recurring result.
Why this feels different from how it actually is
The dominant cultural story about ageing is one of loss — failing health, narrowing options, decline — so improvement in emotional life runs against the script we expect. We notice the visible losses (a body that changes, a shrinking calendar) far more readily than the quieter gains in perspective and emotional steadiness, which do not announce themselves.
Younger people also tend to mispredict their own future wellbeing, generally expecting to be less happy in old age than older people actually report being. So the version of 'old age' most of us carry around is forecast from fear rather than from data, and it skews dark.
And averages hide enormous variation. The midlife dip is real for many, the late-life rise is real for many, but plenty of people do not follow the curve at all. Grief, chronic pain, financial insecurity and isolation can pull an individual life sharply away from the average — which is exactly why a population pattern is context, not a prediction about you.
What the research says to do about it
The factors most consistently linked to better mental health across the lifespan are unglamorous and well-supported: maintaining close social connections, staying physically active within your ability, keeping a sense of purpose and routine, and getting adequate sleep. These are associations rather than guarantees, but they recur across studies and tend to matter more as other supports fall away in later life.
Protecting against isolation appears especially important as people age, because the social losses of later life are among its biggest mental health risks. Research on socioemotional selectivity suggests this happens somewhat naturally — older adults often prune toward fewer but more meaningful relationships — but deliberately sustaining those ties, rather than letting them erode, is the part within reach.
For specific, persistent symptoms, the evidence base for treatment in older adults is strong: both talking therapies and, where appropriate, medication can be effective at any age. Depression and anxiety in later life are treatable and are not a normal, untreatable part of getting old. This is general information, not medical advice — a qualified clinician can assess an individual situation.
What the research says does not help
Assuming low mood in an older person is 'just ageing' is one of the most documented mistakes in this area. Treating depression or anxiety as an inevitable feature of getting old can mean genuinely treatable conditions go unaddressed; later-life mental illness responds to treatment much like it does earlier.
Waiting to feel better through willpower or 'staying busy' alone tends not to resolve a clinical-level problem, just as it does not at younger ages. Distraction can blunt symptoms briefly but does not substitute for support when distress is persistent.
Catastrophising about the future is also counterproductive and, the forecasting research suggests, usually inaccurate: people routinely overestimate how unhappy older age will be. Dreading decline that the average data does not support adds present anxiety without changing anything.
Real numbers in context
The headline to hold onto is directional rather than a precise figure: across many epidemiological studies, anxiety and depressive disorders are no more common — and frequently less common — among older adults than among midlife or younger adults. Prevalence estimates vary widely by country, condition and method, so treat any single percentage with caution; the robust finding is the shape, not an exact number.
On wellbeing, the U-shaped pattern associated with Blanchflower's work suggests life satisfaction often bottoms out somewhere in midlife and rises afterward, with many people in their 60s and beyond reporting wellbeing above their midlife low. The size and even the existence of the curve is contested in some datasets — so the honest summary is that later life is, on average, not the emotional low point people expect, while remaining highly variable from person to person.