What the research really says about loneliness and ageing

Loneliness is not an inevitable part of growing older. On the best Australian evidence, people aged 65 and over are no lonelier than the country as a whole, and the rate among them has fallen over the past two decades rather than risen – but that national average hides a smaller group for whom loneliness is persistent, and for whom the health consequences are real and well documented.

If you have arrived here because you are worried about a parent, that distinction is the useful part. The question worth asking is not “are old people lonely?” – as a generalisation, mostly not – but “is this person, in these circumstances, one of the ones at risk?” The research is genuinely helpful on that second question.

Loneliness and social isolation are not the same thing

Almost every confused conversation about this topic comes from treating two different things as one. The Australian Institute of Health and Welfare (AIHW) keeps them separate, and so should we:

• Social isolation is objective and countable. It describes how much contact a person actually has – how many people they see, how often, how connected they are to a network.
• Loneliness is subjective. It is the distressing gap between the connection someone wants and the connection they have.

The two overlap, but they come apart often enough to matter. A woman who lives alone, sees two people a week and is perfectly content is socially isolated but not lonely. A man in a busy residential aged care home, surrounded by people all day, can be profoundly lonely. Only one of those two situations calls for anything to change, and it is not always the one that looks worse from the outside.

This is why the most common family worry – “Mum lives on her own now” – is a reason to pay attention rather than a diagnosis in itself.

Older Australians are not the loneliest group, and the trend is going the right way

This surprises most people. Drawing on the long-running HILDA survey, the AIHW estimates that in 2024 about 15% of Australians were experiencing loneliness – 15% of men and 16% of women. Among people aged 65 and over, the rate was also around 15%, having declined from about 20% in 2001.

Australians aged 65 and over also consistently report the highest life satisfaction of any age group. Whatever else later life brings, the picture of a uniformly miserable, disconnected old age is not what the national data describes.

You will also see much larger figures quoted, and they are not wrong – they are measuring something broader. The 2026 Ending Loneliness Together State of the Nation report, based on responses from just over 4,000 Australians aged 12 to 95, found that around half reported some form of social disconnection – loneliness, social isolation, or both – with close to one in five reporting both at once. A broader definition produces a bigger number. That is a reason to read any single dramatic statistic carefully, not a reason to distrust the field.

The honest summary: at a population level, things are not deteriorating for older Australians. The risk is concentrated rather than widespread.

Where the risk actually concentrates

The averages hide the people who most need attention. Risk clusters around a handful of recognisable circumstances:

• The oldest old. Around 46% of Australians aged 75 and over rate their level of social support as low – a far higher share than in the wider population.
• Living alone, particularly for women. About a quarter of Australians aged 65 and over live alone. That splits sharply by sex: roughly 31% of older women compared with 18% of older men, and among those aged 85 and over, about 41% of women compared with 25% of men.
• Recent loss or upheaval. Bereavement, a partner moving into residential care, giving up driving, or moving away from a long-established neighbourhood all remove contact that used to happen without anyone arranging it.
• Where someone lives. AIHW notes that loneliness is more common among people who are single, living alone, living in aged care, or in regional and remote areas.

Notice how many of these are about losing incidental contact rather than losing relationships. A great deal of an older person's social life is a by-product of ordinary logistics – the walk to the shops, the drive to church, the chat at the club. When the logistics go, the contact goes with them, quietly, and often before anyone notices.

What the health evidence shows

Here the findings are sobering and reasonably consistent. The most authoritative recent synthesis is the 2020 report from the US National Academies of Sciences, Engineering, and Medicine, Social Isolation and Loneliness in Older Adults, which reviewed the evidence for adults aged 50 and over. Among its findings:

• Social isolation is associated with roughly a 50% increased risk of developing dementia.
• Among people with heart failure, loneliness is associated with a nearly four-fold increased risk of death, a 68% increased risk of hospitalisation, and a 57% increased risk of emergency department visits.

The other landmark is a 2010 meta-analysis by Julianne Holt-Lunstad and colleagues in PLOS Medicine, pooling 148 studies covering 308,849 people followed for an average of 7.5 years. People with stronger social relationships had a 50% greater likelihood of survival over the follow-up period. The authors noted the size of that effect is comparable to giving up smoking, and larger than well-known risks such as obesity and physical inactivity.

Three things the headlines get wrong

Because this evidence is striking, it gets repeated carelessly. Three caveats are worth holding on to.

1. These are associations, not proof of cause

Almost all of this research is observational. It shows that isolation and poor health travel together; it cannot fully separate cause from effect. Poor health, reduced mobility and early cognitive decline all make people withdraw – so some of the association certainly runs the other way. Researchers adjust for this as best they can, and the effect survives adjustment, but “loneliness causes dementia” overstates what has been shown.

2. Living alone is the weakest signal of the lot

This one is genuinely useful for families. Within the Holt-Lunstad analysis, the effect varied a lot depending on what was measured. Complex measures of social integration – the richness of someone's actual connections – showed the strongest association with survival. The weakest measure of all was the simple binary of living alone versus living with others, and its confidence interval crossed the line of no effect.

In plain terms: whether your mother lives by herself is a poor predictor. Whether she still has people she talks to, and things she turns up to, is a much better one. “Dad lives alone” is not the alarm. “Dad hasn't spoken to anyone but me in a fortnight, and he's stopped going to bowls” is.

3. The cigarette comparison is a scale, not an equivalence

You will often see loneliness described as being as harmful as smoking a certain number of cigarettes a day. That comparison comes from lining up effect sizes across different studies. It is a fair way to convey that the effect is not trivial. It is not a finding that loneliness and smoking do the same thing to a body, and it should not be read that way.

What appears to actually help

The intervention research is messier than the epidemiology, and honest reviews – including the National Academies report – find the evidence base thinner than anyone would like. Still, some patterns recur:

• Regularity beats intensity. Something small and predictable tends to do more for a person's sense of connection than an occasional long visit. Having something to expect is part of the benefit.
• Meaning beats headcount. Interventions built around genuine conversation and shared interest do better than those that simply increase the number of interactions.
• Protect the incidental contact. Anything that keeps the ordinary routines going – transport to the shops, staying in the club, remaining somewhere walkable – preserves social contact that nobody has to organise.
• Ask, rather than infer. Because loneliness is subjective, it cannot reliably be read from the outside. “Do you get enough company these days?” is a more informative question than any amount of observation from a distance.

If you are worried about a parent

Put the statistics down for a moment and look at change over time in the specific person. The signals worth taking seriously are not about living arrangements; they are about withdrawal from things that used to matter – dropped activities, unreturned calls, a shrinking set of names in conversation, less interest in food or going out, or a marked change after a bereavement or a move.

And be realistic about the shape of the gap. For most families, the problem is not an absence of love; it is Tuesday. It is the long stretch between Sunday visits, the days when everyone is at work and the phone does not ring. That gap is a practical problem, and practical problems have practical answers – a neighbour who checks in, a standing call from a grandchild, a community visitors scheme, a men's shed, a regular scheduled call.

What matters is that it is reliable. Contact someone can count on does more good than contact that depends on everyone remembering.