Do Dogs Make Their Owners Healthier? What the Studies Really Show

In brief

Almost all studies on dogs and health are observational studies – they show associations, not causes. That doesn’t make the findings worthless, but it does change what you can conclude from them.

The difference is not academic hair-splitting. “Dog owners get more exercise” can mean the dog gets them moving. But it can also mean that people who already like to be active are more likely to get a dog. Both explanations fit the same data – and only the first justifies saying that a dog is good for your health.

That is no reason for disillusionment. It is a reason to look closely at which effects are well supported, which are plausible but unproven, and which are simply oversold.

1. Physical activity: the best-supported association

If any health effect of dog ownership has substance, it is this one.

A study from the University of Liverpool led by Carri Westgarth (2019, Scientific Reports) compared 385 households in West Cheshire: 191 adult dog owners, 455 without a dog, and 46 children. The result was clear – and remarkable in one respect: not only did dog owners walk far more often, but dog walking did not replace other exercise either; it came on top. The rest of their physical activity did not differ between the groups.

Two limitations belong right next to this. First, the main data were self-reported; only a small subsample of 28 people wore activity monitors as a check. Second, this too is a cross-sectional study – it shows that dog owners move more, not that the dog causes it. Self-selection remains a possible explanation.

Still, the fact that a dog has to go outside every day, and that this activity comes in addition to everything else, is the mechanism with the strongest support. If getting more exercise is one of your reasons for getting a dog, you are relying on the most solid part of the research.

👉 Why mental stimulation matters as much as physical exercise: Mental stimulation, scent work and creative activities

2. Cardiovascular health and life expectancy: an instructive dispute

This point deserves a closer look, because it is so often oversimplified.

A widely cited meta-analysis (Kramer and colleagues, 2019, Circulation: Cardiovascular Quality and Outcomes) pooled ten studies with a combined total of about 3.8 million people and found roughly a quarter lower all-cause mortality in dog owners compared with non-owners, with a stronger association in people who had had a heart attack.

So much for the headline. The revealing part comes afterward: a group of researchers reworked the analysis (Bauman and colleagues, 2020, same journal) and pointed out that the original analysis had mostly used unadjusted figures – with no adjustment for known confounders such as baseline fitness, income or general lifestyle. When the adjusted values are used, the association comes out weaker.

What you take away from this is less a number than a principle: an association between dog ownership and a longer life is plausible and supported by data – but how strong it is, and how much of it is due to the dog itself rather than to the kind of people who get dogs, remains open. Whenever you come across a figure like this, ask whether it was adjusted for confounders.

3. Stress and hormones: real, but overloaded

The line “Petting lowers cortisol, contact releases oxytocin” is part of the standard repertoire. It isn’t wrong, but it carries less weight than it is meant to.

For cortisol, there are studies showing a decrease during calm contact with animals – mostly with small samples and over short periods. On top of that, a cortisol level on its own, without the context of behavior, does not distinguish between strain and relaxation.

With oxytocin, caution is called for on two fronts. First, the evidence is inconsistent: the much-cited study by Nagasawa and colleagues (2015) examined 30 human–dog pairs; of the follow-up studies on a mutual rise, only some found confirmation, while others found no effect. Second, a rise does not automatically mean well-being – in humans, higher baseline levels have been linked to increased anxiety and greater susceptibility to stress. As a simple lever – “more oxytocin, more well-being” – the mechanism doesn’t hold up.

👉 In depth: Oxytocin in dogs: bonding, trust and stress regulation

None of this diminishes the well-being many people feel in contact with their dog. It only qualifies the explanation for it. That closeness to a familiar dog feels good doesn’t need a hormone chart.

👉 How finely dogs respond to human feelings: How dogs perceive our emotions

4. Daily structure: easy to understand, hardly quantified

A dog brings fixed routines: feeding, walks, recurring rhythms. That this structure can have a stabilizing effect, especially in difficult phases of life, makes sense from a behavioral-biology point of view and is reported by many owners.

At the same time, this is exactly the area where hard evidence is thin. It therefore belongs under experience and plausibility, not under measured effect. And it has a downside that is rarely mentioned: the same responsibility that stabilizes can also become a burden in difficult phases – when you lack the energy to meet a dog’s needs.

👉 How a sound relationship with your dog develops: Building trust with your dog

5. Social contact: plausible, often reported

Dogs make encounters easier – on walks, in the park, at training class. Owners often report more casual conversations and a stronger sense of belonging. This, too, is mostly self-reported, but it is hardly a surprising effect: a dog is a natural conversation starter.

For completeness, the other side belongs here as well: not every encounter is welcome, and for people who are socially exhausted, a constant conversation starter can also be tiring.

👉 Why social contact matters for the dog, too: Social contact for dogs

6. Immune system and allergies: correct only when worded precisely

This is where claims often go too far. What has actually been studied concerns children who grow up with dogs, and the association with their later risk of allergies and asthma. The mechanism under discussion is early exposure to a more diverse microbial environment.

Two things follow from this. First, it can’t be transferred to adults: “A dog means you’ll catch fewer colds” cannot be derived from this research. Second, even the finding in children is not automatic – the evidence is mixed, and it is about probabilities, not guarantees of protection.

👉 What has actually been studied on allergy and asthma risk: Dogs, allergies and asthma risk

7. Therapy and medicine: a separate chapter, with its own caution

Deliberately involving dogs in therapy, care and education is something different from living with your own dog – and the research on it is mostly of low quality, with small samples and inconsistent methods. The most robust review on dementia, for example, found a small indication of fewer depressive symptoms and otherwise no clear evidence.

It belongs here because the same mechanisms are in the background – but the context is different, and claims about effects should be correspondingly cautious.

👉 Dogs and pain relief

8. The downside that belongs in articles like this

An article about the health benefits of dogs that only looks in one direction would be dishonest. For completeness:

• A dog costs money, time and flexibility. For people whose situation is already strained, that can be an additional burden.
• A dog who can’t be left alone or shows behavior that restricts everyday life creates stress instead of reducing it.
• And the crucial reversal: any health benefit probably depends less on a dog simply being there than on the dog doing well and the relationship being sound. An overwhelmed dog won’t get anyone out on relaxed walks.

👉 Teaching your dog to be home alone · Sleep and learning in dogs

9. Conclusion

Living with a dog goes along with measurably more physical activity, and this activity comes on top of the rest – that is the best-supported point. For stress regulation, social contact and daily structure, the benefit is plausible and widely reported, but less well supported. For cardiovascular health and life expectancy, there are indications whose strength is disputed. And for the immune system, the finding applies only to children, and only with reservations.

“A dog is man’s best friend” therefore remains a good line – it is just less a proven medical principle than a description of what many people experience with their dog. That is reason enough.

The most honest look at the health benefit leads back to the dog anyway: it most likely arises where the relationship is sound and the animal is doing well – not from its mere presence. The most important conclusion remains: a dog is not a health intervention. If you get a dog, do it for reasons that stand on their own – responsibility, time and life circumstances.

References

• Bauman, A., Owen, K. B., Torske, M. O., Ding, D., Krokstad, S., & Stamatakis, E. (2020). Does dog ownership really prolong survival? A revised meta-analysis and reappraisal of the evidence. Circulation: Cardiovascular Quality and Outcomes, 13(10), e006907. https://doi.org/10.1161/CIRCOUTCOMES.120.006907
• Kramer, C. K., Mehmood, S., & Suen, R. S. (2019). Dog ownership and survival: A systematic review and meta-analysis. Circulation: Cardiovascular Quality and Outcomes, 12(10), e005554. https://doi.org/10.1161/CIRCOUTCOMES.119.005554
• Nagasawa, M., Mitsui, S., En, S., Ohtani, N., Ohta, M., Sakuma, Y., Onaka, T., Mogi, K., & Kikusui, T. (2015). Oxytocin-gaze positive loop and the coevolution of human-dog bonds. Science, 348(6232), 333–336. https://doi.org/10.1126/science.1261022
• Westgarth, C., Christley, R. M., Jewell, C., German, A. J., Boddy, L. M., & Christian, H. E. (2019). Dog owners are more likely to meet physical activity guidelines than people without a dog: An investigation of the association between dog ownership and physical activity levels in a UK community. Scientific Reports, 9, 5704. https://doi.org/10.1038/s41598-019-41254-6

Notes on the sources

Nearly all statements in this article are based on observational studies, mostly cross-sectional. Such studies show associations, not cause-and-effect relationships; self-selection (people who get a dog may differ systematically from others) is a possible alternative explanation throughout.

The statements on physical activity are based on Westgarth and colleagues (2019): 385 households in West Cheshire, mostly self-reported data, with objective measurement only in a subsample of 28 people. The finding that dog owners more often meet physical activity recommendations, and that dog walking comes in addition to their other activity, is well documented but remains cross-sectional.

The statements on cardiovascular health and mortality are based on the meta-analysis by Kramer and colleagues (2019) of ten studies with about 3.8 million people. A published reanalysis (Bauman and colleagues, 2020) pointed out that the original analysis mostly used unadjusted estimates; after adjustment for confounders, the association comes out weaker. For that reason, we deliberately give only an order of magnitude and no firm effect size.

The statements on cortisol are based on studies with mostly small samples. The evidence on oxytocin is inconsistent and is presented in detail in the linked article; the point that a rise should not be equated with well-being is based partly on human research.

The statements on structure, social contact and everyday benefits are based mostly on self-reports and practical experience. The statements on the immune system and allergies refer to studies on early exposure in childhood and on allergy and asthma risk, and cannot be transferred to adults. The assessment of animal-assisted therapy is backed by sources in the linked article.

Key takeaways at a glance

Almost all studies are observational – they show associations, not causes. That doesn’t make them worthless, but it changes what can be concluded from them.

The best-supported association is physical activity. Dog ownership is associated with more physical activity – on average, dog owners walk more. Whether the dog is the cause cannot be concluded from these data.

For cardiovascular health and life expectancy, the picture is more contested than headlines suggest.

A dog does not automatically extend your life. People who choose to get a dog differ in many ways from people without one.

The oxytocin findings are inconsistent – and a rise does not automatically mean greater well-being.

Daily structure is easy to understand but hardly quantified. Plausibility is not evidence.

On allergies, the claim is only correct when worded precisely – sensitization, allergy and asthma are different endpoints.

The downside belongs in the picture: costs, restrictions, the strain of illness and loss – if you only mention the positive findings, your advice is incomplete.

The most important conclusion: a dog is not a health intervention. The decision to get a dog should be made for other reasons – responsibility, time and life circumstances carry it, not a possible side effect on your health.

Frequently asked questions