Mubanga M, Byberg L, Nowak C, et al. Dog ownership and the risk of cardiovascular disease and death - a nationwide cohort study. Sci Rep. 2017;7(1):15821.
To investigate the association between dog ownership and risk of cardiovascular disease (CVD) and death.
This study reports on the analysis of 2 separate cohorts: a register-based prospective nationwide cohort (n=3,432,153) and a second smaller cohort, a self-reported group (n=34,202).
All Swedish residents aged 40 to 80 years on January 1, 2001 (N=3,987,937) were eligible for this study. This age range excluded younger individuals, who have a low risk of CVD, and the elderly, who are unlikely to own a dog. People who had not lived continuously in Sweden since 1987 (n=163,156) and individuals whose personal identity numbers were unconfirmed (n=5,057) were excluded from the study. In addition, people who had had inpatient visits (n=387,571) for CVD or CVD certain interventions were excluded. This still yielded an extremely large cohort (n=3,432,153). All Swedish residents are covered by the public health care system, and disease incidence is tracked through medical records.
This Swedish study is by far the largest and most thorough attempt to find an association between dog ownership and health to date.
A second subgroup cohort was created from participants in the “Screening Across the Lifespan Twin study” (SALT), a sub-study of the Swedish Twin Register (STR). A longitudinal study, SALT began in 1958 and has enrolled most twins born in Sweden since 1886. All SALT participants who responded to telephone interviews between 1998 and 2002 and who were between 42 and 80 years old in 2001 were eligible for study analysis (n=41,039).
Sweden requires all dogs in the country to have a unique identifier (tattoo or chip) registered with the Board of Agriculture. In addition, the Swedish Kennel Club registers all dogs with a certified pedigree. Approximately 83% of dogs in Sweden are listed on one of these registers. For the purpose of this study, dog ownership was defined as the dog being registered on one or both of these registers to either the owner or a partner. Information on dog breeds came from these registers.
Public registries and health records provided data on human participants’ death and cause of death. Four incident causes of death were recorded: 1) acute myocardial infarction; 2) heart failure; 3) ischemic stroke; and 4) hemorrhagic stroke.
In both single-person and multiple-person households, dog ownership was associated with significantly lower all-cause mortality and mortality from CVD.
The hazard ratios (HRs) for all-cause risk of death were 0.67 for single-person households (95% confidence interval [CI]: 0.65-0.69) and 0.89 for multiple-person households (95% CI: 0.87-0.91). For death from CVD causes, the HRs were 0.64 for single-person households (95% CI: 0.59-0.70) and 0.85 for multiple-person households (95% CI: 0.81-0.90).
In single-person households, dog ownership was inversely associated with cardiovascular outcomes (HR composite: 0.92; 95% CI: 0.89-0.94). Ownership of hunting breed dogs was associated with lowest risk of CVD.
Ownership of a mixed-breed dog was associated with higher risk of CVD (HR: 1.13; 95% CI: 1.09-1.17). The HR for all-cause mortality was <1 for all breed groups, with pointing dogs associated with the lowest estimate (HR: 0.60; 95% CI: 0.53-0.68) and mixed-breed dogs with estimates closest to 1 (HR: 0.98; 95% CI: 0.94-1.01).
Analysis of data from the twin cohort did not find significant associations between CVD and dog ownership.
Cardiovascular disease is the leading cause of death in the world today. In Europe, nearly half of all deaths result from CVD. Interventions that reduce risk by even small percentages will still have large impact. There is a belief that owning a dog reduces risk of CVD by providing social and emotional support and by increasing the owners’ physical activity. Dog ownership, particularly for single and older adults, may reduce the sense of social isolation and depression.1 It’s easier to make friends when you have a dog,2 and, no surprise, dog owners walk more than non-owners. A 2011 meta-analysis of 11 studies confirmed this;3 not only do people start walking more after they get a dog,4 they do so rain or shine and do not let inclement weather keep them from going out to walk.5
Prior studies that have attempted to demonstrate benefit of dog ownership have reported less than consistent results. Some studies have reported that ownership is inversely associated with CVD risk factors such as dyslipidemia, hypertension, and diabetes,6 but other studies have not.7
A Norwegian prospective cohort study published in June 2017 reported no difference in all-cause mortality between dog owners and non-owners. Dog owners in this study did not report any greater physical activity than non-owners; both groups reported just over 3 hours per week. Owners had virtually the same hazard of dying as non-owners (HR: 1.00; 95% CI: 0.91-1.09).8 Note that the number of participants in this Norwegian study was 28,746; this Swedish study included nearly 4 million participants, making it 133 times as large.
There have been 3 other studies conducted using data from the National Health and Nutrition Survey (NHANES), all relatively small, each including somewhere between 4,000 to 6,000 dog owners; these studies also reported no significant effects of dog ownership on all-cause mortality.9-11
This Swedish study is by far the largest and most thorough attempt to find an association to date between dog ownership and cardiovascular disease. Because of both its size and the greater accuracy of follow-up via government registries, this study makes it possible to give less credence to earlier failed attempts to demonstrate benefits of dog ownership. Health registries allowed assessment of both cardiovascular associated and overall mortality. The large size of this Swedish study cohort also allowed subgroup analysis to reach significance; for example, analyzing data by type of household and, interestingly, by dog breed. The varying outcome by dog breed is intriguing and may explain prior failures to discern benefit when dog types were not taken into account. If some types of dogs actually increase risk, while others lower risk, it may be that studies that analyze effects of ownership of any dog, regardless of breed, find no benefit.
While this paper does not demonstrate causation it is still hard to ignore the difference in HR between dog types. Based on these results, one might be compelled to discourage ownership of the dogs associated with greater risk in favor of those associated with lower risk of death.
The effects on disease risk appear to vary by dog breed. For example, owning a retriever was associated with a 10% decrease in CVD [adj HR 0.90 (0.87–0.94)] and a 26% decrease in all-cause mortality [0.74 (0.71–0.77)]. On the other hand owning a mixed pedigree dog was associated with a 13% increase in risk for CVD [1.13 (1.09–1.17)] and only a 2% decrease in overall mortality [0.98 (0.94–1.01)]. On average, though, dog ownership was associated with a 23% drop in CVD mortality [0.77 (0.73–0.80)] and a 20% drop in all-cause mortality [0.80 (0.79–0.82)]
These tables provide an interesting picture of the data:
- Hazard ratios (HR) and 95% confidence intervals (CI) for associations between dog breed groups and CVD outcomes in the National cohort using Cox proportional hazards regression with attained age as time-scale
- Hazard ratios (HR) and confidence intervals (CI) examining associations between dog ownership and CVD outcomes in the National Cohort (n=3,432,153) using Cox proportional hazards regression with attained age as time-scale