Cardiovascular disease is the leading cause of death in the United States.1 Results of research concerning the effect of pet ownership on cardiovascular health have been inconsistent.2–4 A clinical trial of 48 hypertensive stockbrokers from New York found that pet ownership was more effective than angiotensin-converting enzyme (ACE) inhibitors in blocking the increase of blood pressure (BP) with mental stress.2 Similarly, a clinical trial in 240 married couples found that those with pets had significantly lower BP levels, less BP reactivity, and a quicker recovery from stressors.3 Another clinical trial in 23 healthy children, aged 3-6 years, reported that BP was reduced in the presence of a dog during a doctor's office visit.5 Two other small intervention trials of pet snakes showed mixed results.6,7 Although 3 of these studies reported significant BP benefits,2,3,5 they were relatively small studies that did not include older adults, and most were of dog and cat owners only.
In contrast, a cross-sectional study of 5079 Australian men and women aged 40–64 found no evidence that pet (dog/cat/other) ownership (of any type of animal) was associated with cardiac risk factors4; in fact, participants with pets had higher diastolic BP than those without pets, even after adjusting for confounders. Editorials and letters have since debated these results.8,9 Although some argue that pet ownership benefits cardiovascular health,8 others state that there is no clear evidence that this is the case.9 There have been no other large population-based studies examining the association of pet ownership and BP in older men and women. We examined the association between pet ownership and BP before and after adjustment for confounders in a large population-based study of older men and women.
Between 1972 and 1974, 82% of all adults age 30 to 79 residing in the southern California community of Rancho Bernardo were enrolled in a study of heart disease risk factors. These 6339 individuals have been followed with periodic clinic visits and yearly mailed questionnaires; vital status is also obtained annually. Participants are white, middle class, and relatively well-educated.
Between 1992 and 1996, members of this cohort age 50 years and older were invited to attend a follow-up clinic visit that included BP measurements; 1782 participated. Of these individuals, 1322 had responded to a mailed survey in 1991–1992 on pet ownership. After eliminating 29 individuals for lack of data on pet ownership and 7 whose BP was not measured, there were 1179 (498 men and 681 women) age 50 and older who provided data for this analysis. This study was approved by the University of California, San Diego (UCSD) Human Subjects Protections Program; all participants were ambulatory, and written informed consent was obtained prior to participation.
During the 1992–1996 clinic visit, 2 BP measurements were obtained using a standard mercury sphygmomanometer, after the participant had been seated quietly for at least 5 minutes, by a specially trained nurse using the Hypertension Detection and Follow-up Program protocol.10 A 30-second pulse rate was measured at the wrist.
Weight and height were assessed with the participants wearing light clothing and no shoes; body-mass index (BMI; kg/m2) was calculated as an estimate of obesity. Percent body fat was estimated using bioelectrical impedance (model 1990B, Valhalla Scientific Inc., San Diego, CA). Waist and hip girths were measured, and waist-hip ratio was calculated ([waist (cm)/hip (cm)] × 100). A blood sample was obtained in the morning after a requested 12- to 16-hour overnight fast; total cholesterol was measured in a Centers for Disease Control-certified Lipid Research Clinic laboratory using enzymatic techniques and an ABA-200 biochromatic analyzer (Abbott Laboratories, Irving, TX).
Participants answered standardized questions on current exercise 3 or more times per week (no/yes), the frequency of alcohol consumption 3 or more times per week (no/yes), and cigarette smoking history (never/past/current). Participants also reported whether they had ever been diagnosed by a physician as having diabetes or hypertension. Current medication use (including use of betablockers, thiazides, and other antihypertensives) was queried and validated by a nurse who examined pills and prescriptions brought to the clinic for that purpose.
On the mailed questionnaire in 1991–1992, participants were asked about lifetime and current pet ownership, including the kinds of indoor pets they had ever owned (dogs, cats, birds, hamsters/gerbils, or others). Pet ownership was defined as current ownership of a pet. Lifetime pet ownership was defined by ever owning an indoor pet.
Means of the 2 systolic BP measurements (SBP) and the 2 diastolic BP (DBP) measurements were used to analyze BP as a continuous variable. In categorical analyses, individuals were categorized as hypertensive if their mean SBP was ≥140 mm Hg, their mean DBP was ≥90 mm Hg, or they reported current use of antihypertensive medication. Because of changes in recent guidelines for hypertension, a second analysis was performed in which individuals were categorized as hypertensive if they had mean SBP ≥ 130 mm Hg or mean DBP ≥ 85 mm Hg, or they reported current use of antihypertensive medication. Pulse pressure was calculated as SBP minus DBP. Mean arterial pressure was calculated as (SBP + DBP)/2.
Participants were dichotomized on pet ownership based on whether they currently owned a pet. Because it was thought that current pet ownership would have a greater effect on current BP than lifetime pet ownership, current ownership was used in most analyses. However, analyses were also performed after categorizing participants based on lifetime pet ownership and type of pet owned. Participants were categorized as lifetime dog owners if they owned only a dog or if they owned a dog and any other pet; as lifetime cat owners if they owned only a cat or a cat and any pet other than a dog; and as lifetime other pet owners if they owned only pets other than a dog or cat.
Comparisons of descriptive statistics by pet ownership controlling for age were performed using analysis of covariance (ANCOVA). Multiple linear regression analysis was used to examine the associations of pet ownership with SBP, DBP, pulse pressure and mean arterial pressure after adjustment for potential confounders. Multiple logistic regression analysis was used to determine whether pet ownership was associated with risk of categorically defined hypertension after adjustment for potential confounders. Analyses were also performed stratified by age, sex, pet type (dog owners/other pet owners/nonowners) and time when pets were owned (current/past/never). Multiple logistic regression analysis was used to determine survival bias by examining associations between pet ownership and vital status adjusted for potential confounders. Statistical analyses were performed with SAS statistical software (version 8.1, SAS Institute, Cary, NC).
Of these 1179 participants, 42% (n = 498) were men and 57% (n = 681) were women. Age ranged from 50.2 to 95.4 years with mean ± SD of 70.4 ± 10.8 years. Overall, 30% (n = 354) reported current pet ownership and 80% (n = 947) reported ever owning a pet. Mean (±SD) SBP was 137.5 (±21.4) mm Hg and DBP was 76.1 (±9.3) mm Hg. Using a categorical definition based on SBP ≥ 140, DBP ≥ 90, or use of antihypertensive medication, 55% were hypertensive; based on SBP ≥ 130, DBP ≥ 85, or use of antihypertensive medication, 71% were defined as hypertensive. Among the 947 participants who ever owned a pet, 80% reported ever owning a dog, 38% a cat, 20% a bird, 8% a hamster or gerbil, and 5% another type of pet.
As shown in Table 1, current pet owners were younger (mean = 64 years) than those who did not own pets (mean = 73.0 years). Age-adjusted comparisons of other characteristics by current pet ownership are also shown in Table 1. Current pet owners were somewhat more likely to be overweight (defined as BMI ≥ 25.0) than those who did not own pets (58% vs. 46% although mean BMI was similar between groups (mean = 25.4 and 25.7, respectively). Current pet owners were also less likely than nonowners to exercise 3 or more times per week (66% compared with 73%), slightly more likely to report diabetes (7% compared with 6%), and slightly more likely to take beta-blockers (23% compared with 22%). Those who currently owned pets did not differ from nonowners by sex, percent body fat, smoking status, alcohol consumption, physical health status, or other BP medication use. Results were similar in comparisons of those who were lifetime ever versus never pet owners (data not shown).
Table 2 shows unadjusted, age-adjusted, and multivariate-adjusted comparisons of mean BP by current pet ownership. Based on the unadjusted analyses, current pet owners had lower SBP than nonowners (mean = 132.8 mm Hg [130.5–135.0] vs. 139.5 mm Hg [138.1–141.0]), lower pulse pressure (mean = 55.5 mm Hg [53.6–57.5] vs. 63.9 mm Hg [62.6–65.1], lower mean arterial pressure (mean = 105.0 mm Hg [103.6–106.4] vs. 107.6 mm Hg [106.7–108.5]), but somewhat higher DBP (mean = 77.2 mm Hg [76.3–78.2] vs. 75.7 mm Hg [75.0–76.3]), compared with those who did not own pets. However, these differences were attenuated after adjustment for age, or for age, sex, BMI, antihypertensive medication use, exercise, and diabetes diagnosis. Stratification by sex, by pet type (dog owners/other pet owners/nonowners), or by time when pets were owned (current/past/never) yielded similarly minor differences in SBP, DBP, PP, and MAP (data not shown). In addition, similar results were obtained when comparisons were made between those who had ever versus never owned a pet and when age and BMI interactions were considered (data not shown).
Table 3 shows the risk of hypertension in current pet owners as compared with nonowners overall and in sex-specific models after adjustment for age, or for age and other covariates. Before adjustment for age, the risk of hypertension among all pet owners (men and women combined) was lower than for nonowners, whether using the more stringent criteria (OR = 0.62; 95% CI = 0.49–0.80) or the less stringent criteria (0.55; 0.43–0.72). However, after adjustment for age, or for age and the other covariates, there were no substantial differences by pet ownership in risk of hypertension. Similar results were obtained when stratified by sex, time when pets were owned (current/past/never) or when lifetime pet ownership (ever/never) was used instead of current pet ownership (data not shown). Similar results were obtained when BP data collected at a 1988–1991 clinic visit was used instead of data from 1992–1996.
Results of analyses examining the association of type of pet ever owned with BP, pulse pressure, and mean arterial pressure are presented in Table 4. In unadjusted analyses, participants who had ever owned a dog had a lower SBP (mean = 136.2 mm Hg [134.7–137.7] vs. 141.9 mm Hg [139.3–144.6]) and pulse pressure (mean = 59.9 [58.6–61.2] vs. 66.3 [63.9–68.6]) than those who never owned pets. There were no other pairwise differences in SBP, DBP, pulse pressure, or mean arterial pressure by type of pet ever owned (dogs vs. cats vs. other vs. none). There similarly were no important pairwise differences after adjustment for age alone or for age, sex, exercise, self-reported physical health, beta-blocker use, other antihypertensive medication use, and diabetes diagnosis.
Overall, 415 (35.2%) of this sample died between the 1991–1992 questionnaire and 2004. In an age-adjusted analysis of pet ownership from 1991–1992 with vital status through 2004 (data not shown), those who did not own pets were more likely to have died than those who did own pets (OR = 2.09; 95% CI = 1.61–2.72).
Millions of Americans own pets. Results from some studies have suggested that pets can reduce BP.2,3,5 Plausible mechanisms have been suggested, including reduced stress, companion support, and walking a pet (usually a dog) for exercise. In the present study, current pet ownership was associated with both lower BP and hypertension in unadjusted analyses, but this association was explained by age, and not modified by further adjustment for covariates. Similarly, unadjusted analyses suggested that dog ownership conferred some benefit with regard to systolic BP and pulse pressure. However, after adjustment for age, there were no important differences between dog owners and nonowners on these variables. These results show the importance of adjustment for confounders, especially age, when examining BP and related outcomes. It is of interest that some, though not all, previous studies that suggested a reduction in BP with pets ownership2,5 did not adjust for confounders.
The results of this study are in accord with those of a randomized study in 72 healthy, middle-aged Australian men and women, which showed that the presence of a dog had no effect on BP during a task inducing mental stress.11 Mental stress increased BP in the absence of a dog, but BP was not affected by the presence of a dog. A large cross-sectional study performed with more than 5000 middle-aged men and women drawn from the electoral roll in Australia4 also has reported that pet ownership did not improve BP. In that study, current pet owners had higher DBP than those who did not currently own pets after adjustment for age, BMI, alcohol consumption, smoking status, physical activity, diabetes diagnosis, education, and socioeconomic status.
The results of the present study are in contrast to those of a number of small studies.2,3,5 For example, Allen and colleagues2 found that pet ownership reduced BP increases in response to mental stress in 48 hypertensive stockbrokers, but only persons who agreed to acquire a pet were included in the study, possibly biasing the results. The same research group found that among 240 healthy, middle-aged married couples, those who owned pets had lower BP levels, greater reactivity, and a quicker recovery from stressors.3 These results persisted even after adjustment for age, sex, BMI, education, income, presence of children in the home, mental health, or interpersonal support, although the data were not adjusted for physical activity.
Several sources of confounding in the present study were considered. Obesity is associated with high BP12 but is unlikely to explain the similarity of pet owners and nonowners observed here. Pet owners actually had a slightly higher BMI and exercised less than those who did not own pets, and adjusting for BMI or exercise 3 or more times per week did not change the associations. Furthermore, pet owners and those who did not own pets did not differ on either waist-hip ratio or percent body fat, and adjusting for these variables did not materially alter the results. In western cultures, BP, especially SBP, increases with age.13 The favorable associations of pets with BP in the present study were not independent of age and thus do not support the thesis that current pet ownership has an independent favorable effect on SBP or any other BP variable in older adults.
It is plausible that the type of pet owned or the owner's affinity for the pet could affect BP. For example, a dog owner might be more physically fit than a cat owner, because dog owners might walk more than cat owners. However, no differences by type of pet owned were observed in this study. Those who have a greater affinity for their pet and perceive their pets as having an affinity for them may have different brain levels of serotonin or cortisol14; but these variables were not assessed in the present study. Likewise, duration of pet ownership could have an effect but was not assessed.
Several potential limitations of this study should be considered. The outcome measures of BP, pulse pressure were all objectively assessed, thereby reducing the risk of response bias. However, we cannot exclude the possibility of survival bias, whereby those without pets had the highest BPs but died prior to the study. An age-adjusted analysis of pet ownership (1991–1992) with current vital status through 2004 showed that those who did not own pets were more likely to have died than those who did own pets. Selection bias may also have occurred; healthier participants may have been more likely to answer questionnaires and attend clinic visits. However, if this had occurred, we would have been more likely to find differences between pet and nonpet groups after adjustment for covariates. There is also the possibility that participants may have given up their pets as illness limited their ability to care for pets, but participants were not queried about reasons for not owning pets. Finally, results from this study of well-educated, middle-class men and women with good access to medical care may not generalize to other samples of men and women. Previously published papers have shown that, as compared with national, representative samples of persons their age, participants in the Rancho Bernardo cohort are on average somewhat leaner,15 but are no more or less likely to have been cigarette smokers,16 and have similar levels of alcohol consumption, SBP, diabetes, impaired glucose tolerance, and plasma total cholesterol.17–22
In conclusion, the overall results of the present cross-sectional study suggest that BP, hypertension, and vascular reactivity do not vary by pet ownership. The similarity of BP levels in those who did and did not own pets does not support the thesis that pet ownership lowers BP and risk of hypertension. The overall results of this study suggest that there is no beneficial effect of pet ownership on BP in old age.
1. National Center for Health Statistics. Health, United States, 2004. Chartbook on Trends in the Health of Americans
. Hyattsville, MD: National Center for Health Statistics; 2004:172.
2. Allen A, Shykoff B, Izzo JL, Jr. Pet ownership, but not ACE inhibitor therapy, blunts home BP responses to mental stress. Hypertension
3. Allen A, Blascovich J, Mendes WB. Cardiovascular reactivity and the presence of pets, friends, and spouses: the truth about cats and dogs. Psychosom Med
4. Parslow RA, Jorm AF. Pet ownership and risk factors for cardiovascular disease: another look. Med J Aust
5. Nagengast SL, Baun MM, Megel M, Leibowitz JM. The effects of the presence of a companion animal on physiological arousal and behavioral distress in children during a physical examination. J Pediatr Nurs
6. Eddy TJ. RM and Beaux: reductions in cardiac activity in response to a pet snake. J Nerv Ment Dis
7. Alonso Y. Cardiovascular responses to a pet snake. J Nerv Ment Dis
8. Headey B. Pet ownership: good for health (Editorial)? Med J Aust
9. Nair BR, Flynn B. Pet owners and risk factors in cardiovascular disease (Letter). Med J Aust
10. Curb JD, Ford C, Hawkins CM, Smith EO, Zimbaldi N, Carter B, Cooper C. A coordinating center in a clinical trial: the Hypertension Detection and Followup Program. Control Clin Trials
11. Kingwell BA, Lomdahl A, Anderson WP. Presence of a pet dog and human cardiovascular responses to mild mental stress. Clin Auton Res
12. Grundy SM, Pasternak R, Greenland P, Smith S Jr., Fuster V. Assessment of cardiovascular risk by use of multiple-risk-factor assessment equations: a statement for healthcare professionals from the American Heart Association and the American College of Cardiology. Circulation
13. Hahn RA, Heath GW, Cheng MH. Cardiovascular disease risk factors and preventive practices among adults – United States, 1994: a behavioral risk factor atlas. Behavioral Risk Factor Surveillance System State Coordinators. MMWR CDC Surveill Summ
14. Asberg M, Thoren P, Traskman L, et al. “Serotonin depression”—a biochemical subgroup within the affective disorders? Science
15. Barrett-Connor E. The prevalence of diabetes mellitus in an adult community as determined by history or fasting hyperglycemia. Am J Epidemiol
16. Holdbrook MJ, Austin MS, Criqui MH, et al. Cigarette smoking cessation in an upper middle class population. J Cardiac Rehab
17. Jones BR, Barrett-Connor E, Criqui MH, et al. A community study of calorie and nutrient intake in drinkers and nondrinkers of alcohol. Am J Clin Nutr
18. Barrett-Connor E, Criqui MH, Klauber MR, et al. Diabetes and hypertension in a community of older adults. Am J Epidemiol
19. Harris MI, Hadden WC, Knowler WC, et al. Prevalence of diabetes and impaired glucose tolerance and plasma glucose levels in U.S. population aged 20–74 years. Diabetes
20. Harris MI. Impaired glucose tolerance in the U.S. population. Diabetes Care
21. Wingard DL, Sinsheimer P, Barrett-Connor EL, et al. Community-based study of prevalence of NIDDM in older adults. Diabetes Care
. 1990;13(Suppl 2):3–8.
22. National Institutes of Health. The Lipid Research Clinics Population Studies Data Book. Vol 1. The Prevalence Study
. Washington, DC: US Department of Health and Human Services, GPO; July 1980. (NIH publication no. 80–1527).