Pet Ownership and Quality of Life: A Systematic Review of the Literature

Pet ownership is the most common form of human–animal interaction, and anecdotally, pet ownership can lead to improved physical and mental health for owners. However, scant research is available validating these claims. This study aimed to review the recent peer reviewed literature to better describe the body of knowledge surrounding the relationship between pet ownership and mental health. A literature search was conducted in May 2020 using two databases to identify articles that met inclusion/exclusion criteria. After title review, abstract review, and then full article review, 54 articles were included in the final analysis. Of the 54 studies, 18 were conducted in the general population, 15 were conducted in an older adult population, eight were conducted in children and adolescents, nine focused on people with chronic disease, and four examined a specific unique population. Forty-one of the studies were cross-sectional, 11 were prospective longitudinal cohorts, and two were other study designs. For each of the articles, the impact of pet ownership on the mental health of owners was divided into four categories: positive impact (n = 17), mixed impact (n = 19), no impact (n = 13), and negative impact (n = 5). Among the reviewed articles, there was much variation in population studied and study design, and these differences make direct comparison challenging. However, when focusing on the impact of pet ownership on mental health, the results were variable and not wholly supportive of the benefit of pets on mental health. Future research should use more consistent methods across broader populations and the development of a pet-ownership survey module for use in broad, population surveys would afford a better description of the true relationship of pet ownership and mental health.


Introduction
Throughout history, animals have played a significant role in society including in agriculture and pet ownership. A recent survey conducted in the United States estimated that approximately 67% of homes had at least one pet, equaling about 63 million homes with at least one dog and 42 million homes with at least one cat [1]. Pets can constitute a connection to nature, function in recreational and work activities, and provide companionship in our homes [2][3][4]. The importance of animals in our lives is founded on the human-animal bond concept, which is the "mutually beneficial and dynamic relationship that exists between people and other animals that is influenced by behaviors that are essential to the health and well-being of both" [5]. This concept has championed animals as companions and family members, leading to their essential part of everyday life for many. The human-animal bond has additionally driven the common belief that pets are good for human health, both physical and mental [6][7][8].
While there are some qualitative [9,10] studies that claim that pet ownership benefits people, particularly in regard to improved mental health, there are few studies with substantial evidence from large, diverse population samples to support this theory. The studies that have been published are often not substantiated with regard to study populations or methods, making broad conclusions difficult. Furthermore, some studies that have investigated the correlation between pet ownership and mental health have revealed no effect, or even worse, negative effects of pet ownership [11][12][13][14][15]. The inconsistencies in the literature and limitations of these studies warrant a thorough exploration of the effect of pet ownership on mental health outcomes among large, diverse population samples.
Two previous systematic reviews of the literature did examine the relationship between pet ownership and mental health/well-being [16,17]. Islam and Towel [16] did not find a clear relationship between pet ownership and well-being in the 11 studies included in their review. Similarly, Brooks et al. [17] examined the role of pets in owners with diagnosed mental health problems and found mixed results across the 17 studies included in the review. The purpose of this study was to perform a systematic review of the peerreviewed published literature containing original research that examined the relationship between pet ownership and mental health for people in any population. Previous reviews included a smaller sample of research articles, often limited to a specific population of pet owners. By describing the relationship between pet ownership and mental health across all examined populations, this study will better inform whether pets could be recommended to help with mental health and whether promotion of the human-animal bond is generally beneficial.

Materials and Methods
The systematic review process involved a literature search, screening, extraction, and an assessment of the remaining articles by four researchers and three graduate students. For the purpose of this study, pet ownership was limited to dogs and cats. Our research team sought to answer, "How does ownership of a dog or cat influence the mental health or quality of life of pet owners?" In May of 2020, the following databases were searched for peer-reviewed articles on pet ownership and mental health: PubMed and Web of Science. Utilizing Boolean search terms, the literature search was conducted using the terms: anxiety OR depressi* OR bipolar OR (mental* AND (health OR disease* OR disorder* OR condition* OR ill*) for the problem, (dog OR dogs OR cat OR cats OR canine* OR feline*) AND ((pet OR pets)) AND (owner* OR companion* OR interact* OR bond* OR "human animal bond" OR "animal human bond" OR "animal assisted") for the intervention and health* AND (impact* OR outcome* OR status OR effect* OR affect* OR consequen* OR result*) for the outcome.
Although there was not an approved PRISMA protocol, the research team used Covidence (Melbourne, Australia), a software program that tracks the systematic review screening process. Identified articles were imported into Covidence, duplicates were removed, and the remaining articles were screened by the research team. Through random assignment, each article was independently reviewed by one faculty member and one graduate student. Each reviewer indicated in Covidence if the article should be included or excluded according to established criteria (Table 1). When there was a conflict between reviewers, a third reviewer (non-student) resolved the conflict. The full review process is shown in Figure 1. At the final review stage, two researchers independently extracted specific information (Table 2) from each article. The type of impact on mental health was determined based on the results reported in each article.   In addition to extracting the information outlined in Table 2, an index (Appendix A) was created to assess article quality. The index was based on two previous systematic reviews of mental health in veterinary science [17,18]. Each dichotomous index question assigned a 0 if the article did not meet criteria and a 1 if the article did meet criteria. The higher the score an article received (0-9 points), the higher the quality of the article.
Interventionary studies involving animals or humans, and other studies that require ethical approval, must list the authority that provided approval and the corresponding ethical approval code.

Results
The article review process and number of articles in each step are shown in Figure 1. A total of 54 articles met the inclusion and exclusion criteria (Table 1) and were systematically extracted (Table 2). These articles were then divided into four categories based on the type of overall impact pets had on the mental health of owners: (1) positive impact (n = 17); (2) mixed impact (n = 19); (3) no impact (n = 13); and (4) negative impact (n = 5). Factors that influenced mental health include (a) age (middle-aged female caregivers had more psychological stress than young female and male caregivers), (b) obedience and aggressiveness of the pet, (c) marital status (single women who owned a dog were less lonely and socially isolated than women without pets), and (d) attachment to the pet (high level of bonding has lower anxiety and depression scores than lower level of bonding) [19][20][21][22][23][24]. A few representative studies with mixed results include one examining the general population, which found that unmarried men who live with a pet had the most depressive symptoms and unmarried women who live with a pet had the fewest [19]. Another study examining the impact of companion animals on cancer patients found that mental health was associated with the status of cancer treatment, with those receiving intense treatment having poorer mental health [20]. In addition to overall impact, the study population, study type, population size, year of publication and article quality are reported (Appendix B).
Of the 54 articles, 19 (35%) were studies conducted in the general population, 15 (28%) were studies in older adult individuals, eight (15%) were in children and adolescents, six (11%) focused on people with some type of chronic physical illness/disease, three (6%) were studies in people with severe mental illness, and three (6%) studies examined unique populations. Of the 15 studies that had only older adult participants, none of them reported a positive impact. Seven of the articles reported mixed impact based on type of pet, gender, companionship, or another demographic. Six of the studies had no impact and two had a negative impact. Of the eight studies that involved children and adolescents, six of them indicated a clear positive impact, one indicated mixed impact, and one indicated no impact. Of the three studies that involved those with severe mental illness, two indicated clear positive impact and one indicated mixed impact.
Research studies either compared mental health outcomes in pet owners versus nonpet owners (n = 41) or with regard to owner attachment to the pet (n = 13). Similar to the overall distribution, the outcomes within these two different types of studies were distributed across all four categories (Tables 3 and 4). In 38% (five of 13) of the studies, attachment to a cat or dog was associated with a positive impact on mental health in 38% of the studies. Four of the 13 studies (31%) indicated mixed results, meaning that humananimal attachment sometimes was associated with better mental health and sometimes it was not. One example of higher attachment leading to worse mental health was for those amid cancer treatment [20]. There was no clear trend towards attachment and better mental health.  The study types included 41 (76%) cross-sectional studies, 11 (20%) prospective cohort longitudinal studies, and two (4%) other study designs. Of the cross-sectional studies, 27 (66%) found that companion animals had no or negative impact on mental health and 14 (34%) found mixed or positive impact on mental health. Of the 11 articles that reported on a longitudinal study design, five (45%) demonstrated no or negative impact and six (55%) demonstrated mixed or positive impact. Among the 54 studies, sample size ranged from 30 to 68,362.
To measure mental health constructs, 75 different validated scales were used (Table 5). Eight scales were used to measure human attachment to pets. The most common scales used across studies were the CES-D (13 studies) to measure depression and the ULS (10 studies) to measure loneliness. Two scales were used by four studies each (DASS and any variation of GHQ). Three scales were used by three studies each (GDS, CABS, and any variation of PHQ). The remaining scales were used only once or twice across the studies assessed.
Regarding the study quality scores (Appendix A), no articles received a quality score of 9, six (11%) received a score of 8, 11 (20%) received a score of 7, 20 (37%) received a score of 6, and 17 (31%) received a score of 5 or below. Of the articles with a quality scale score of 5 or lower, 18% (3) articles had no or negative impact and 82% (n = 14) had mixed or positive impact on owner mental health. Articles with a quality scale score of 6 or higher, 43% (n = 16) showed no or negative impact and 57% (n = 21) showed mixed or positive impact.

Discussion
Understanding the nature of the relationship between mental health and pet ownership is important for both human and animal welfare and to better determine the impact of human-animal interactions. Over the years, the perspective that "pets are good for you" has become an assumption [25] and when negative implications are recognized it often relates to zoonotic diseases rather than human-animal interactions [26]. This belief in the positive aspects of the human-animal bond is strengthened by marketing tools used by the pet industry [27]. While there certainly is evidence that supports the benefits of the human-animal bond to people's mental health [28,29], there is also clear and consistent evidence that the relationship is complex and sometimes negative [30,31]. The question of whether pets should be prescribed by health professionals is an especially important one. Recent qualitative research supports that attending to a pet can help a person manage mental health crises [32], however, doing so can also cause a person to rely on the pet instead of other evidenced based methods of seeking mental health support. The recommendation of obtaining a pet in the presence of mental illness ought to be coupled with other evidenced based strategies for mental health recovery such as increasing social support and engaging in third wave behaviorally based interventions such as Acceptance and Commitment Therapy or Dialectical Behavior Therapy.
The broad perspectives that pets are good for mental health may cause people to place false expectations on the role a dog or cat must play in their lives [33]. The anthropomorphism of pets (people placing human cognitive motivations on pets' behavior and treating pets as people) can in fact have a negative impact on the animal's welfare [34]. The untreated stress of people who turn to their pets instead of their human social supports and health professionals may in fact be causing pets to be more stressed [35]. Although initial data suggest relinquishment rates were not higher after COVID-19 lockdowns were lifted [36], some still have concerns that the recent increase in pet adoptions from shelters may result in pet relinquishment once the pandemic is more managed and people return to their daily work environments [37] (J. Schumacher personal communication, 5 May 2021). Developing clear guidelines about the benefits and liabilities of pet ownership and mental health is important to mitigate the public halo effect that suggests that simply acquiring a pet will improve your mental health.
Previous systematic reviews of the literature have found mixed results regarding the relationship between mental health and pet ownership [16,17]. Our search and review methodology was similar to Islam and Towel [16], which yielded 11 studies compared to the 54 studies compiled in this review. Although the Brooks et al. [17] review yielded 17 studies, they limited their search to studies only including people diagnosed with mental health conditions. While the current study did examine a larger body of research that covered broader populations and more recent publications than previous reviews, the findings were similar in that results varied across outcomes including positive, negative, mixed, and negligible. Unlike previous studies, this review also differentiated studies that compared pet owners to non-pet owners and studies that examined the level of attachment with a pet as a predictor of the mental health of the owner. Islam and Towel [16] argued that the definition of pet ownership needs to be defined across all studies, including aspects of length of ownership, time spent with the animal, and perceived quality of the interaction. Within these two categories of study types, the outcomes still varied and showed no consistent evidence that pet ownership is a positive contributor to mental health. The lack of consensus from these studies was not surprising. While popular literature and media consistently highlight the positive, it rarely highlights the negative aspects of pet ownership. In fact, studies with negative or non-significant findings are often subject to the "file drawer" effect, in which authors ultimately decide not to publish their studies [15]. In this review, we did find and include studies that reported negative or mixed findings.
The authors made the decision a priori to divide the results into categories based on the type of impact each study had on mental health. Among the 17 studies that were determined to have positive results, most of the studies were with children and adolescents (n = 6) and the general adult population (n = 6). There were some challenges to identifying these studies as clearly positive. Because a variety of different variables and a variety of different methodologies were used based on the specific purpose of each study, they could not be directly or easily compared to one another. Many of the positive impact studies investigated additional variables that could be better predictors of positive mental health than dog/cat ownership. For example, several studies indicated that children or adolescents with a dog had less depression and/or less anxiety than peers without a dog. However, family dynamics such as single parent or two parent households, time parents spend at work, presence of siblings, and family dysfunction [2,8] may be more significant contributors to child mental health than dog ownership.
The 19 mixed impact studies were easier to categorize because of conflicting outcomes, particularly for studies with an older adult or general adult population. In each of these studies, the direction of the outcome was influenced by demographic variables (such as gender) or the type of pet (cat or dog). For example, one general population study determined that women with pets had lower levels of depression whereas men with pets had higher levels of depression [19]. Another example is that pet-owning individuals with severe mental illness had less psychiatric hospitalizations than non-pet owning peers, however, they also had higher levels of substance use [38]. Another reason why a study would be categorized as mixed impact is if mental health was assessed using multiple instruments and yielded conflicting results. For instance, one study indicated that when compared to people without pets, those with pets had no difference in anxiety or stress scores yet had higher depression scores [22].
For the 13 studies that had no impact, most were with the older adult (n = 6) and general adult (n = 4) population. These studies concluded that when comparing pet ownership to non-pet ownership or when comparing attachment levels, the pet had no correlation with positive or negative mental health. Many of these studies controlled for demographic variables such as age, gender, and socioeconomic status in their statistical models. One challenge to categorizing the studies was that study participants subjectively believed their pets were helpful to their mental health despite what validated measures showed. The inclusion of these biased observations in an attempt to still put a positive spin on the study may reflect the conflict a researcher has in publishing negative results. An additional challenge is that studies that included non-mental health measures (such as physical health) showed that those with pets did better than those without. Expert reviews of pet ownership on cardiovascular health have demonstrated a significant challenge to reach a definitive conclusion of the impact of pet ownership on health based on the current evidence [39].
Five studies demonstrated a clear negative impact between pet ownership and mental health. The sample populations were general (n = 2), older adults (n = 2), and single adults living alone (n = 1). In these studies, pet ownership was associated with higher levels of depression, loneliness, and other psychological symptoms across all demographic variables and type of pet (dog or cat). Again, the challenge to classifying these studies as negative impact suggests that pet ownership causes increased levels of mental health illnesses, when in reality, the studies are about correlation, not causation. There may be other factors that cause the samples in these studies to have worse mental health. As indicated by Mullersdorf et al. [40], the presence of a psychological condition could predispose individuals to become pet owners, making it difficult to truly know if pet ownership causes a negative impact on mental health. These studies, regardless of type of outcome, only indicate association of pet ownership and mental health.
Another challenge in comparing the 54 studies was the difference in methodology and quality of each study. Due to this, our methods did not evaluate the individual and overall power and effect sizes of study results. Quantitative methodologies are warranted in this field, particularly prospective, randomized, double-blind, placebo-controlled intervention trials that are longitudinal in design to provide evidence of the impact of animal ownership over time while eliminating as many extraneous and confounding variables as possible [41]. Ideally, this truly experimental model of pet ownership would include random assignment of companion animals in a closed system to eliminate as many sources of error variance as possible [42]. However, due to the nature of pet ownership being integrated as a part of daily life on a voluntary basis, this experimental model would be difficult to achieve. Perhaps the most compelling of all studies that comes closest to this design was a prospective interventional study in which 71 previous non-pet owners were given a cat or dog; results demonstrated mild benefits in mental health and behavior after 10 months of pet ownership compared to the 26 non-pet owners [43]. While noteworthy, there was lack of randomization, so the pet ownership group consisted of a relatively small number of subjects who were searching for a pet to adopt rather than receiving it on random chance. Regardless, this study still reports an improvement in mental health in this specific population. Future studies should strive to achieve this prospective, controlled, experimental methodology to more compellingly connect pet ownership with mental health.
A quality index attempted to rate the rigor of each study, but the index was subjective and based on questions that could be asked without statistical analysis (e.g., does this study include a comparison population?). The higher the score on the quality index, the more likely the study was scientifically rigorous. The lower the score, the more likely the study was to demonstrate a positive or mixed impact on the pet owner's mental health. While both previous literature reviews critiqued the rigor of the studies reviewed and remarked upon the consistent methodological flaws, Islam and Towel did not assign objective scores to the 11 studies reviewed. Brooks et al. [17] did assign quality scores to each of the 17 studies reviewed but did not evaluate the impact of the quality of the study on its results. The quality scores in the current review varied across all four outcome categories and did not give any indication of quality impacting the overall outcome. Still, it is important that researchers strive for higher quality research that carries more weight in the question of whether pet ownership truly impacts mental health. Additionally, we recommend that studies be replicated in an attempt to corroborate previous findings, which contribute to the overall understanding of the phenomenon.
Lastly, this study also examined how mental health was evaluated across the studies. For the 54 studies included in this review, 75 different scales (Table 5) were used with many research studies implementing more than one scale (Appendix B). While most of the scales used have been previously validated, the inconsistent use of scales makes comparison of results across studies challenging. While it is common to utilize an instrument that is a validated self-report of depression, it is likely that researchers often utilize other scales because they are investigating other aspects of mental health such as loneliness, stress, and anxiety. Many scales also rely on self-reporting of mental health indicators, which can be affected by inherent bias, especially when completing a survey regarding mental health and pet ownership. To allow for better comparison of future studies, researchers should attempt to use consistent measures of mental health across studies, such as the CES-D [44], which was the most commonly used scale in 13 of the 54 examined studies.
In addition to consistent use of mental health scales across studies, the development of a module for use in wide-scale population surveys with a focus on pet-ownership would benefit future research examining the relationship between pet ownership and health. The Behavioral Risk Factor Surveillance System (BRFSS) [45] is an annual questionnaire administered by the US Centers for Disease Control and Prevention. There are 14 core sections that are administered to all participants and 31 optional modules [45]. None of these modules focuses on pet ownership and the addition of such a module would allow for a more in-depth evaluation of the relationship between pet ownership and health, both mental and physical, across large populations. While pets can play a significant role in the owner's health, it can be difficult to differentiate the effects of pet ownership from the many other factors that contribute to one's mental and physical health. The addition of a pet ownership module to the BRFSS would allow researchers to examine the role of pet ownership in tandem with other factors that contribute to health. On a smaller scale (approximately 3000 participants), the General Social Survey (GSS) is a representative survey that monitors trends in opinions, behaviors, and demographics among Americans [46]. Though not a main focus, the GSS does include pet ownership and mental health variables. Including pet ownership allows researchers who study the relationship of ownership with humans to have a large, representative dataset to analyze correlations. For example, a recent study used the GSS 2018 to examine demographics of pet ownership [46]. In their conclusion, the authors of this study indicated that the strengths of using the GSS to study pet ownership characteristics are high quality data, multiple covariates, sound methodology, and easy access [47]. Including pet ownership questions in multi-wave, representative studies would further the work of human animal relationship research.
This systematic review was limited due to only searching two databases and only evaluating research published in English. The majority of studies focused on pet-owners in Western cultures. The human-animal bond may differ across cultures and future studies should include pet-owners in non-Western cultures. However, a large number of articles were identified, and the total number of articles included in final extraction was greater than similar previous systematic reviews. More consistent methods across research that evaluates the relationship between pet ownership and mental health might allow for more extensive comparison of studies.

Conclusions
Previous research examining the impact of pet ownership on mental health has shown mixed results and the results of this study were the same. While there were more absolute numbers of studies to demonstrate a positive impact (n = 17) compared to negative impact (n = 5) on mental health, the overall results indicate a much more complicated picture. While 17 of the 54 studies had a clear association of pet ownership and positive mental health, the remaining 37 articles show a mixed association, no association, or a negative association. Comparing these studies is quite challenging due to the number of measures used to assess mental health, the differences in study quality, and the variety of variables that were controlled for. While research studies can be improved by addressing limitations as described, a more comprehensive evaluation of behavior and its association with health outcomes is warranted. We also cannot ignore that mental health is multifactorial. Pet ownership and the resulting human-animal interaction is a single factor; other factors that also contribute to mental health should be examined in large populations of pet-owners and non-pet-owners. The addition of a pet-ownership specific module to the BRFSS, as previously described, would allow for prospective research that can be replicated, and eventually retrospective research, that will also allow for inclusion of other factors that contribute to health.

Conflicts of Interest:
The authors declare no conflict of interest.

Appendix A
Article Quality Index 1 Was the study purpose reported? 1 = yes, 0 = no 2 Did the study design include a comparison group? 1 = yes, 0 = no 3 Was the recruitment method reported? 1 = yes, 0 = no 4 Was the sample size response rate over 50%? 1 = yes, 0 = no (or unreported) 5 Were sample demographics reported? (3 or more demographic categories reported)1 = yes, 0 = no 6 Was the sample representative (not self-selected)? 1 = yes, 0 = no (or not reported) 7 Did mental health diagnosis occur through standardized scale or mental health professionals? 1 = yes, 0 = no 8 Was the validity and/or reliability of scales reported? 1 = yes, 0 = no 9 Was there a study limitation section? 1 = yes, 0 = no

Appendix B
Following a literature review and data extraction of research articles that examined the relationship between pet ownership and mental health, the following articles were found to meet inclusion and exclusion criteria as outlined in Table 1. There was no relationship between companion animal ownership and stress or anxiety, however, owners had higher levels of depression than non-owners. Depression among those who perceived their animal as more friendly was lower and for those who perceived their animal as more disobedient stress was higher. Participants who were exposed to pet aggression in childhood and reported medium level bonds with animals also reported more depression and anxiety symptoms in early adulthood. Those who were not exposed to pet aggression reported fewer internalizing symptoms. There was no indication for an association between dog ownership and mental health in groups with high or low access to natural outdoor environment (NOE) and with high or low residential surrounding greenness on the whole.