This review followed established guidelines (15) and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for scoping reviews (PRISMA-ScR) (16). We consulted with an expert librarian from our university in May 2022 for feedback on search terms, truncation, Boolean phrasing, and databases best suited for this review. The final search terms are included in Supplementary Table S1a. Six databases were searched in June 2022 using EBSCOhost (APA PsycInfo, PubMed [includes MEDLINE], Sociological Abstracts, SocINDEX with Full Text, Social Science Full Text, and Psychology and Behavioral Sciences Collection) as well as one publicly available platform (HABRI Central). Google Scholar and Web of Science Collection searches were attempted but inaccessible due to imposed character limits. The reference lists of assessed articles were also screened for inclusion. Our PRISMA-ScR Flow diagram is provided in Figure 1.

Articles were required to provide a frailty definition and relevant instrumentation or otherwise consider it an outcome. Those that only addressed human-animal observations (i.e., aviaries, fish tanks) were excluded due to the notable emphasis on physical conceptualizations and measurements of frailty in the literature (17). All forms of academic manuscripts published in English were considered. Participants must be 50+ but could have reported any form of housing or living arrangement (e.g., multigenerational; retirement community residents), any health status (i.e., healthy, unhealthy) or any type of physical, cognitive, or psychiatric diagnosis or comorbidity (e.g., cognitive impairment; mental illnesses).

After downloading full texts to Zotero for screening, core data was extracted (e.g., frailty instrumentation; animal species) and further assessed for eligibility. From the articles included for review via database searches, the first author reviewed each reference list and screened titles and abstracts to identify other relevant articles. Reported key findings and implications were charted and thereupon analyzed for inductive thematic analysis.

Overall, 3,407 sources were retrieved. After 289 duplicates were automatically removed and titles and abstracts of the remaining 3,118 unique articles were assessed, the first two authors independently read through 34 full texts to determine eligibility. Thereafter, 43 articles identified via reference reviews were screened and three were assessed for eligibility. Any inclusion uncertainty was resolved through discussion among the first two authors. A total of four articles are included: two were retrieved from PubMed (includes MED-LINE) (4, 14), one from HABRI (Human Animal Bond Research Institute) Central (18), and one from reference reviews (19).

The following results are organized by article and in chronological order.

Survey responses of Ota Genki Senior Project participants (N = 11,233) were analyzed regarding their experiences of “dog/cat ownership” to examine physical, psychological, and social functioning among current, past, and never pet owners (19). While no significant differences in frailty (i.e., score > 4) were found among current, past, and never owners through this scoring mechanism, compared to never owners, past and current ownership was associated with benefits in physical, social, and psychological functioning domains (19). This work suggests that never owning a dog/cat is associated with lower levels of walking, lesser degrees of light physical activity, and poorer social functioning (i.e., feelings of trust in neighbors) (p < 0.001) (19). Finally, certain health characteristics were reported as associated with either current or past dog ownership [Motor Fitness Scale (p < 0.001), interaction with neighbors (p < 0.001), trust in neighbors (p < 0.001), social isolation (p < 0.001)] and cat ownership [interaction with neighbors (p < 0.001), trust in neighbors (p < 0.005), social isolation (p < 0.002)] (19).

The study concerned with incident frailty (N = 6,197) found greater reduced risks among past dog/cat owners (OR = 0.85, CI: 0.71–0.99, p < 0.05) than current (OR = 0.90, CI: 0.72–1.13) or never (OR = 1) owners (18). When separating dog from cat owners, past dog owners were significantly less likely to experience incident frailty (OR = 0.84, CI: 0.70–0.99, p < 0.05) than never (OR = 1) or current (OR = 0.86, CI: 0.65–1.13) dog owners. No statistically significant associations were found between incident frailty and cat ownership, but past cat owners (OR = 0.89, CI: 0.70–1.12) experienced slightly lower incident frailty risks than never (OR = 1) or current (OR = 1.04, CI: 0.77–1.40) owners. Compared to those who never owned a dog/cat, social function (i.e., interaction with neighbors) had a stronger negative association with incident frailty than those who currently or previously owned either pet species. They also found current dog/cat owners to be significantly younger than past and never owners and more often married [p < 0.001] with higher education [p < 0.001] and income rates [p < 0.001], obtained higher Motor Fitness Scale scores [p < 0.001] and frequencies of going outside [p < 0.001], and interacted with neighbors more often [p < 0.003] than past or never owners (18).

While the systematic review reported no statistically significant differences in frailty percentages among never, past, and current dog or cat owners, calculated odds ratios across the three included studies suggest that cat and/or dog ownership may be associated with benefits in the physical, psychological, and social health domains, which constitutes frailty (14). Additionally, an article concerned with psychological frailty (i.e., depressed mood and phenotypic frailty) (21) found the risks to be 40% lower for participants rearing grandchildren or pets (OR = 0.60, 95% CI: 0.47–0.76, p < 0.001) than those who did not (14, p. 4).

Finally, significant differences in cognitive frailty risk were uncovered among those (N = 2,638) who experienced social loss and did or did not own pets (4). Regardless of gender, experiencing either form of social loss had higher cognitive frailty rates than those who did not, but women with social loss and did not own a pet experienced the highest risk (15.5%) (OR = 2.06, 95% CI: 1.20–3.54, p < 0.01), followed by men with social loss who did own a pet (13.5%) (OR = 4.20, CI: 1.38–12.77, p < 0.05) (4). Conversely, rates were lowest among those who did not experience social loss and did own a pet [women: OR = 1.63 (CI: 0.99–2.71); men: OR = 1.50 (CI: 0.72–3.11)] (4). Overall, the prevalence of cognitive frailty was higher for women (9.3%) than for men (6.1%), suggesting sex differences may be at play and warrant consideration. Those deemed cognitively frail were older [p < 0.001], had lower education [p < 0.001] and income [p < 0.001] rates, higher psychological distress scores [p < 0.001], did not drink alcohol [p < 0.01], and did not own a pet [p < 0.01] (4).

Physical, psychological, cognitive, and social health benefits can result from PO (4, 18, 25, 26) and yet, how these health domains affect one another are rarely assessed through the interdisciplinary perspectives of gerontology and anthrozoology. Each included article spoke to benefits in the physical, psychological, and social domains, but only one spoke to the influence of PO on cognitive health. Understanding the directionality behind the interdependent effects of these health domains is bound to result in more appropriate instrumentation and interventions regarding this incipient topic (27). One article spoke to social frailty presaging physical frailty (18). Another argued that dog ownership is associated with greater degrees of walking and social functioning, suggesting physical activity can promote social interaction (19). Relatedly, poor social health can result in social isolation, consequently impeding psychological health. The conceptualizations of cognitive frailty (4) and psychological frailty (14) build this argument further.

One article examining frailty differences among men and women spoke to the possibility of social roles, like gender norms, being influential to differences in cognitive frailty rates (4), further highlighting the interconnectedness of psychological, social, cognitive, and physical health. Cultural differences, including but not limited to social norms, warrant increased attention to better understand how these concepts might influence individual health domains and, consequently, frailty risk.

Evidence of PO as a protective mechanism against frailty was found, with dog ownership possibly having more influence than cat ownership. This may be because current and past dog owners reported greater degrees of physical activity via walking compared to those who never owned a dog (19). Or this finding could be due to the phenomenon known as the “pet ownership effect,” which argues that physically active individuals are more likely to own a pet than those who are not (18). Despite these causality concerns, this work argues that older adults who never owned a pet are consistently more likely to be frail than those who currently or previously owned a pet (18, 19). The varied effectiveness of past and current dog/cat ownership on frailty risk should be further explored.

This work supports the notion that PO can result in increased meaning and purpose, which can positively influence psychological health (26, 28). Experiencing a sense of meaning and purpose in life is similarly associated with lower risks of ill-being (29). This is especially true for older adults with low socioeconomic status or weak social networks (28). The included review (14) revealed psychological frailty as significantly less prevalent in older individuals who rear either grandchildren or pets (21). Feeling intrinsically valued and purposeful through PO can afford “social situatedness” to older adults even in light of fluctuating autonomy (28). Interventions to increase meaning and purpose in older individuals should be developed and assessed for appropriateness and effectiveness.

Meta-analysis was unattainable due to the inconsistent collection of frailty-based variables. Although we aimed to synthesize the largest body of literature possible, only incorporating articles published in English likely limited the number of hits, particularly considering the limited geographical diversity identified in this review. The generalizability of these findings is further tethered by all articles pertaining to older individuals in East Asia, predominantly Japan, potentially limiting the applicability of these findings to other older adult populations.

Varied findings across the included articles necessitate investigation into the heterogeneity of effects that PO may have on frailty for 50+ adults. Further, whether said associations are direct or indirect, and the specific mechanisms by which older adults benefit from HAIs regarding frailty risk remain unclear. Improved development and assessment of relevant studies and interventions may determine whether (and which) HAIs are successful in averting frailty in older adults, what health domains are affected when, and the directionality between any interconnected health effects.

Future investigation should consider longitudinal analysis, companion animals outside of dogs and cats, and HAI studies and interventions outside of PO. Other studies and interventions that could glean meaning and purpose for older adults should be explored, like interacting with visiting animals or caring for houseplants.

More research is needed on PO in various housing environments. While most prefer to age in place (28), retirement community enrollment will likely increase as global societies continue aging. Related housing policies that prevent PO for older adults—in and outside of retirement communities—require review to ensure these regulations foster autonomy, avoid ageist and ableist language and intentions, and to eliminate the possibility of one choosing between housing or their pet (28). When older adults are better supported at the individual, communal, and policy levels to age in place the beneficial effects of PO are likely to ensue, including increased degrees of well-being and QOL (26).

Social determinants of health, like one’s built environment, are known to significantly impact QOL (30). These multifaceted health determinants likely contribute to frailty risk over time but are not yet adequately integrated into frailty assessments. Future research should also integrate the environmental health domain when assessing and defining multidimensional frailty.