Table 3.

Key findings

First author	Country	Assessment of pet ownership	Assessment of loneliness or social isolation	Study type	Sample characteristics				Results
					Sample description	Sample size	Age	Females in total sample (%)
Adult population prior to the pandemic
Antonacopoulos (2010) [50]	Canada	Pet ownership (dichotomous)	UCLA Loneliness Scale (Version 3) (20 items)	Cross-sectional	Individuals who are living alone	n = 132	M: 39.4 SD: 14.4 18–78	73.3	According to hierarchical regression, there was no significant association between pet ownership (ref.: non-pet ownership) and loneliness However, pet ownership × social support was associated with decreased levels of loneliness (ß = – 0.32, p < 0.05)
Antonacopoulos (2017) [49]	Canada	Having acquired a dog (dichotomous)	Feelings of loneliness during the last week, rated on a four-point scale UCLA Loneliness Scale (Version 3) (20 items)	Longitudinal (two waves during 8 months)	Individuals who are living in a town and do not have a dog at the baseline	n = 139	M: 36.8 SD: 14.4 18–68	64.0	Regarding the UCLA Loneliness Scale, acquiring a dog (ref.: not acquiring a dog) was not related to diverging levels of loneliness With respect to the single item, ANOVA revealed that acquiring a dog (ref.: not acquiring a dog) was associated with decreased levels of loneliness (p < 0.05)
Bennett (2015) [38]	Australia	Pet ownership (dichotomous) Dog ownership (dichotomous) Cat ownership (dichotomous)	UCLA Loneliness Scale-Revised (20 items)	Cross-sectional	Community-dwelling individuals	n = 68	M: 71.6 SD: 5.6 65–80	72.1	T tests revealed no significant differences between pet owners and non-pet owners, dog owners and non-dog owners, and cat owners and non-cat owners
Branson (2019) [43]	United States	Cat ownership (dichotomous)	UCLA Loneliness Scale Revised (20 items)	Cross-sectional	Community-dwelling individuals without a dog	n = 96	M: 76.6 SD: 9.5 60–100	74.0	Logistic regression did not detect loneliness as a significant covariate of cat ownership (ref.: non-cat ownership)
Carr (2020) [44]	United States	Pet ownership (dichotomous) × social loss (dichotomous)	Composite measure (UCLA Loneliness Scale, Health and Retirement Study Psychosocial and Lifestyle Questionnaire) (three items)	Longitudinal (three waves during 8 years)	Health and Retirement Study	n = 437	M: 65.6 SD: 10.1 37–88	56	Pet ownership (ref.: non-pet ownership) did not significantly affect changes in loneliness following a social loss
Enmarker (2015) [19]	Norway	Pet ownership (dichotomous)	Loneliness: four-point scale	Cross-sectional	Nord-Trøndelag Health Study	n = 12,093	M: 74.8 SD: 6.5 65–101	54.3	“There was a slight difference in pet ownership in relation to loneliness: 16.5% of participants who indicated that they were lonely owned a pet compared with 18% of participants who indicated that they were not lonely.”
Gulick (2012) [52]	United States	Dog ownership (dichotomous) Cat ownership (dichotomous)	UCLA Loneliness Scale (20 items)	Cross-sectional	Individuals who own a dog or cat, utilize services for older people and can communicate in English	n = 159	55–72: 50.9% 73–84: 49.1%	100.0	There were no significant differences among loneliness between cat and dog owners
Hajek (2020) [21]	Germany	Dog ownership (dichotomous) cat ownership (dichotomous)	Loneliness: De Jong Gierveld Loneliness Scale (11 items) Social isolation: scale from Bude & Lantermann, 2006 (four items)	Cross-sectional	German Ageing Survey	n = 1,160	M: 75.1 SD: 6.4 65–95	65.4	Linear regression showed that dog ownership (ref.: not owning a pet) was related to decreased levels of social isolation (ß = – 0.16, p < 0.05) and loneliness (ß = – 0.12, p < 0.1). Cat ownership (ref.: not owning a pet) remained insignificant
McConnell (2011) [53]	United States	Pet ownership (dichotomous)	UCLA Loneliness Scale (20 items)	Cross-sectional	Community sample	n = 217	M: 31 SD not specified Range not specified	79	Regarding t tests, pet owners (ref.: non-pet ownership) had lower loneliness scores (p < 0.08)
Pikhartova (2014) [39]	United Kingdom	Pet ownership (dichotomous)	Revised UCLA Loneliness Scale (three items)	Longitudinal (five waves during 9 years)	English Longitudinal Study of Ageing	n = 5,210	M: 61.4 SD not specified Range not specified	55.8	According to logistic regression, pet ownership (ref.: non-pet ownership) was associated with increased odds of loneliness in the cross-sectional analysis (OR: 1.24, 95% CI: 1.06–1.47) In the longitudinal analysis, pet ownership (ref.: non-pet ownership) was also related to higher chances of loneliness (e.g., wave 0 to wave 5: OR: 1.31, 95% CI: 1.03–1.68)
Powell (2018) [20]	Australia	Dog ownership (current or past or not)	Expectation that dog ownership would result in a decrease among loneliness	Cross-sectional	Potential dog owners	n = 3,465	18–44: 52.0% 45–64: 39.0%  ≥ 65: 9.0%	85.0	According to logistic regression, current dog ownership (ref.: never owned a dog) was significantly related to higher expectations that a dog benefits to a decrease in loneliness (OR: 1.61, 95% CI: 1.19–2.20). Past dog ownership remained insignificant
Rijken (2011) [48]	Netherlands	Pet ownership (dichotomous) Dog ownership (dichotomous) Cat ownership (dichotomous) Dog and cat ownership (dichotomous) Other pet ownership (no cats or dogs) (dichotomous)	UCLA Loneliness Scale Revised (six items)	Cross-sectional	National Panel of People with Chronic Illness or Disability	n = 1,410	M: 74.6 SD: 6.4 Range not specified	60.0	With respect to ANOVA, there were no significant differences between the different types of pet ownership among loneliness
Stanley (2014) [18]	United States	Pet ownership (dichotomous)	Felts of loneliness during the last 2 weeks (dichotomized)	Cross-sectional	Primary care patients	n = 830	M: 72.2 SD: 8.3 Range not specified	57.8	According to logistic regression, pet ownership (ref.: non-pet ownership) was not significantly associated with loneliness Though, living alone x pet ownership was significantly related to decreased odds of loneliness (OR: 0.20, 95% CI: 0.08–0.50)
Taniguchi (2018) [7]	Japan	Pet ownership (current or past vs. not)	Social isolation: having contact with others less than once a week	Cross-sectional	Ota Genki Senior Project	n = 11,233	65–74: 47.7% 75–84: 52.3%	51.6	Referring to mixed-effects cumulative logistic regression models, social isolation was related to decreased chances of current or past pet ownership (ref.: non-pet ownership) (OR: 0.74, 95% CI: 0.66–0.80)
Zasloff (1994) [17]	United States	Pet ownership (dichotomous)	UCLA Loneliness Scale-Revised (number of items not specified)	Cross-sectional	Single students who do not live with a mate, a significant other, or children under the age of 18	n = 148	M: 28.4 SD: 8.3 21–53	100.0	There were no significant differences in loneliness among pet owners and non-pet owners
Children/adolescent prior to the pandemic
Black (2012) [36]	United States	Pet ownership (dichotomous)	UCLA Loneliness Scale Revised (20 items)	Cross-sectional	Rural adolescents who visit public high schools	n = 293	M: 15.8 SD: 1.3 13–19	54.1	An ANOVA showed that individuals with pets (ref.: non-pet ownership) had significantly lower loneliness scores (p < .001)
Charmaraman (2020) [51]	United States	Pet ownership (dichotomous) dog ownership (dichotomous)	Social isolation: two items	Cross-sectional	Middle school students	n = 700	M: 12.7 SD not specified 11–16	52	Social isolation was negatively associated with dog ownership (ß = – 0.23, p < 0.05, ref.: pet, but non-dog ownership), but not with pet ownership (ref.: non-pet ownership) in general, according to regression analysis
Mueller (2021) [45]	United States	Pet ownership (dichotomous) Dog ownership (dichotomous)	Three-point scale	Longitudinal (two waves in 10 months)	Adolescents visiting Middle schools	n = 1,033	M: 12.69 SD: 1.21	50	Dog ownership (ref.: non-dog pet ownership) was related to decreased levels of loneliness (ß = – 0.1, p < 0.05), according to regression analysis
Rhoades (2015) [37]	United States	Pet ownership (dichotomous)	UCLA Loneliness Scale (three items)	Cross-sectional	Homeless youth who utilize drop-in centers	n = 398	M: 21.3 SD: 2.1 Range not specified	27.4	Regarding Chi-square tests, pet ownership (ref.: non-pet ownership) was associated with decreased levels of loneliness (p < .05)
Adult population during the pandemic
Carr (2021) [22]	United States	Dog ownership (dichotomous) Cat ownership (dichotomous)	Composite measure (UCLA Loneliness Scale, Health and Retirement Study Psychosocial and Lifestyle Questionnaire) (three items)	Longitudinal (two waves during 2 years)	Community-based sample	n = 473	M: 69.4 SD: 6.1 60–92	66.0	According to the fully adjusted regression model, neither dog ownership (ref.: non-dog ownership) nor cat ownership (ref.: non-cat ownership) was significantly associated with loneliness
Kogan (2021) [23]	Mostly United States	Dog ownership (dichotomous) Cat ownership (dichotomous)	Loneliness: five-point scale Social isolation: five-point scale	Cross-sectional	Dog or cat owners who participated in an online survey	n = 5,061	≤ 39: 30% 40–59: 43%  ≥ 60: 27%	89	Most of the pet owners reported that their pet would decrease their loneliness (66%) and their feelings of isolation (64%) Regarding binary regression, cat owners were less likely to feel isolated than dog owners (OR: 0.74, 95% CI: 0.64–0.86). Concerning loneliness, there no significant differences between these groups were revealed
Oliva (2021) [46]	Australia	Dog ownership (dichotomous) Cat ownership (dichotomous)	UCLA Loneliness Scale (three items) Loneliness during COVID-19 lockdown, rated on a four-point scale	Cross-sectional	Individuals living alone	n = 384	M: 50.9 SD: 15.1 23–89	85.4	Referring to hierarchical logistic regression, dog ownership (ref.: non-dog ownership) was associated with decreased levels of loneliness among both measures (e.g., UCLA Loneliness Scale: ß = – 0.71, p < 0.05). Cat ownership (vs. non-cat ownership) remained insignificant
Phillipou (2021) [24]	Australia	Pet ownership (dichotomous)	UCLA Loneliness Scale-Revised (number of items not specified)	Cross-sectional	Covid-19 and you: mentaL heaLth in AusTralia now survEy	n = 263	M: 25.1 SD: 14.2 range not specified	84.2	Pet ownership (ref.: non-pet ownership) was not significantly related to loneliness
Ratschen (2020) [47]	United Kingdom	Pet ownership (dichotomous)	UCLA Loneliness Scale (three items)	Cross-sectional	General population	n = 5,926	18–24: 7.1% 25–34: 17.5% 35–44: 16.8% 45–54: 23.8% 55–64: 22.2% 65–70: 7.1%  ≥ 70: 5.6%	78.6	Looking at linear regression models, pet ownership (ref.: non-pet ownership) was associated with a decreased height of loneliness (p < 0.01)
Children/adolescent during the pandemic
Mueller (2021) [45]	United States	Pet ownership (dichotomous) Dog ownership (dichotomous)	Three-point scale	Longitudinal (two waves in 10 months)	Adolescents visiting Middle schools	n = 357	M: 12.69 SD: 1.21	50	Pet ownership (ref.: non-pet ownership) was significantly associated with increased loneliness during COVID-19 (ß = 0.12, p < 0.05) Pet owners (ref.: non-pet owners) reported significantly higher increases in loneliness during COVID-19 (ß = 0.14, p < 0.01)