Pet ownership and risk of depression: a systematic review and meta-analysis

Abstract

Background

Pet ownership is often believed to confer psychological benefits, such as reducing loneliness and providing emotional support. However, evidence on its relationship with depression is mixed, and no clear consensus currently exists. This systematic review and meta-analysis aimed to evaluate the association between pet ownership and the risk of depression.

Methods

A comprehensive systematic review and meta-analysis were performed following PRISMA guidelines. Three electronic databases (PubMed, Scopus, Web of Science) were searched for observational studies assessing the impact of pet ownership on depression. Two independent reviewers screened and extracted data, and study quality was evaluated using the Newcastle–Ottawa Scale. Random-effects models were used to compute pooled odds ratios (ORs) and 95% confidence intervals (CIs) using STATA-17.

Results

A total of 21 studies involving 159,322 participants were included. Overall, pet ownership was not associated with a significant change in depression risk compared to non-ownership (OR: 1.03; 95% CI: 0.995–1.07). However, sensitivity analyses by pet type revealed that cat ownership was associated with a modestly increased risk of depression (OR: 1.06; 95% CI: 1.02–1.09), whereas dog ownership showed no significant association (OR: 0.93; 95% CI: 0.789–1.10).

Conclusion

This study reveals a complex relationship between pet ownership and depression. Cat ownership is linked to a higher risk, while dog ownership shows mixed results. Overall, pet ownership isn’t significantly associated with depression, highlighting the need for further research into its psychosocial dynamics and mental health implications.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12991-025-00600-x.

Introduction

Pet ownership has shown promising psychological benefits, including reduced loneliness, enhanced social support, and improved mental well-being [1]. Research highlights that pet ownership has been associated with lower levels of loneliness, depression, and improved well-being in older adults [2]. Pets like cats and dogs foster connection and reduce isolation. Activities such as petting or being near animals lower blood pressure and decrease stress, reinforcing pet ownership’s potential to prevent psychological distress, including depression [1, 3, 4]. In the U.S., approximately 94 million pets are owned, up from 82 million in 2023, according to the American Pet Products Association (APPA) [5]. Through companionship, a connection to nature, and the human-animal bond—mutually beneficial for humans and animals—pets significantly contribute to physical and mental well-being [6].

The relationship between pet ownership and mental health is shaped by various advantages and challenges. Positively, pets offer emotional connection, social support, and stress relief through physical contact. Dog walking promotes physical activity, yielding both physical and psychological gains [7, 8]. Pet care routines also encourage psychological stability. However, caregiving stress, financial burden, pet loss, or behavioral problems may reduce these benefits, revealing the complexity of pet ownership’s impact on mental health [9].

Studies examining the link between pet ownership and depression have produced inconsistent findings. While some report no significant difference in depression levels between pet owners and non-owners [10, 11], while others suggest a positive [12] or negative [13, 14] associations depending on context. Additionally, secure attachment to pets is linked to lower depression levels, whereas insecure attachment may increase depression risk [6, 15].

However, pet ownership is generally believed to improve mental well-being, current research indicates that its psychological impacts may differ depending on the type of pet. For instance, dogs are often linked to increased physical activity and social interaction, while cats display greater independence, potentially providing a different form of companionship that may provide different emotional effects [6, 16]. Studies of human–animal interaction during the COVID-19 pandemic have showed potential benefits, like reduced loneliness and depressive symptoms in the general population, but these advantages were not reported among pet owners themselves and had no significant association between the pet ownership and mental health, especially after adjusting for confounding factors such as social support and demographic characteristics; in some cases, pet ownership was even associated with greater stress and poorer perceived mental health [17–19]. These effects appear to depend on several factors, including pet type, level of attachment, and demographic or health-related characteristics [20]. For instance, one study found that individuals living alone with a pet had a higher risk of depression compared to those living with both people and pets [14], while another reported that dog walking may improve mental health through physical activity and social engagement, although these benefits varied by age and context [21]. Further contributing to the mixed evidence, a study has reported that cat ownership may be associated with increased risk of certain psychiatric outcomes, including schizophrenia [22].

The psychological impact of pet ownership also appears to differ across clinical and vulnerable populations; for example, ICU survivors may experience distinct benefits or challenges, suggesting that life circumstances and health status substantially shape outcomes [23].

These findings highlight that pet ownership is not uniformly protective and justify the need to examine pet types separately when evaluating their associations with mental health, including depression and highlight the importance of comprehensive systematic reviews and meta-analyses to better understand pet ownership’s influence on depression. Despite the volume of studies, findings remain unclear due to demographic factors, methodological limitations, and inconsistent study designs [6]. Existing reviews have not sufficiently investigated the strength of this association. This study addresses these gaps by conducting a systematic review and meta-analysis to clarify the relationship between pet ownership and depression risk, aiming to support public health professionals and clinicians with deeper insight into pet ownership’s psychological benefits.

Methods

This review followed 2020 PRISMA guidelines and was registered in PROSPERO with the registration number CRD42024618041 to ensure methodological transparency [24].

Search strategy

We identified relevant studies using the primary keywords “Pet ownership,” “Cat ownership,” “Dog ownership,” and “Depression,” along with their synonyms and MeSH terms. Searches were conducted from inception until November 2024, focusing exclusively on peer-reviewed, English-language studies. The databases searched included PubMed, Scopus, and Web of Science, (search details are available in Supplementary Material 1). To minimize the risk of missing relevant literature, grey literature was reviewed by examining references from included studies, manually searching Google Scholar, and assessing studies cited in relevant reviews.

Eligibility criteria

To develop the PECO (Population, Exposure, Comparator, Outcome) framework for the study, we defined each component as follows:

• Population: Participants of all ages and genders who owned pets, particularly those at risk of depression, were included. Individuals with pre-existing mental health disorders were excluded.

• Exposure: Ownership of pets, specifically cats or dogs (and other pets), and the frequency of interaction with these pets. Eligible participants had direct engagement with their pets.

• Comparator: Non-pet owners or individuals with limited interaction with pets served as the baseline group to evaluate the impact of pet ownership on mental health.

• Outcomes: Depressive symptom levels measured using standardized tools such as the Beck Depression Inventory (BDI) and the Hamilton Depression Rating Scale (HADS). Studies relying solely on self-reported depression without validated questionnaires were excluded.

Design: Observational studies case-control, cohort and cross-sectional designs that reported odds ratios (OR), risk ratios (RR), beta coefficients, or effect sizes convertible to ORs were included.

Study selection and data extraction

Search results from the databases were imported into EndNote for duplicate removal. Subsequently, Rayyan [25] was used for systematic screening. Titles and abstracts were screened during the initial phase, followed by a thorough review of full texts to finalize study selection.

Data extraction was conducted meticulously, with results documented in two summary tables:

• Table 1: General characteristics, including author, country, design, ethnicity, participant data (pet and non-pet owners: N, mean age, M/F), and Newcastle–Ottawa Quality Assessment Scale (NOS) scores.

• Table 2: Study details such as participant definitions, depression assessment tools, key findings, and limitations.

Table 1.

General characteristics of all included studies

Author	Country	Design	Ethnicity	Non-pet (N. mean age. M/F)	Pet (N. mean age. M/F)	NOS
Siegel 1999 [29]	USA	Cohort	Non-Hispanic white (96%), other (4%)	965, 381, 342/0	Any pet: 907, 381, 366/0 Dog: 422, 381, 179/0 Cat: 502, 381, 203/0	5/moderate
Irani 2006 [30]	Switzerland	CS	NR	43, 46 (12), 25/18	Any pet: 46, 43 (14), 23/23	5/moderate
Lentino 2012 [31]	USA	CS	Caucasian (86%), other (14%)	380, 39 (14), 114/266	Dog: 399, 40 (13), 64/335	8/low
Enmarker 2015 [32]	Norway	CS	NR	10,196, 74.8 (6.45)1, 4557/5639	Dog: 814, 74.8 (6.45)1, 439/375 Cat: 1083, 74.8 (6.45)1, 536/547	7/low
Lem 2016 [33]	Canada	CS	Caucasian (73%), African/Caribbean (13.5%), First nation/Aboriginal (6.7%), Asian (4.4%), other (2.2%)	100, 20.79 (2.4), 69/31	Any pet: 89, 20.9 (2.4), 54/35	6/moderate
Batty 2017 [34]	UK	Cohort	NR	5815, 69 (9), 2616/3199	Dog: 1619, 65 (8), 719/900 Cat: 1077, 65 (8), 460/617 Other: 274, 64 (8), 127/147	7/low
Muldoon 2017 [35]	USA	CS	Hispanic/Spanish/Latino (9%), none Hispanic/Spanish/Latino (91%)	63, 48.72 (10.68), 172/272	Dog: 136, 48.72 (10.68), 172/272	7/low
Mein 2018 [36]	UK	CS	NR	4646, 66.4 (6), NR	Any pet: 1929, 64.9 (5.7), NR Dog: 718, 65.1 (5.7), NR Cat: 1199, 64.9 (5.7), NR	7/low
Branson 2019 [37]	USA	CS	White (88%) and non-white (12%)	55, 79.62 (9.87), 13/42	Cat: 41, 72.56 (7.30), 12/29	5/moderate
Sharpley 2020 [38]	UK	CS	White (97.2%), non-white (2.8%)	5042, 68.02 (8.89), 2395/2647	Dog: 1381, 64.33 (7.57), 634/747 Cat: 959, 64.40 (7.98), 450/509 Other: 235, 63.24 (7.79), 118/117	7/low
Bohn 2021 [12]	Brazil	CS	White (34.8%), other (65.2%)	339, 67.68 (5.91), 104/10192	Dog: 363, 67.68 (5.91), 104/10192 Cat: 253, 67.68 (5.91), 104/10192 Other: 168, 67.68 (5.91), 104/10192	6/moderate
Chakma 2021 [39]	Bangladesh	CS	NR	140, 26.52 (5.41), 73/67	Any pet: 140, 26.61 (6.18), 44/96	7/low
Cui 2021 [40]	Australia	Cohort	NR	546, 43.6 (13.11), 224/322	Dog: 477, 42.1 (11.13), 167/310	8/low
Opdebeeck 2021 [41]	UK	Cohort	NR	1075, 77.80, 588/487	Any pet: 467, 73.01, 278/189 Dog: 271, NR, NR	8/low
Silva 2021 [42]	Portugal	CS	NR	42, 44 (11.11), 0/42	Dog: 64, 44 (8.89), 0/64	5/moderate
Xin 2021 [43]	China	CS	NR	575, 57.8 1, 173/402	Any pet: 181, 57.8 1, 36/145	8/low
Lima 2022 [44]	Portugal	CS	NR	164, 40 (12), 51/113	Dog: 345, 38 (12), 70/275	8/low
Matsumura 2022 [13]	Japan	Cohort	NR	66,468, 30.9 (4.93) 1, 0/66,468	Dog: 8958, 30.9 (4.93)1, 0/14,338 Cat: 4093, 30.9 (4.93)1, 0/4093 Both: 1295, 30.9 (4.93)1, 0/1295	8/low
Miyake 2023 [14]	Japan	CS	NR	10,504, 41.48 (12.53), 9235/1269	Any pet: 2259, 46.9 (10.8), 1988/271	7/low
Z˙ebrowska 2023 [11]	USA	Cohort	White (93.7%), non-white (6.3%)	1339, 14.9 (1.6), 617/729	Any pet: 8292, 14.9 (1.6), 3474/4818	8/low
Hajek 2024 [10]	Germany	CS	NR	2611, 85.6 (4.2), 993/1618	Any pet: 256, 84.88 (3.71), 126/130	7/low

Abbreviations: CS: Cross Sectional; NR: Not Reported; N: Number; M/F: Male/Female; NOS: Newcastle-Ottawa Scale

¹Median age was reported for all participants

²Mean age and M/F reported for all participants

Table 2.

Details of all included studies

Author	Definition participants	Depression tool	Finding	Limitations
Siegel 1999 [29]	Homo- and bisexual HIV patients ≥ 18 y/o	CES-D	Pet owner =+non-pet	-non-representative population -no randomization
Irani 2006 [30]	Lung transplant recipients	HADS	Pet owner = non-pet	-low sample size -CS design -non-representative population -observational study -absence of a healthy control group -pets was mainly cats and dogs
Lentino 2012 [31]	General population ≥ 18 y/o	CESD-10	Dog owner < non-pet	-CS design -self reported medical information -non-representative population -low sample size
Enmarker 2015 [32]	General population ≥ 65y/o	HADS-D	Dog owner = non-pet Cat owner > non-pet	-not inclusion of most severe case of depression -CS design -only inclusion of rural residents
Lem 2016 [33]	16y/o ≤ street-involved youths ≤ 24y/o	CES-D	Pet owner < non-pet	-low sample size -CS design -non-representative -self reported information -selection bias
Batty 2017 [34]	General population ≥ 50y/o	CES-D	Dog owner = non-petCat owner = non-petOther pet = non-pet	-underestimation of true effect of pet ownership due to measuring on one occasion -confounding result by unknown factor
Muldoon 2017 [35]	HIV patients ≥ 18 y/o	CESD-10	Dog owner < non-pet	-low sample size -CS design -non-representative population
Mein 2018 [36]	59y/o ≤ civil servants working ≤ 79y/o	CES-D	Pet owner = non-petDog owner = non-petCat owner = non-pet	-CS design -the possibility of recal bias -non-representative population
Branson 2019 [37]	General population ≥ 60y/o	GDS-SF	Cat owner = non-pet	-CS design -non-representative
Sharpley 2020 [38]	General population ≥ 50y/o	CES-D	Dog owner > non-pet Cat owner = non-pet Other pet = non-pet	-small sample size -self reported data
Bohn 2021 [12]	General population ≥ 60y/o	GDS-15	Dog owner < non dog owner Cat owner = non-pet Other pet = non-pet	-CS design -non-representative
Chakma 2021 [39]	General population ≥ 13y/o	PHQ-9	Pet owner = non-pet	-CS design -small sample size -recall bias
Cui 2021 [40]	Participants taking part in the RESIDential Environments project	6 item Kessler Psychological Distress Scale	Dog owner = non-pet	-small number of stress and depression cases -low statistical power -self reported data -non-representative (predominately female, married and of middle age).
Opdebeeck 2021 [41]	Patients with mild to moderate dementia	GDS-10	Dog owner = non-pet Pet owner = non-pet	-non-representative
Silva 2021 [42]	Fibromyalgia patients	HADS	Dog owner = non-pet	-CS design -small sample size -non-representative -low response rate
Xin 2021 [43]	General population ≥ 12 y/o	DASS-21	Pet owner = non-pet	-CS design -self report
Lima 2022 [44]	General population ≥ 18y/o	HADS-D	Pet owner = non-pet	-CS design -non-representative
Matsumura 2022 [13]	Mothers and pregnants in an ongoing nationwide birth cohort study (JEXS)	EPDS	Dog owner < non-pet Cat owner > non-pet	-non-representative -self report
Miyake 2023 [14]	Employees of five companies in 2018–2021	CES-D-11	Pet owner = non-pet	-CS design -non-representative -self report
Z˙ebrowska 2023 [11]	Adolescents and young adults (participants in GUTS cohort)	MRFS	Pet owner = non-pet	-non-representative -self report
Hajek 2024 [10]	Participants in “Old Age in Germany (D80+)” study	DIA-S4	Pet owner = non-pet	-CS design -non-representative -moderate response rate

Abbreviations: CS: Cross Sectional; NR: Not Reported; N: Number; M/F: Male/Female; HIV: The human immunodeficiency viruses; CES-D: Center for Epidemiologic Studies Depression Scale; HADS: Hospital Anxiety and Depression Scale; CESD: Center for Epidemiologic Studies Depression Scale; GDS: Geriatric Depression Scale; GDS-SF: Geriatric Depression Scale - Short Form; PHQ: Patient Health Questionnaire; DASS-D: Depression Anxiety Stress Scales - Depression subscale; MRFS: Modified Resilience Framework Scale; DIA-S4: Dementia Impact Assessment Scale − 4 item version

Two authors independently conducted the screening and data extraction processes, and disagreements were resolved by a third author.

Risk of bias (ROB) assessment

The Newcastle–Ottawa Scale [26] was employed to assess the risk of bias, categorizing studies as high (≤ 4 stars), moderate (5–6 stars), or low risk (≥ 7 stars). The NOS assigns a maximum of 9 stars across selection, comparability, and outcome/exposure assessment domains.

Quantitative analysis

OR and standard errors (SE) were extracted for the meta-analysis. Transformations were applied where necessary to standardize effect sizes [27]. Random-effects models were employed using maximum likelihood and restricted maximum likelihood methods to calculate pooled effect sizes.

Analyses were conducted in two phases:

1. Overall analysis: Pooled ORs compared depression rates between pet owners and non-owners.

2. Subgroup analysis: Specific comparisons of cat and dog owners with non-owners.

Sensitivity analyses employed the leave-one-study-out method to evaluate the influence of individual studies on the results. Heterogeneity was assessed using the chi-square test and I² statistic, where values of 0%, 25%, 50%, and 75% corresponded to no, low, moderate, and high heterogeneity, respectively [28].

Publication bias was evaluated using funnel plots, Egger’s test, and Begg’s test. If bias was detected, the trim-and-fill method was applied to estimate missing studies and adjust the effect size.

Subgroup and meta-regression analyses

Subgroup meta-analyses were conducted based on:

• Pet type: Any pet, dog, cat, or other.

• Geographical region: North America and Australia, Europe, Asia, and South America.

• Study design: Case-control or cohort.

• Sample size: Cutoff of 10,000 participants.

• Mean age: Cutoff of 50 years.

• Presence of comorbidities.

Meta-regression analyses

Meta-regression analyses explored variables, including mean age, male percentage, smoking and alcohol consumption rates, marital status, loneliness prevalence, and body mass index (BMI).

All statistical analyses were performed using STATA-17, with a significance threshold of p < 0.05.

Results

Selection of studies

From an initial pool of 525 studies, 195 duplicates were automatically removed using EndNote. This left 330 articles for further consideration. Of these, 63 underwent full-text screening, and 21 studies [10–14, 29–44] were ultimately included in the analysis (Fig. 1).

Fig. 1.

The Preferred Reporting Items for Systematic Reviews and meta-analysis (PRISMA) flow diagram of search results

Study characteristics

We identified 21 articles conducted from 1999 to 2024. Five research were performed in USA [11, 29, 31, 35, 37], four in UK [34, 36, 38, 41], two in Japan [13, 14], two in Portugal [42, 44], one each in Australia [40], Bangladesh [39], Brazil [12], Canada [33], China [43], Germany [10], Norway [31], and Switzerland [30]. Almost all studies were cross-sectional, except for six studies [11, 13, 29, 34, 40, 41], which were cohort studies. A total of 159,322 participants were included in our analysis. The number of pet owners and non-pet owners included in each study summarized in Table 1. One study [29] included only male participants, while two studies [13, 42] included only female participants. The age range of participants was from 14.9 years [11] to 85.6 years [10]. Six studies reported comorbidities including HIV [29, 35], lung transplant [30], mild to moderate dementia [41], fibromyalgia [42], and pregnancy [13]. The remaining 15 studies [10–12, 14]31– [34]36– [40, 43, 44] had no reported comorbidities. Depression assessment tools used across these studies were the Center for Epidemiologic Studies Depression Scale (CES-D) [14, 29, 31]33– [36, 38], HADS [30, 32, 42, 44], Geriatric Depression Scale (GDS) [12, 37, 41], Patient Health Questionaire (PHQ) [39], 6-item Kessler Psychological Distress Scale [40], Depression Anxiety Stress Scale-21 (DASS-21) [43], Edinburgh Postnatal Depression Scale (EPDS) [13], McKnight Risk Factor Survey (MRFS) [11], and short form of the Depression in Old Age Scale (DIA-S4) [10] (Table 1).

Included studies demonstrated that in most of studies [10, 11, 14, 29, 30, 36, 39, 41, 43, 44] the risk of depression in pet owners is equal to that of non-pet owners. One study indicated that depression is less prevalent in pet owners than in non-pet owners [33]. Two studies[31, 35] found that the occurrence of depression is lower in dog owners compared to non-pet owners. Two studies [40, 42] revealed that dog owners have similar depression risk to non-pet owners. One study [37] found that cat owners have a similar OR to non-pet owners. Enmarker et al. [32] demonstrated that dog owners have depression rates similar to non-pet owners, while cat owners are more prone to depression than non-pet owners. Sharply et al. [38] showed that depression is more common in dog owners than in non-pet owners, but the occurrence of depression in cat and other pet owners is similar to non-pet owners. Bohn et al. [12] found that dog owners are less prone to depression compared to non-pet owners, while cat and other pet owners have similar depression rates to non-pet owners. Opdebeek et al. [41] revealed that the OR of depression in pet and dog owners are is comparable to non-pet owners. Matsumura et al. [13] demonstrated that depression is more common in cat owners, while dog owners have less depression rates compared to non-pet owners. Mein et al. [36] found that pet, dog, and cat owners have similar risk of depression in comparison to non-pet owners. Batty et al. showed that likelihood of depression among dog, cat, and other pet owners did not differ from that of non-pet owners [34] (Table 2).

Risk of bias within studies

Out of the included studies, 15 were assessed as having a low risk of bias, while six were determined to have a moderate risk of bias (Table 1).

Synthesis of results

Pet owners vs. Non-owners

A total of 20 studies [10–14, 30–44], comprising 32 effect sizes and 142,251 participants, were included in the analysis. The findings indicated that pet ownership did not significantly increase the risk of depression compared to non-ownership (OR: 1.039; 95% CI: 0.995–1.07; p = 0.09; I² = 47.88%) (Fig. 2).

Fig. 2.

Meta-analysis comparing depression risk between pet owners and non-pet owners

However, sensitivity analyses suggested that excluding the studies by Lentino et al. [31] (OR: 1.03; 95% CI: 1.00–1.07), Matsumura et al. (OR: 1.04; 95% CI: 1.02-1.07) and Bohn et al. [12] (OR: 1.04; 95% CI:1.01 –1.07) could potentially shift the non-significant findings to a significantly higher risk of depression among pet owners compared to non-owners (SM2, Figure S1).

Publication bias was assessed using funnel plots and counter-enhanced funnel plots, alongside Egger’s test (p = 0.005) and Begg’s test (p = 0.192). These analyses indicated no evidence of publication bias (SM2, Figures S2–S3).

Subgroup meta-analysis revealed no significant differences between the groups p(Fig. 3).

Fig. 3.

Subgroup meta-analysis comparing depression risk between pet owners and non-pet owners

Meta-regression analysis did not identify any variables that significantly contributed to the observed heterogeneity (Table 3).

Table 3.

Meta-regression analysis of key variables for all studies included, focusing on all studies, dog ownership and Cat ownership

Variables	All			Dog			Cat
	N of effect sizes	Coefficient [95% CI]	I2	N of effect sizes	Coefficient (95% CI)	I2	N of effect sizes	Coefficient (95% CI)	I2
Mean age	32	0.074 [−0.021, 0.169]	64.78	12	0.182 [−0.173, 0.536]	88.05	7	−0.012 [−0.107, 0.084]	14.41
Male (%)	28	−0.000 [−0.004, 0.003]	88.48	10	−0.003 [−0.013, 0.007]	91.77	6	−0.000 [−0.003, 0.002]	24.37
Current smoking (%)	18	−0.005 [−0.011, 0.002]	73.65	5	0.034 [−0.000, 0.067]	88.21	5	−0.005 [−0.012, 0.002]	0.00
Marital statues (%)	16	−0.000 [−0.003, 0.002]	45.12	5	−0.004 [−0.005, −0.002]	0.01	5	−0.000 [−0.005, 0.004]	31.14
Current alcohol use (%)	9	0.000 [−0.001, 0.001]	67.05	NA	NA	NA	NA	NA	NA
Loneliness (%)	12	0.009 [−0.009, 0.027]	73.62	4	0.049 [0.025, 0.073]	8.49	4	−0.022 [−0.057, 0.012]	45.72
BMI	11	0.006 [−0.011, 0.023]	46.77	4	0.009 [−0.074, 0.091]	81.84	4	−0.003 [−0.021, 0.015]	0.00

Dog owners vs. Non pet owners

A total of 12 studies [12, 13, 31, 32, 34–36, 38, 40–42, 44], comprising 12 effect sizes, were included in the analysis. The findings indicated that owning a dog did not significantly increase the risk of depression compared to persons without pet (OR: 0.929; 95% CI: 0.724–1.10; p = 0.39; I² = 95.32%) (Fig. 4).

Fig. 4.

Meta-analysis comparing depression risk between dog owners and non-pet owners

The sensitivity analysis indicated that removing any single study did not alter the overall pooled OR (SM2, Figure S4).

Publication bias was assessed using funnel plots and adjusted funnel plots, along with Egger’s test (p = 0.380) and Begg’s test (p = 0.582). These analyses showed no evidence of publication bias (SM2, Figures S5–S6).

Subgroup meta-analysis revealed significant results based on geographic location (Fig. 5). Additionally, meta-regression analysis identified loneliness (coefficient: 0.049; 95% CI: 0.025, 0.073) and marital statues (coefficient: −0.004; 95% CI [−0.005, −0.002])as a significant factor contributing to the observed heterogeneity (Table 3).

Fig. 5.

Subgroup meta-analysis comparing depression risk between dog owners and non-pet owners

Cat owners vs. Non pet owners

In this analysis, seven studies [12, 13, 32, 34, 36–38] comprising seven effect sizes were included. The results indicated that cat ownership was associated with an increased risk of depression compared to persons without pet (OR: 1.06; 95% CI: 1.02–1.09; p < 0.001; I² = 10.31%) (Fig. 6). However, sensitivity analyses suggested that excluding the studies by Matsumura et al. (OR: 1.06; 95% CI: 0.999-1.12) could potentially shift the significant findings to a non-significant risk of depression among cat owners compared to non-owners(SM2, Figure S7).

Fig. 6.

Meta-analysis comparing depression risk between cat owners and non-pet owners

Publication bias was assessed using funnel plots and adjusted funnel plots, along with Egger’s test (p = 0.39) and Begg’s test (p = 0.54)which revealed no evidence of bias (SM2, Figures S8–S9).

Subgroup meta-analysis revealed no significant differences between the groupsp (Fig. 7).

Fig. 7.

Subgroup meta-analysis comparing depression risk between cat owners and non-pet owners

Meta-regression analysis did not identify any variables that significantly contributed to the observed heterogeneityTable 3).

Discussion

This systematic review and meta-analysis, utilizing data from three databases and 21 studies, explored the relationship between pet ownership and depression. Our findings revealed a nuanced association, with cat ownership linked to a modest but significant increase in the risk of depression (p < 0.001), while no significant relationship was found for dog ownership. Cat owners exhibited a higher prevalence of depressive symptoms compared to non-owners, supported by Enmarker et al., which highlighted increased susceptibility to depression among cat owners [32]. Matsumura et al. identified cat ownership as a risk factor for postpartum depression [13], while Connell et al. found that cat owners reported more severe depressive symptoms than both dog owners and individuals without pets [45].

Several mechanisms may explain the differential impact of pet ownership on depression. Cats, for instance, exhibit less social dependency and emotional responsiveness compared to dogs, which may limit their ability to alleviate loneliness or provide consistent emotional support [46]. Sleep disturbances caused by cats’ nocturnal behavior could further contribute to depression, as highlighted in Van Egmond et al.’s study [47]. Conversely, dogs often promote regular physical activity through walking and offer emotional support that alleviates loneliness and depression [48–50]. Additionally, socioeconomic stressors, such as the financial burden of pet care and concerns about pets’ health, may offset the psychological benefits of pet ownership. Studies have shown that heightened anxiety among pet owners facing economic challenges could counteract the potential benefits of companionship [51–53]. The link between cat ownership and depression should be interpreted cautiously, as self-selection bias may influence results. People with a predisposition to depression might be more inclined to own certain pets or none at all, which limits causal interpretations and may explain the observed differences in depression risk.

Our analysis revealed significant heterogeneity among the included studies, driven by differences in geographic location. Geographic variation played a role, with studies conducted in North America and Australia reporting different outcomes compared to those in Europe, Asia, and South America. These differences emphasize the need for standardized methodologies and careful adjustment for confounders in future research. Also, meta-regression analysis showed that loneliness and marital status were significant sources of heterogeneity.

Owning a dog is associated with notable cardiovascular health benefits. Studies show that dog owners experience a 24% decrease in overall mortality and a 31% reduction in cardiovascular-related deaths [54]. Furthermore, having a dog lowers the risk of metabolic syndrome and obesity, which enhances health outcomes for these owners [55]. Additionally, pet ownership influences gut microbiota composition, potentially decreasing the likelihood of metabolic disorders, as pet owners exhibit different microbial profiles compared to those without pets [55]. In terms of mental health, research indicates a significant correlation between cat ownership and an increased risk of schizophrenia. Individuals who have been around cats are more than twice as likely to develop related disorders [56] although some studies do not find a clear link [57]. The findings regarding dog ownership and mental health are mixed; while some studies suggest that dog owners may experience higher levels of psychopathological symptoms, the overall effect of pet ownership on mental health is complex and necessitates further investigation [58]. Overall, the impact of pet ownership on mental health remains intricate and requires additional research to clarify its effects [59, 60].

Research has shown that owning pets, especially cats and dogs, may enhances long-term mental health and overall quality of life. Numerous studies indicate that pet owners often report lower levels of anxiety, depression, and loneliness, which contributes positively to their mental health, even among individuals with severe mental illnesses [61, 62]. Pets, especially dogs, offer emotional support, help build social connections, and encourage increased physical activity—elements that are vital for maintaining good mental health [62]. Additionally, older adults who own dogs tend to exhibit fewer symptoms of psychological disorders, leading to an improved quality of life [63]. A 2023 Morning Consult poll of 2,200 adults found that 86% of pet owners—equally among dog and cat owners—reported positive mental health effects from their pets. Common benefits included reduced stress, emotional support, and companionship [64]. The bond between humans and animals seems to play a crucial role in fostering mental health and life satisfaction across diverse groups [63].

This study has several strengths. We conducted a comprehensive search across multiple databases and grey literature, minimizing the risk of missing relevant studies. Our methodology included subgroup and sensitivity analyses to address heterogeneity and identify key moderators such as pet type and participant demographics. Study quality was assessed using the NOS, ensuring the inclusion of high-quality data. However, several limitations should be acknowledged. All included studies were observational, limiting causal inferences due to potential confounding. Pet ownership and depression were often assessed at single time points, with depression predominantly self-reported, potentially inflating prevalence [29, 32]. Socioeconomic status, regional laws, and cultural practices in pet care may impact both the likelihood of owning a pet and mental health outcomes. Exclusion of non-English studies could have omitted relevant data. Moreover, few studies examined specific subgroups like older adults or individuals with chronic illnesses, limiting broader applicability [44, 45].

Future research should prioritize longitudinal designs to establish causality between pet ownership and mental health [62, 65]. Investigating aspects such as interaction frequency and quality could offer deeper psychological insights [49, 50]. Studies should include diverse populations and consider cultural contexts to enhance generalizability [44]. Employing standardized tools for assessing pet ownership and depression would improve cross-study comparability. Evaluating differential effects of various pets on mental health across demographics could further inform interventions promoting companion animals for well-being.

Conclusions

In conclusion, this systematic review and meta-analysis provides valuable insights into the complex relationship between pet ownership and depression. While cat ownership is associated with an increased risk of depression, dog ownership demonstrates mixed effects. These findings highlight the need for personalized approaches when considering pet ownership as a strategy to improve mental health. By addressing the current limitations and advancing research methodologies, future studies can better inform interventions aimed at leveraging the benefits of companion animals for mental health.

Author contributions

RM initiates the concept and leads the initial investigation. SR and MK are tasked with data extraction and screening. SM and FG handle the evaluation and integrate the findings, while RM and GA are responsible for drafting the initial document. RM also conducts the analysis. JA and RM provide oversight throughout all phases of the project. The final version for publication receives approval from all authors.

Funding

Not applicable.

Data availability

All necessary information is provided in the text and materials; additional details can be requested via email from the corresponding author.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Clinical trial number

Not applicable.