Pet ownership and lifestyle behaviours of immunosuppressed individuals and their relatives in the context of COVID-19 pandemic

Highlights

• •Immunosuppressed individuals reduce physical activity and increase sedentary time.
• •Pet ownership can alleviate the negative impacts of unhealthy lifestyles changes.
• •Pet ownership is a potential strategy for initiatives preserving and promoting healthier behaviours.

Abstract

The COVID-19 pandemic and containment measures will likely have a detrimental impact on immunosuppressed individuals’ lifestyle behaviours. Increasing evidence suggests that pet ownership is positively associated with healthier lifestyle. Yet, no study has investigated the potential benefits of pet ownership on lifestyle behaviours of immunosuppressed individuals, a population at increased risk of COVID-19 complications. This study aims to examine 1) changes in light, moderate and vigorous intensity physical activity (LPA, MPA, VPA), sedentary time (SED), and sleep duration, assessed by comparing “before COVID-19 pandemic” and “past 7 days” (i.e., current, during pandemic) self-reported behaviours in immunosuppressed individuals and their relatives; 2) to assess if changes in lifestyle behaviours are associated with pet ownership status and whether age is a moderator of these associations. A convenience sample of 132 participants (65.2% female, 41.3% ≥55 years of age) provided self-reported LPA, MPA, VPA (days/week), SED and sleep (min/day) and pet ownership status using an online questionnaire (May-August 2020). Descriptive analyses, paired T-tests, Cohen’s d effect size and linear regressions were conducted. Results show that participants reported a decrease in VPA (−0.56 days/week, d = 0.34; p < 0.01) and an increase in SED (106.79 min/day, d = −0.81; p < 0.01). Stratified analysis revealed that having at least one dog, compared to not owning pets, is associated with a reduced decline in LPA, MPA and VPA and an increase in sleep in participants aged < 55 years old only. Having a dog appears to be positively associated with healthy lifestyle behaviours in younger and middle age immunosuppressed individuals.

1. Introduction

Immunosuppressed individuals have a weakened immune system resulting from certain chronic illness, disease, and medication. Many immunosuppressed individuals live with side effects including an increased risk of developing infectious and cardiovascular disease, weight gain and decreased quality of life (Sommerer and Zeier, 2016). It is important to identify modifiable factors that alleviate these side effects. Lifestyle behaviours such as increased physical activity, decreased sedentary time (SED) and sufficient sleep contribute positively to cardiovascular health (Bailey et al., 2019, Brindle et al., 2019, Janssen et al., 2020, Matricciani et al., 2017), reduce risks of obesity (Janssen et al., 2020, Panahi and Tremblay, 2018), and anxiety and depressive symptoms (Allen et al., 2019, Janssen et al., 2020, McDowell et al., 2019, Pearce et al., 2022, Zhai et al., 2015b, Zhai et al., 2015a). Among immunosuppressed individuals, regular physical activity is an essential strategy for controlling symptoms and disease (Oberoi et al., 2018, Radtke et al., 2020), reducing mortality risk and improving quality of life (Radtke et al., 2020, Takahashi et al., 2018).

Stressful life events such as the COVID-19 pandemic disrupted lifestyle behaviours (Hargreaves et al., 2021). Physical and social distancing (PSD) measures introduced in Canada from mid-March 2020 and intermittently thereafter for several months included closures and lack of or limited access to public places and professionals. In the general population, PSD measures appear to have negatively impacted daily number of steps (i.e. light intensity physical activity (LPA)) (Fitbit Inc, 2020), time spent in moderate and vigorous intensity physical activity (MPA and VPA) (Rhodes et al., 2020), have increased SED (Stockwell et al., 2021), and decreased sleep quality (Flanagan et al., 2021). Younger adults showed greater lifestyle behaviour changes compared to older adults (Castañeda-Babarro et al., 2020, Zajacova et al., 2020). Alternatively, people over 60 years of age were identified as being at increased risk of developing a severe form of COVID-19 and were strongly encouraged to strictly adhere to PSD measures (Public Health Agency of Canada, 2022) which could potentially translate into greater lifestyle behaviour changes.

It is unclear how immunosuppressed individuals’ lifestyle behaviours are affected by stressful life events such as COVID-19 pandemic and its related PSD measures. Immunosuppressed individuals were given strict instructions regarding self-isolation and avoidance of social contact given their increased risk of severe complications due to COVID-19 (Banerjee et al., 2020, Bhoori et al., 2020), which represented additional barriers to maintain healthy lifestyle behaviours during the pandemic (Ammitzbøll et al., 2021, Lupi et al., 2020, Radtke et al., 2021, Radtke et al., 2020). These behaviours likely extend to close relatives of immunosuppressed individuals as they might adopt stricter PSD measures to protect their loved ones (Chan et al., 2020). As such, relatives’ movement behaviours’ may be similarly affected by the pandemic.

Research on human-animal interactions is still relatively new but pet ownership might prevent or mitigate the impact of stressful life events on human lifestyle behaviours. In the general population, pet owners tend to engage in more PA than non-owners, but the results on the intensity of activity are mixed (Christian et al., 2013, Mein and Grant, 2018, Westgarth et al., 2019). A study by Mein and Grant (2018) showed that dog owners engaged in more mild and MPA than non-owners, but there were no significant differences in VPA. Similarly, Westgarth et al. (2019) found that dog owners are more likely to walk in their leisure time, walk for a longer period, and are four times more likely to meet the 150-minute PA guidelines. During the COVID-19 pandemic, pet owners experienced a lower decrease in LPA than non-owners (Rhodes et al., 2020, Tan et al., 2021) and these benefits seemed to increase with age (Tan et al., 2021). Based on the Health Belief Model (Janz and Becker, 1984), we could argue that dogs can provide an incentive or act as a prompt for their owners to walk more often, an activity that remained possible during the COVID-19 pandemic (Mein and Grant, 2018, Tan et al., 2021, Westgarth et al., 2019). Pet ownership also appears to improve sleep quality; dog owners seem to be able to fall asleep more easily than other animal owners and non-owners (Mein and Grant, 2018). However, pet ownership can also bring a set of stressors such as fear of walking the dog in parks that had become crowded due to PSD measures (Applebaum et al., 2020b). Owning a horse also appear beneficial for physical activity (Machová et al., 2019) but the health benefits of owning other types of animals (e.g., cats, birds, lizards, rodents) is sparse (Barroso et al., 2021). In immunosuppressed populations, pet ownership improves quality of life, facilitates symptom management, and provides support and companionship (Chan and Tapia Rico, 2019, Irani et al., 2006, Oliver-Hall et al., 2021).

To date, very few studies have examined the impact of the COVID-19 pandemic on the lifestyle behaviours of immunosuppressed individuals and the association between pet ownership and lifestyle behaviours are unknown. The main objectives of this study were 1) to examine the changes in lifestyle behaviours – LPA, MPA, VPA, SED and sleep – assessed by comparing “before COVID-19 pandemic” and “past 7 days” (i.e., current, during pandemic) behaviours among immunosuppressed individuals and their relatives; 2) to assess if changes in lifestyle behaviours are associated with pet ownership status, and whether age is a moderator of these associations. We hypothesized that the COVID-19 pandemic has reduce physical activity and sleep, while leading to an increase in SED in immunosuppressed people. We also expected that pet ownership, and particularly dog ownership, would have a protective effect, attenuating those aforementioned negative changes and that this protective effect of will vary with the owner’s age. Given the scarse literature, the direction of modification effect is unknown.

2. Materials and methods

2.1. Study design, study population, and sampling strategy

The COVID-Immuno study is part of Projet Laurent, a multidisciplinary research program in Canada initiated and co-developed by patient and caregiver partners and researchers that aims to evaluate and understand the risks and benefits of pet ownership in transplanted and immunosuppressed populations (https://www.projetlaurent.org). In this study, we use cross-sectional data collected immediately following the first COVID-19 outbreak, from May to August 2020, among a sample of immunosuppressed individuals and their relatives. Participants were recruited via an invitation advertised on partner organizations’ websites and social media (Canadian Donation and Transplantation Research Program (CDTRP), Kidney Foundation of Canada (KFC), Canadian Transplant Association (CTA) and Cystic Fibrosis Canada (CFC)). Inclusion criteria included: (i) aged 15 years and older; (ii) English or French-speaking; (iii) being transplanted, immunosuppressed, a close relative (spouse, children, parent) or a donor; and (iv) having access to the internet. Eligible participants were invited to complete an online consent form and questionnaire using Qualtrics (Provo, UT, USA, 2020). The study met the Centre de Recherche du Centre Hospitalier de l’Université de Montréal (CRCHUM) ethical requirements for research involving human subjects and received ethics approval from the Ethics Research Committee of the CRCHUM (#20.018).

2.2. Measures

2.2.1. Light, moderate and vigorous intensity physical activity

Physical activity was assessed using the International Physical Activity Questionnaire (IPAQ), which has shown good test–retest reliability (Spearman ρ = 0.8) and acceptable criterion validity against accelerometers (ρ = 0.30) (Craig et al., 2003). Participants were asked “How many days per week did you do vigorous aerobic activity, like aerobics or fast bicycling?”, “How many days per week did you do moderate aerobic activity, like jogging, bicycling at a regular pace, or doubles tennis?”, and “How many days per week did you walk for at least 10 min?” to assess VPA, MPA, LPA, respectively. IPAQ questions were adapted so that participants reported each activity frequency for two reference periods: before and during the COVID-19 pandemic (i.e. in the past 7 days). To help participants situate the pandemic inception the following information was provided: “The next questions ask about your physical activity, sedentary behaviour, and sleep. For each question, we will ask you to think about a typical week before* the COVID-19 pandemic and The past 7 days. *Note: COVID-19 was declared a global pandemic on March 11, 2020.” Change in LPA, MPA and VPA was calculated as the difference between the number of days per week reported before the COVID-19 pandemic and in the last 7 days, for each of LPA, MPA and VPA, and used as continuous variables.

2.2.2. Sedentary time

SED was also assessed using the IPAQ (Craig et al., 2003) adapted for the two reference periods. Participants were asked to report how much awake time they usually spent sitting, reclining or lying down on a weekday and on a weekend day before and during the COVID-19 pandemic (i.e. in the last 7 days). Average daily SED was then calculated using the following formula for both periods:

$$\left(weekdaysedentarytime\times 5\right)+\left(weekendsedentarytime\times 2\right)/7$$

Change in SED was calculated as the difference between average daily minutes of SED before the COVID-19 pandemic and in the last 7 days and used as a continuous variable in all analysis.

2.2.3. Sleep

Sleep was also assessed using the IPAQ (Craig et al., 2003) adapted for the two reference periods. Participants were asked to report “On average, how long did you sleep on a typical weekday night?” and “on a weekend night? before and during the COVID-19 pandemic (i.e. in the last 7 days). Average sleep time was calculated using the same formula as average daily SED. Change in sleep represents the difference between average minutes of sleep per night before the COVID-19 pandemic and in the last 7 days and used as a continuous variable in all analysis.

2.2.4. Pet ownership

Participants were asked to report whether they own pet(s). Participants that declared having at least one pet were asked to specify what type of pet(s) (i.e. cat(s), dog(s), bird(s), horse(s), rodent(s) and/or lizards). Since literature suggests that dog owners are more likely to engage in physical activity than owners of other types of animals (Hoffman, 2021, Rhodes et al., 2020, Tan et al., 2021), we categorized pet ownership using three mutually exclusive categories: no pet, at least one dog, other pet-no dog. This categorization strategy has been used in previous studies (Mein and Grant, 2018, Thorpe et al., 2006).

2.2.5. Sociodemographic and clinical profile

Based on a Directed Acyclic Graph (DAG) created from a literature review, sex, age and household composition (living with spouse/partner, not living with spouse/partner), used as a proxy of marital status, were considered potential confounders (Alomari et al., 2020, Applebaum et al., 2020a) of the association between pet ownership and lifestyle behaviours. Age, a variable initially comprising 5 categories, was dichotomized into younger participants (<55 years) and older participants (≥55 years) who were strongly encouraged to comply with the PSD measures. All potential confounders were included in the analyses as categorical variables. To further describe the sample, information on employment status and immunosuppression status (transplant, tissue or stem cell recipient, a family member/relative, donor or immunosuppressed for another reason) was also collected.

2.3. Statistical analyses

Descriptive statistics were first conducted to assess distributions, identify outliers and compute proportions, means and standard deviations. Paired t-tests were conducted to assess changes in lifestyle behaviours before and during COVID-19 for the total sample and according to each pet ownership status and Cohen’s d effect size were computed. Using the Bonferroni correction for multiple testing (n = 20 tests), a p-value of 0.0025 was required for statistical significance. Then, separate linear regression models were fitted to assess the association between pet ownership (a categorical variable including three categories – no pet, other pet(s) – no dog, at least one dog - with no pet as the reference category) and each lifestyle behaviour (continuous variables), adjusting for potential confounders (sex and household composition) and including an interaction term for pet ownership and age. In the presence of a statistically significant interaction, age-stratified linear regressions were used to assess the association between pet ownership and change in lifestyle behaviour in younger (<55 years old) and older (55 years and older) participants. We had a specific hypothesis for each outcome; only one regression model was computed per lifestyle behaviour, then no multiple testing correction was used. Assumptions for linear regression were checked with residuals plots and all were met. Statistical analyses were performed with R Software Version 4.1.2 (R Core Team, 2021). Sensitivity analyses were performed on a subsample of immunosuppressed individuals (excluding donors and relatives). These results are presented in supplementary material (Table S2).

3. Results

A total of 138 participants (65.2% female) completed the online questionnaire. Table 1 shows sociodemographic and clinical characteristics of participants according to their pet-ownership status. The majority of participants were organ-, tissue- and cell-transplant recipients (76.1%). Just over 64% of the participants were pet owners, with 43.5% having at least one dog. There was a greater proportion of participants aged ≥ 55 (59.2%) in the no pet category, compared to participants reporting having other pets and at least one dog (31.0% and 31.7%, respectively). Only 6 participants (4.3%) did not provide complete data on variables of interest and were removed from the analytical sample (n = 132, 65.2% female).

Table 1.

Distribution (n (%)) of COVID-Immuno study participants' characteristics (n = 138) by pet ownership status (Canada, May to August 2020).

Characteristics	Overall n = 138	No pet n = 49	Other pets n = 29	At least one dog n = 60
Sex, female	90 (65.2)	31 (63.3)	18 (62.1)	41 (68.3)
Age, years
< 55	81 (58.7)	20 (40.8)	20 (68.9)	41 (68.3)
≥ 55	57 (41.3)	29 (59.2)	9 (31.0)	19 (31.7)
Household composition
Alone	14 (10.1)	12 (24.5)	1 (3.4)	1 (1.7)
With spouse/partner	65 (47.1)	17 (34.7)	16 (55.2)	32 (53.3)
With spouse/partner and child(ren)	30 (21.7)	8 (16.3)	6 (20.7)	16 (26.7)
With one or several other family members	25 (18.1)	10 (20.4)	4 (13.8)	11 (18.3)
With one or multiple friend(s)/roommate(s)	2 (1.4)	2 (4.1)	0 (0)	0 (0)
Missing	2 (1.4)	0 (0)	2 (6.9)	0 (0)
Employment status
Employed full time	42 (30.4)	9 (18.4)	9 (31.0)	24 (40.0)
Employed part-time	13 (9.4)	3 (6.1)	1 (3.4)	9 (15.0)
Self-employed	11 (8.0)	4 (8.2)	3 (10.3)	4 (6.7)
Unemployed (any reason)	7 (5.1)	3 (6.1)	1 (3.4)	3 (5.0)
Retired	26 (18.8)	12 (24.5)	6 (20.7)	8 (13.3)
On temporary leave	13 (9.4)	7 (14.3)	4 (13.8)	2 (3.3)
Student	12 (8.7)	5 (10.2)	1 (3.4)	6 (10.0)
Other	13 (9.4)	5 (10.2)	4 (13.8)	4 (6.7)
Missing	1 (0.7)	1 (2.0)	0 (0)	0 (0)
Immunosuppression status
A transplant, tissue or stem cell recipient	105 (76.1)	37 (75.5)	23 (79.3)	45 (75.0)
A family member/relative	24 (17.4)	6 (12.2))	3 (10.3)	15 (25.0)
An organ, tissue or stem cell donor	4 (2.9)	2 (4.1)	2 (6.9)	0 (0%)
Immunosuppressed (other reasons)	4 (2.9)	4 (8.2)	0 (0)	0 (0%)
Missing	1 (0.7)	0 (0)	1 (3.4)	0 (0)

Table 2 present results from paired t-tests and effect sizes comparing before- and during-COVID-19 means change in lifestyle behaviours for the total sample and for each group of pet ownership status. A small decrease in VPA (−0.56 days/wk, effect size = 0.34) and large increase in SED (107 min/day, effect size = −0.81) was observed for the total sample. For the no pet and other pets groups, a medium decrease in VPA (−0.94 days/wk, effect size = 0.59 and −0.69 days/wk, effect size = 0.55, respectively) and a large increase in SED (156 min/day, effect size = −1.01 and 103 min/day, effect size = −0.88, respectively) were observed. In the at least one dog group, a medium increased in ST was observed (67 min/day, effect size = −0.65). Similar trends were observed in sensitivity analyses restricted to immunosuppressed individuals (n = 103) (see supplementary material).

Table 2.

Results of paired t-tests and effect sizes comparing movement behaviours before and during the COVID-19 pandemic in the COVID-Immuno study (Canada, May to August 2020).

	Before COVID-19			Past 7 days
	M	SD		M	SD		Change	p	Cohen’s d
All sample (n = 132)
LPA (days/wk)	5.11	2.12		4.67	2.31		−0.43	0.06	0.16
MPA (days/wk)	2.46	2.05		2.02	2.08		−0.44	0.01	0.24
VPA (days/wk)	1.82	1.84		1.26	1.78		−0.56	< 0.01	0.34
SED (min/day)	356.10	232.92		462.89	238.63		106.79	< 0.01	−0.81
Sleep (min/day)	457.92	67.95		456.55	97.10		−1.37	0.85	0.02
No pet (n = 49)
LPA (days/wk)	5.51	2.09		4.33	2.59		−1.18	0.01	0.40
MPA (days/wk)	2.82	2.17		2.10	2.28		−0.71	0.03	0.32
VPA (days/wk)	2.24	1.98		1.31	1.84		−0.94	< 0.01	0.59
SED (min/day)	332.89	207.44		488.13	247.39		155.25	< 0.01	−1.01
Sleep (min/day)	458.41	79.70		451.65	107.44		−6.85	0.57	0.08
Other pets (n = 29)
LPA (days/wk)	4.38	2.26		3.92	2.21		−0.46	0.28	0.21
MPA (days/wk)	2.50	1.75		2.12	1.88		−0.38	0.27	0.22
VPA (days/wk)	1.73	1.80		1.04	1.40		−0.69	< 0.01	0.55
SED (min/day)	359.01	218.29		462.20	218.20		103.19	< 0.01	−0.88
Sleep (min/day)	479.67	60.58		467.56	88.44		−12.11	0.38	0.17
At least one dog (n = 60)
LPA (days/wk)	5.09	2.02		5.32	2.05		0.23	0.47	−0.10
MPA (days/wk)	2.14	2.05		1.91	2.02		−0.23	0.26	0.15
VPA (days/wk)	1.49	1.67		1.32	1.89		−0.17	0.47	0.10
SED (min/day)	374.72	260.62		441.50	241.82		66.78	< 0.01	−0.65
Sleep (min/day)	447.57	58.21		455.81	92.69		8.25	0.46	−0.10

Bold Indicate statistically significant change as suggest by paired t-Test at p < 0.0025.

LPA = Light intensity physical activity, MPA = Moderate intensity physical activity, VPA = Vigorous intensity physical activity, SED = Sedentary Time, M = Mean, SD = Standard deviation, p = p-value.

Comparison of lifestyle behaviours before COVID-19 according to age reveals few differences between younger (< 55) and older (≥ 55) participants. Younger participants reported slightly more LPA (5.19 days/week vs. 4.98 days/week), MPA (2.47 days/week vs. 2.45 days/week) and VPA (2.14 days/week vs. 1.36 days/week) than older participants. No difference in SED and sleep before COVID-19 were found between younger and older participants.

Unstratified linear regression models showed that having at least one dog was associated with each of the lifestyle behaviours compared to having no pets while controlling for age, sex and household composition. Statistically significant interaction term for LPA*age (b = −3.66, SE = 1.17, p < 0.01), MPA*age (b = −2.62, SE = 0.92, p < 0.01, VPA*age (b = −2.19, SE = 0.82, p < 0.01) were identified. Age-stratified regressions are presented in Table 3.

Table 3.

Results of age-stratified linear regressions showing associations between pet ownership status and movement behaviours in the COVID-Immuno study (Canada, May to August 2020).

	Younger participants < 55 years n = 76				Older participants, ≥ 55 years n = 55
	b	(95% CI)	p-value		b	(95% CI)	p-value
LPA change, days/wk
Intercept (no pet)	−2.79	(−4.19, −1.39)	<0.001		−1.41	(−3.27, 0.44)	0.131
Pet (no dog)	0.63	(−1.00, 2.27)	0.442		0.68	(−1.61, 2.97)	0.555
Pet (at least one dog)	2.39	(0.98, 3.80)	0.001		−0.62	(−2.29, 1.05)	0.461
MPA change, days/wk
Intercept (no pet)	−1.80	(−2.90, −0.71)	0.002		−0.21	(−1.45, 1.04)	0.741
Pet (no dog)	1.33	(0.05, 2.61)	0.042		−0.47	(−2.01, 1.08)	0.547
Pet (at least one dog)	1.76	(0.65, 2.86)	0.002		−0.65	(−1.77, 0.48)	0.255
VPA change, days/wk
Intercept (no pet)	−1.82	(−2.88, −0.76)	0.001		−0.76	(−1.74, 0.23)	0.129
Pet (no dog)	1.13	(−0.11, 2.37)	0.072		−0.31	(−1.53, 0.91)	0.611
Pet (at least one dog)	1.92	(0.85, 2.98)	0.001		−0.42	(−1.31, 0.47)	0.344
SED change, min/day
Intercept (no pet)	184.11	(110.26, 257.97)	<0.001		161.95	(74.01, 249.89)	0.001
Pet (no dog)	−57.37	(−143.58, 28.84)	0.189		−36.38	(−145.18, 72.42)	0.505
Pet (at least one dog)	−108.55	(−182.89, −34.21)	0.005		−68.99	(−148.34, 10.36)	0.087
Sleep change, min/wk
Intercept (no pet)	−17.09	(−62.78, 28.60)	0.458		−15.08	(−74.55, 44.40)	0.613
Pet (no dog)	42.21	(−11.13, 95.54)	0.119		−35.32	(−108.90, 38.26)	0.340
Pet (at least one dog)	61.28	(15.28, 107.27)	0.010		−12.28	(−65.95, 41.38)	0.648

LPA = Light intensity physical activity, MPA = Moderate intensity physical activity, VPA = Vigorous intensity physical activity, SED = Sedentary Time, CI = Confidence Interval, b = Unstandardized regression coefficient.

Models adjusted for sex and household composition.

Bold indicates statistically significant results at p < 0.05.

Younger participants (< 55 years old) that have no pets report a decrease of 2.79 days/week in LPA (95 %CI = −4.19, -1.39), 1.80 day/week in MPA (95 %CI = −2.90, −0.71) and 1.82 day/week in VPA (95 %CI = −2.88, -0.76). Comparatively, younger participants with at least one dog report only a small decrease of 0.4 day/week in LPA and 0.04 day/week in MPA, and an increase of 0.1 days/week in VPA. Younger participants that owned other types of pets (e.g. cats, birds) also reported a smaller decrease of 0.4 days/week MPA compared to younger participants that have no pets. Results also show that SED increased by 184 min/day among younger participants who don't have pets (95 %CI = 110.26, 257.97). In comparison, younger participants with at least one dog report a smaller increase of 76 min/day in SED. Younger participants without pet decreased their daily sleep time by 17 min (95 %CI = −62.78, 28.60) while that of younger dog owners increased by 44 min per day. Results show no significant association between pet ownership and changes in lifestyle behaviours in participants 55 years and older. See Fig. 1. In models restricted to immunosuppressed individuals (n = 103), younger participants that owned a pet but no dog reported a small but non statistically significant decrease in MPA (see supplementary material).

Fig. 1.

Line plots and 95% confidence interval error bars of age-stratified linear regressions showing associations between pet ownership and change in light intensity physical activity (A), moderate intensity physical activity (B), vigorous intensity physical activity (C), sedentary time (D) and sleep time (E) in the COVID-Immuno study (Canada, May to August 2020).

4. Discussion

This study assessed changes in lifestyle behaviours in immunosuppressed individuals and their relatives and whether these changes vary according to pet ownership in younger and older adults. When comparing changes in lifestyle behaviours from before to during the COVID-19 pandemic, our results show, based on the effect sizes, a small but significant decrease in VPA and a large increase in SED. These results are consistent with other studies suggesting a decrease in MPA and VPA in the general population (Lesser and Nienhuis, 2020, Wunsch et al., 2022) and in immunosuppressed populations with lupus, rheumatoid arthritis, or cystic fibrosis (Ammitzbøll et al., 2021, Radtke et al., 2021) during the COVID-19 pandemic. In two studies among immunosuppressed individuals, 45–46% of participants reported being less active since the start of the pandemic (Ammitzbøll et al., 2021, Radtke et al., 2021). In both healthy adult populations and populations with chronic disease, SED was reported to increase during the COVID-19 pandemic (Stockwell et al., 2021) similarly to what is observed in the present study. A reduction in LPA and sleep was also observed in the present study, though not statistically significant. The literature shows that for the general population, utilitarian walking has declined significantly, but recreational walking has for some people exceeded pre-pandemic levels (Hunter et al., 2021). In the present study, we used a self-reported measure of LPA that did not distinguish between utilitarian and recreational walking, which could explain why LPA in our sample did not decrease significantly compared to before the pandemic. Finally, participants in our sample reported no significant changes in sleep. Similar results were found in a study on sleep during the pandemic among 6800 people across 59 countries (i.e., 43% of participants reported similar sleep as before the pandemic; (Yuksel et al., 2021)). The authors note that participants who slept 6–8 h per night before the pandemic more often reported no change in sleep time during the pandemic.

In the present study, changes in lifestyle behaviours varied according to pet ownership status; having a pet, and more specifically a dog, was protective against the negative impact of the COVID-19 pandemic on lifestyle behaviours. Based on the effect sizes, results showed that dog owners slightly increased LPA during the pandemic in contrast to the other groups (i.e other pets, no pet) who had a moderate decrease in LPA. However, pet ownership wasn’t associated with changes in MPA nor sleep. These results are similar to those in the literature which indicates that owning a pet would have an effect on the practice of LPA, such as walking, but less on the practice of MPA or VPA for which the results are mixed (Mein and Grant, 2018, Tan et al., 2021, Westgarth et al., 2019). LPA (e.g. walking) is associated with a reduction in obesity (Füzéki et al., 2017), metabolic disease (Füzéki et al., 2017), cardiovascular disease (Dohrn et al., 2018), frailty (Mañas et al., 2018) and better mental health (Sylvester et al., 2017) in adults and older adults, especially if it replaces SED (Piercy et al., 2018). Replacing SED with LPA is more accessible than replacing it with MPA or VPA, especially in older individuals or those with multiple comorbidities (Mañas et al., 2018). Dogs can act as an accessible gateway to physical activity promotion and lifestyle changes.

Although SED increased for all groups during the pandemic, dog owners had a moderate increase compared to the other groups that had a large increase. Prior to the pandemic, Dall and colleagues (Dall et al., 2017) found that dog owners had fewer sedentary bouts – period of time in continuous SED - than non-owners, but no difference was observed in total SED between the two groups. Koohsari and colleagues (2020) observed that SED was lower for dog owners compared with non-owners. The transplanted and immunosuppressed individuals in our sample appear to have benefited from the presence of a dog but this benefit was not observed in owners of other pets (e.g., cats, birds, etc.).

Finally, we examined whether age moderates the associations between pet ownership and lifestyle behaviours. The results show that the protective effect of owning a dog is only present in younger participants (under 55 years of age). Younger participants reported being slightly more active than older participants before the COVID-19 pandemic thus giving them room to decrease their physical activity during the pandemic. Also, unlike younger people, older individuals, usually over the age of 60, were identified by public health authorities as having an increased risk of complications due to COVID-19 and were encouraged to take precautions and follow PSD measures closely (Government of Canada and Canada, 2021). The cumulative vulnerabilities of being immunosuppressed and an older adult might explain that having a dog was not sufficient to protect older immunosuppressed individuals against decline in health lifestyles in the COVID-19 context.

Dog owners highly involved in their pet's care generally have better lifestyle habits (Tan et al., 2021); they walk their dogs more often, which breaks up SED and improve their sleep (Dall et al., 2017, Koohsari et al., 2020, Mein and Grant, 2018). At the height of the COVID-19 pandemic health restrictions, walking with a dog was still permitted, potentially allowing dog owners to maintain LPA (Mein and Grant, 2018, Tan et al., 2021, Westgarth et al., 2019). From a Social Cognitive Theory perspective, the dog can potentially act as a reinforcer to go for a walk and the owner is likely to feel more capable of maintaining this behaviour (Richards et al., 2017) compared to a non-owner. Based on the Health Belief Model (Janz and Becker, 1984), the dog could also act as a prompt (i.e. cue to action) and help the owner maintain an active behaviour in the context of COVID-19.

4.1. Study Limitations and strengths

This study is one of the very few that examine lifestyle behaviours and pet ownership in immunosuppressed individuals. We used a cross-sectional design, which doesn’t allow us to conclude causal associations or directionality of the associations observed. However, our questionnaire was administered early at the beginning of the pandemic to capture subtle changes in lifestyle behaviour following the abrupt imposition of PSD measures. Limitations include self-reported and retrospective data which could introduce recall bias and may result in misclassification of lifestyle behaviours. Moreover, the IPAQ is known to overestimate physical activity and underestimate SED (Crutzen and Göritz, 2011, Ryan et al., 2018). The small sample size may also limit the ability to detect statistically significant associations.

5. Conclusions

Pet ownership appears to alleviate the negative impacts of the COVID-19 pandemic on lifestyle behaviours in some immunosuppressed individuals. Pet ownership may act as a catalyst for clinical and public health initiatives to preserve and promote healthy behaviours in these populations. This study focused primarily on the impact of the COVID-19 pandemic on lifestyle behaviours, but there are likely other stressful life events that influence movement behaviours among immunosuppressed individuals and their relatives. Future research could explore other factors such as social support, and access to parks and safe built environments, and examine how they interact with pet ownership and other individual-level factors to influence health behaviours.

Ethical approval

This study is part of the Projet Laurent and has received ethics approval from the Ethics Research Committee of the Centre de Recherche du Centre Hospitalier de l’Université de Montréal (#20.018).

Informed consent

All participants provided written informed consent.

Funding

This work was supported by the Université de Montréal Appui aux initiatives intersectorielles program (grant numbers 2020AII-009, 2020) and Boehringer Ingelheim (Canada). The University of Montreal and Boehringer had no involvement in the conduct of the study or in the preparation of the manuscript.

CRediT authorship contribution statement

Audrey Plante Plante: Formal analysis, Writing – original draft, Visualization. Nathalie Bedrossian: Writing – review & editing. Gabrielle Cadotte: Writing – review & editing. Alexia Piché: Writing – review & editing. Fady Michael: Formal analysis, Writing – review & editing. Sylvain Bédard: . Hélène Tessier: . Christopher Fernandez-Prada: Writing – review & editing. Catherine M. Sabiston: Writing – review & editing. Mélanie Dieudé: Conceptualization, Methodology, Writing – review & editing. Isabelle Doré: Conceptualization, Methodology, Supervision, Writing – review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article:

Data availability

Data will be made available on request.