Changes in physical activity and sedentary behaviours of children living in Italy due to the COVID-19 pandemic: the EPaS-ISS study

Abstract

Background

Most governments worldwide introduced containment measures to protect people from the risk of SARS-CoV-2 virus infection which also affected children. The COVID-19 pandemic and the resulting confinement at home may have affected physical health and mental well-being especially among vulnerable young people, widening health disparities. Therefore, the aim of the EPaSS-ISS study was to study the changes in physical activity and sedentary behaviours among 8–9 years old Italian students during the pandemic of COVID-19 compared with before the pandemic.

Methods

Data were collected using a web questionnaire. The target population was parents of children attending third-grade primary schools and living in Italy. A cluster sample design was used. Logistic regression models were used to explore changes between the pre-pandemic and pandemic periods in children’s studying hours, active play and screen time. The possible effects of gender, geographical area of residence, parents’ nationality and educational level, family structure, family-perceived wealth, availability of home outdoor spaces and house size were investigated. A stratification by sociodemographic characteristics of children and parents was also included in the analysis.

Results

4863 children were included: 47.9% were females, the average age was 8 years and 9 months (SD ± 5 months). The parents reported that 44.3% of their children increased the time spent studying at home, 52.7% the screen time for non-educational purposes on weekdays, and 47.3% at weekends; 44% of the children decreased their time spent playing actively outside school hours on weekdays and 39.3% at weekends. A larger increase of sedentary lifestyle was found among children with at least one parent with a medium (OR:2.57; 95%CI:1.78–3.72) or high educational level (OR:3.32, 95%CI:2.29–4.82), with low family-perceived wealth (OR:1.21; 95%CI:1.03–1.41), with both Italian parents (OR:1.44; 95%CI:1.10–1.90), resident in Centre (OR:1.25; 95%CI:1.04–1.50) or South (OR:1.26; 95%CI:1.05–1.51) Italy. No outdoor spaces (OR:2.14; 95%CI:1.82–2.51) and home size < = 90m2 (OR:1.41; 95%CI:1.05–1.68) had also a negative effect on children’s behaviours.

Conclusions

In Italy physical activity decreased and sedentary behaviours increased among 8–9 years old children during the COVID 19 pandemic.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13690-025-01701-5.

Text box 1. Contributions to the literature
• The EPaS-ISS study offers the first national-level data from Italy examining how children’s physical activity and screen time habits changed during the COVID-19 pandemic, providing a snapshot of the pandemic impact on children’s daily routines.
• The study enabled to compare the Italian results with those from other 16 countries of the WHO European region.
•Low levels of active play and sport activities along with high levels of screen use among children was observed by the Italian COSI “OKkio alla SALUTE” surveillance system also before the pandemic. The EPaS-ISS study highlighted a worsening in these behaviours during the COVID-19 pandemic.
• The produced knowledge supports the importance to develop focused strategies to improve health-related behaviours among children both in everyday life and during health emergencies.

Background

Over the past few years, the COVID-19 pandemic has dramatically affected individual behaviours and public health systems across the globe. Following the declaration of the pandemic by the World Health Organization (WHO) on 11 March 2020 [1], many governments worldwide enacted different containment measures (school closures, physical distancing, lockdown and quarantine at home) to reduce the risk of SARS-CoV-2 virus infection.

Although these interventions were critical for infection control, children were not indifferent to the dramatic impact of the COVID-19 pandemic. According to The United Nations Educational, Scientific and Cultural Organization [2], in early April 2020 more than 160 countries closed their schools to over 90% of the world’s student population (1.5 billion young people). In Italy, schools were closed in February/March 2020; all students continued their educational activities at home until the end of the academic year. Later, schools were closed to students and school personnel positive for the SARS-CoV-2 virus and distance learning was adopted. The physical distancing measures changed the daily routines of children undermining their education, physical activity and opportunities to socialize, with long-term consequences [3]. The COVID-19 pandemic and the resulting confinement at home may have affected physical health and mental well-being especially among vulnerable young people widening health disparities [4].

An adequate amount of physical activity and a low level of sedentary time are associated with health benefits such as the reduction of non-communicable diseases (NCDs) (i.e., cardiovascular disease, diabetes, obesity and cancer) and NCD risk factors, such as hypertension, coronary heart disease and overweight [5]. An adequate level of physical activity also positively affects mental health, quality of life and well-being [6]. The WHO and other international organisations have emphasised the importance of increasing or maintaining engagement in physical activity and reducing sedentary time during the COVID-19 pandemic [7, 8]. Despite these recommendations, several studies have shown that COVID-19 led to a decrease in movement, physical activity and outdoor activities [9–12]. Some factors were positively associated with the maintenance of an adequate level of movement activities during the pandemic period, e.g. being younger, having a big outdoor space, parental support and encouragement, and having at least one adult free from working at home [13–17].

Furthermore, an increase in sedentary behaviours, in particular screen time activities, changed in children’s life during the COVID-19 pandemic. Screen use may have provided opportunities for children’s educational and social experiences during the pandemic; in contrast, its marked increase suggested the importance to investigate the amount and quality of time that children spent in this activity outside school hours [18]. Parents’ support, high level of physical activity and screen-free recreational activities might have played an important role in a positive use of screens among young people [18, 19]. High level of screen-based behaviours among young people were also observed before the pandemic [20] and factors such as children’s age, socioeconomic status of families, caregivers’ attitude to screen use were noticed to be associated with young people screen time both before and after the pandemic [21].

In Italy, a web survey was carried out in the framework of the Project EPaS-ISS “Effects of COVID-19 pandemic on health behaviour and lifestyle of children and their families living in Italy” which was promoted and funded by the Italian National Institute of Health (Istituto Superiore di Sanità, ISS, Rome, Italy) and coordinated by its National Centre for Disease Prevention and Health Promotion (Centro Nazionale per la Prevenzione delle Malattie e la Promozione della Salute, CNaPPS, Rome, Italy) [22]. The study is part of the WHO European Childhood Obesity Surveillance Initiative (COSI) and was conducted in 17 countries [23].

It is expected that during the COVID-19 pandemic, levels of physical activity will significantly decrease, and sedentary behaviours will increase, especially in vulnerable family contexts. Furthermore, we speculated a greater decline in physical activities during the pandemic among children living in households lacking outdoor areas, in smaller homes, and in families facing economic difficulties.

This article presents the results for Italy on the changes between the pre-pandemic period (before February/March 2020) and the pandemic period (from February/March 2020 to April 2022 based on the periods of suspension of face-to-face school activities) in some children’s habits that might have increased their sedentary activities: time spent studying at home (more than 3 h per day, including hours of online schooling), time spent playing actively/energetically outside school hours (on weekdays and at weekends) and time spent watching TV, playing videogames or using social media for non-educational purposes (on weekdays and at weekends). Concerning the exploration of health inequalities, data from the Italian COSI “OKkio alla SALUTE” highlights the important role of social determinants in children’s lifestyles [24]. During the COVID-19 pandemic, these determinants might have limited children’s opportunities to maintain an adequate level of physical activity and increased screen time activities. Moreover, recent evidence has highlighted the persistence of sedentary behaviours even after the end of restrictive measures, with children showing reduced spontaneous physical activity and increased dependence on structured [25, 26]. Consequently, stratification by sociodemographic characteristics such as children’s gender, families’ area of residence, parents’ educational level and nationality, family structure and family-perceived wealth was also included in the analysis to identify possible differential impacts of the pandemic on children’s behaviours.

Methods

Data collection

To gather information, a digital questionnaire was administered to parents of children enrolled in the third grade of primary school (typically aged 8–9 years). The study was conducted in accordance with a standardized protocol jointly developed by the WHO and countries participating in the surveillance initiative.

The sample size was determined to guarantee that the margin of error does not exceed 0.03 when estimating a proportion of 0.50, using a 95% confidence level. This estimate was based on a design effect (Deff) of 2 and an anticipated response rate of at least 50% among invited participants. The sampling strategy was designed to produce reliable national estimates, with proportional allocation across Italian regions and autonomous provinces (AP) according to the number of third-grade pupils.

As recommended by the COSI framework, a two-stage stratified cluster sampling design was employed. Schools were designated as the primary sampling units and classes as the secondary units [22, 23]. The list of schools used for sampling was drawn from those involved in the 2019 round of the national COSI survey “OKkio alla SALUTE” [24, 27].

Out of the 19 Italian regions, 17 participated in the study along with the two AP. Activities at the local level were coordinated and supported by regional and local health unit (LHU) coordinators. Regional and LHU coordinators, with support from provincial school offices, presented the EPaS-ISS to schools. Parents of participating students received study information and the survey link through school staff, including head teachers and classroom teachers. The web questionnaire was developed using LimeSurvey software (version 6.1.7), and its layout was optimized for different devices (e.g., smartphones, tablets, or computers).

All parents of the selected classes (typically two or three classes per school) were invited to participate. Data collection occurred from early April 2022 to the end of September 2022. Additional methodological details are available in previous publications [28, 29]. Parents were asked to report any changes in their children’s behaviours during the COVID-19 pandemic period (defined from February/March 2020 to April 2022, coinciding with school closures) compared to the pre-pandemic period (prior to February/March 2020). Information about children behaviours included time spent studying at home (over 3 h/day including online learning), playing actively or energetically outside school hours (during weekdays and weekends), and engaging in screen-based activities such as watching TV, playing video games, or using social media for non-school purposes. Table 1 reported these questionnaire items. The full list of items used in the analysis is presented in Table A1 (Additional file 1).

Table 1.

Items from questionnaire about changes in children’s behaviours (studying hours, movement, screen time activities)

Question	Response
Please indicate whether any of the child’s habits have changed during the COVID-19 period compared to the PRE-COVID period:	Decreased, Stayed the same, Increased, Don’t Know
1. More than 3 h per day of time spent studying at home (including hours of online schooling) 2. Time spent playing actively/energetically (e.g. running or jumping outdoors or moving and playing games indoors) outside school hours, on weekdays 3. Time spent playing actively/energetically (e.g. running or jumping outdoors or moving and playing games indoors) outside school hours, at weekends 4. Time spent watching TV, playing videogames or using social media for non-educational purposes on weekdays 5. Time spent watching TV, playing videogames or using social media for non-educational purposes at weekends.

Information on the sociodemographic background of the families was also obtained through the online questionnaire. The variables included in the study were: child’s gender (male or female), families’ area of residence (North, Centre, or South of Italy), and parents’ nationality (classified as both parents Italian or at least one parent of foreign origin). Educational attainment was recorded for both parents, and the highest level achieved was used to define three categories: “low” (neither parent completed high school), “medium” (at least one parent completed high school), and “high” (at least one parent held a university degree or higher qualification, including master’s, doctoral, or specialized degrees). Family structure was categorized as following: two-parent household and single-parent household. Economic situation perceived by family during the COVID-19 period was reported as very easy, quite easy, slightly hard, very hard, or “don’t know”. Another item investigated the availability of outdoor spaces (such as gardens, balconies, or terraces) at the home where the child spent most of their time during the pandemic. Families were categorized as having or not having access to outdoor areas (grouping small non-habitable spaces and the complete absence of outdoor areas together under “not available”). Lastly, the size of the dwelling, expressed in square meters, was categorized as ≤ 90 m², 91–150 m², or > 150 m².

Statistical analysis

Descriptive statistics, including frequencies and 95% confidence intervals (CIs), were used to summarize outcomes overall and by sociodemographic characteristics of children and families. Pearson’s design-based χ² test was applied to examine differences in the prevalence of outcomes across demographic subgroups.

Logistic regression analyses were performed to evaluate associations between various independent variables—namely, gender, geographical area of residence, parental nationality and educational attainment, family structure, perceived economic hardship, availability of outdoor spaces, and home size and changes in selected behaviours, including time spent studying at home, weekday physical activity, and weekday screen use. Each dependent variable was dichotomized based on directional change during the pandemic: either “increased vs. remained the same or decreased” (regarding the duration of studying and screen time; Models 1 and 3) or “decreased vs. remained the same or increased” (regarding active play on weekdays; Model 2).

Considering sedentary lifestyle as a multidimensional construct encompassing both inadequate time spent practicing sports and actively playing, as well as high sedentary habits such as screen-based activities, the WHO recommended children to enhance physical activity and decrease sedentary behaviours [5]. The COVID-19 pandemic limited the opportunity for sports and movement, increasing use of screen devices for more free time at home and impossibility to join outdoor activities. For these reasons, we considered the combination of two dimensions (reduction in physical activity and increase in screen device use during the COVID-19) as a proxy of unhealthy behaviour. Variables for the logistic regression model (MODEL 4) was defined as follows:

• sociodemographic characteristics: independent variables.

• sedentary lifestyle (which included decrease in movement activities together with the increase in screen time activities, both on weekdays): (dependent variable) considering for each response variable the increase vs. stayed the same or decreased (yes/no) of sedentary lifestyle.

To strengthen the statistical robustness of our analysis, we incorporated global adjustment measures by applying the Hosmer-Lemeshow test to evaluate the goodness-of-fit of the four logistic regression models. Results are expressed as odds ratios (ORs) with 95% confidence intervals (CI). Missing data and “Don’t know” responses were excluded from the analysis. For variables with higher non-response rates (e.g. perceived wealth), a complete-case approach without imputation was adopted, in line with the descriptive and exploratory nature of the research. Sampling weights were applied to adjust for oversampling and nonresponse. All analyses took into account the clustered and stratified structure of the data. The statistical analyses were performed by means of STATA software (version 15, StataCorp, College Station, TX, USA). A p-value of less than 0.05 was considered statistically significant.

Results

Characteristics of children and parents

Around 5,900 parents gave their consent to participate in the survey, equal to 46.6% of those that were invited. Children with parents who answered only the initial few questions were excluded from the analysis. A total of 4,863 children were included: 47.9% were females, the average age was 8 years and 9 months (SD ± 5 months) and they were geographically distributed as follows: North: 47.5%; Centre: 24.4%; South: 28.1%. This geographic distribution is like that reported by the Italian National Institute of Statistics (ISTAT) for Italian 8 years-old children [30]. Overall, the percentage of questions with missing responses ranged from 0 to 3.5%, with the exceptions of questions about parents’ educational level (mother educational level: 6.9%; father educational level: 8.1%), parents’ nationality (mother nationality: 6.6%; father nationality: 6.8%) and family-perceived wealth during the COVID-19 pandemic (17.1%).

The characteristics of parents and families are reported in Table 2.

Table 2.

General characteristics of parents and families at the interview

Sociodemographic variables		Cases	%	IC 95%
Mother educational level
	None	19	0.4%	0.25–0.66
	Primary school (6 to 11 years)	22	0.5%	0.31–0.74
	Secondary school (from 11 to 14 years)	541	12.0%	11.09–13.01
	High school (from 14 to 19 years)	2,110	46.9%	45.44–48.38
	Degree	1,430	31.8%	30.43–33.18
	Master/Doctorate/Specialisation	376	8.4%	7.57–9.21
Mother nationality
	Italian	4,066	90.0%	89.12–90.89
	Foreign	441	9.8%	8.91–10.67
Mother not present		9	0.2%	0.09–0.38
Father educational level
	None	22	0.5%	0.31–0.75
	Primary school (6 to 11 years)	52	1.2%	0.88–1.53
	Secondary school (from 11 to 14 years)	986	22.2%	20.98–23.45
	High school (from 14 to 19 years)	2,238	50.4%	48.90-51.86
	Degree	931	21.0%	19.77–22.19
	Master/Doctorate/Specialisation	213	4.8%	4.19–5.47
Father nationality
	Italian	4,122	91.5%	90.65–92.30
	Foreign	331	7.3%	6.60-8-15
Father not present		52	1.1%	0.09–1.51
Family-perceived wealth
	Very easy/quite easy	2,027	50.6%	49.0-52.2
	Slightly hard/very hard	1,655	41.3%	39.8–42.9
	Don’t know	324	8.1%	7.3-9.0
Availability of home outdoor spaces
	Large habitable spaces	3,287	70.5%	69.2–71.8
	Small non-habitable spaces	1,005	21.5%	20.4–22.8
	No outdoor spaces	371	8.0%	7.2–8.8
House size
	> 150 m2	670	14.7%	13.6–15.7
	91–150 m2	2,240	48.9%	47.5–50.4
	<=90 m2	1,668	36.4%	35.0-37.8

The majority of parents were Italian. Considering the highest level of education between mother and father, most of the parents reported having a medium or high educational level. Regarding the perception of family wealth, one-half of the families claimed to easily pass the month with their earnings, and 41.3% reported to have difficulties making ends meet each month. A not negligible number of “Don’t know” responses (8.1%) was observed for the question on family-perceived wealth. Concerning outdoor spaces (e.g. garden, balcony, terrace) of home in which children spent most of their time during the COVID-19 pandemic, 70.5% of parents declared living in a house with large habitable space, 21.5% with small non-habitable space and 8.0% without outdoor space. Almost half of the families lived in 91–150 m² house during the pandemic, followed by families reported to live in < = 90 m² house (36.4%).

Change in children’s behaviours during COVID-19 pandemic

Table 3 shows the changes between the pre-pandemic period and the pandemic period in some children’s habits: time spent studying at home (more than 3 h per day, including hours of online schooling), time spent playing actively/energetically outside school hours (on weekdays and at weekends) and time spent watching TV, playing videogames or using social media for non-educational purposes (on weekdays and at weekends).

Table 3.

Change in children’s behaviours between the pre-COVID-19 period and the COVID-19 period

Change in child behaviours	Decreased n = 469	Stayed the same n = 1,937	Increased n = 2,117	Don’t Know n = 246
	(%, CI 95%)	(%, CI 95%)	(%, CI 95%)	(%, CI 95%)
More than 3 h per day of time spent studying at home (including hours of online schooling)	9.7 (8.9–10.7)	40.9 (39.3–42.5)	44.3 (42.5–46.1)	5.1 (4.4–5.8)
Time spent playing actively/energetically (e.g. running or jumping outdoors or moving and playing games indoors) outside school hours, on weekdays	44.0 (42-45.9)	35.9 (34.4–37.5)	17.8 (16.6–19.1)	2.3 (1.9–2.8)
Time spent playing actively/energetically (e.g. running or jumping outdoors or moving and playing games indoors) outside school hours, at weekends	39.3 (37.4–41.2)	42.5 (40.9–44.2)	15.7 (14.5–16.9)	2.5 (2.0-3.1)
Time spent watching TV, playing videogames or using social media for non-educational purposes on weekdays	5.4 (4.8–6.2)	39.6 (38.1–41.2)	52.7 (51.1–54.3)	2.3 (1.8–2.8)
Time spent watching TV, playing videogames or using social media for non-educational purposes at weekends.	6.0 (5.3–6.7)	44.1 (42.6–45.6)	47.3 (45.7–48.9)	2.7 (2.2–3.2)

These changes stratified by children’s gender, geographical area of residence, parents’ nationality and educational level, family structure, family-perceived wealth, availability of home outdoor spaces and house size are reported in Table A2 (Additional file 2).

An increase in time spent studying at home was reported by 44.3% of parents. A greater increase was revealed among children from Northern regions (North:47.2%; Centre: 42.4%; South: 41.7%), with at least one parent with medium and high educational level (low: 25.0%; medium: 41.7%; high: 52.5%), with both Italian parents (both Italian parents: 46.0%; at least one foreign parent: 39.0%), living in two-parent families (two-parent family: 46.1%; single-parent family: 40.4%) and children from families claimed to easily pass the month with their earnings (very and quite easy: 50.6%; slightly and very hard: 44.3%). A large-small house size increasing gradient in time spent studying at home was also observed ( < = 90 m²: 42.9%; 91–150 m²: 46.0%; >150 m²: 48.3%).

The time spent playing actively/energetically by children outside school hours decreased during both weekdays (44.0%) and weekends (39.3%). Stratified data showed differences according to geographical area of residence, parents’ educational level and nationality. Higher percentages of decrease were observed in Southern and Central regions both for weekdays (North: 41.1%; Centre: 46.3%; South: 46.3%) and weekends (North: 35.7%; Centre: 39.5%; South: 43.7%). A pronounced decreasing low-high parents’ educational level trend was observed for weekdays (low: 23.6%; medium: 43.4%; high: 49.6%) and weekends (low: 22.6%; medium: 39.9%; high: 42.7%). Moreover, children with both parents Italian decreased the time spent playing actively/energetically more than those with at least one foreign parent on weekdays (both Italian parents: 45.6%; at least one foreign parent: 37.4%) as well as at weekends (both Italian parents: 40.7%; at least one foreign parent: 33.2%). A greater decrease in the time spent playing actively/energetically occurred among children from families reporting difficulty to pass the month with their earnings, both on weekdays (very and quite easy: 45.1%; slightly and very hard: 47.8%) and at weekends (very and quite easy: 38.9%; slightly and very hard: 43.1%). Regarding home characteristics, both availability of home outdoor spaces and home size affected the decrease in movement behaviour among children; a greater decrease was found among children living in a small/medium home ( < = 90 m²: 46.5%; 91–150 m²: 46.3%; >150 m²: 35.1%, on weekdays) and without outdoor spaces (outdoor spaces: 38.8%, no outdoor spaces: 57.9%, on weekdays). Comparable results were also observed for the time spent actively playing at weekends.

Parents reported a marked increase in time spent by children watching TV, playing videogames or using social media for non-educational purposes on weekdays (52.7%) and weekends (47.3%). A greater increase was noticed in the Centre on weekdays (North: 51.2%; Centre: 57.5%; South: 51.8%) and in the Centre and South at weekends (North: 44.9%; Centre: 49.6%; South: 49.0%). After stratification by educational level, the data showed a low-high parents’ educational level increasing gradient for both weekdays (low: 36.0%; medium: 51.9%; high: 58.3%) and weekends (low: 33.1%; medium: 48.3%; high: 50.1%). Marked differences were also found for parents’ nationality; children with both Italian parents during the COVID-19 period spent more time watching TV, playing videogames or using social media than those with at least one foreign parent and this increase was observed on weekdays (both Italian parents: 54.1%; at least one foreign parent: 47.4%) and at weekends (both Italian parents: 48.4%; at least one foreign parent: 43.7%). Concerning family-perceived wealth, greater increases in screen time activities were observed among children from families reporting difficulty to pass the month with their earnings, both on weekdays (very and quite easy: 54.5%; slightly and very hard: 57.8%) as well as at weekends (very and quite easy: 47.8%; slightly and very hard: 53.2%). Stratification by house size did not show differences in children’s screen time activities on weekdays or at weekends. In contrast, the availability of home outdoor spaces highlighted pronounced differences on the increase in screen time activities on weekdays (outdoor spaces: 49.9%, no outdoor spaces: 62.0%) and at weekends (outdoor spaces: 43.8%, no outdoor spaces: 57.1%) during the COVID-19 pandemic.

Multivariate analysis

The results of the logistic regression models on changes in children’s studying hours, active playing and screen time activities on weekdays between pre-COVID-19 period and COVID-19 period are shown in Table 4.

Table 4.

Logistic regression models for the change in children’s behaviours between PRE-COVID-19 period and COVID-19 period

Indipendent variables	MODEL 1 Daily hours spent studying Increased (yes/no)	MODEL 2 Time spent in active play Decreased (yes/no)	MODEL 3 Time spent watching TV, playing video games or using social media Increased (yes/no)	MODEL 4 Sedentary lifestyle Increased (yes/no)
	OR* (95% CI)	OR* (95% CI)	OR* (95% CI)	OR* (95% CI)
Child’s Gender
Male	1	1	1	1
Female	1.10 (0.96–1.26)	0.96 (0.83–1.11)	0.86 (0.75–0.99)	0.89 (0.77–1.03)
Residence area
North	1	1	1	1
Centre	0.79 (0.66–0.94)	1.14 (0.95–1.36)	1.26 (1.05–1.50)	1.25 (1.04–1.50)
South	0.90 (0.76–1.07)	1.36 (1.14–1.62)	1.00 (0.84.1.19)	1.26 (1.05–1.51)
Parents’ educational level**
Low level of education	1	1	1	1
Medium level of education	1.79 (1.30–2.44)	2.46 (1.78–2.41)	1.61 (1.22–2.17)	2.57 (1.78–3.72)
High level of education	2.55 (1.86–3.51)	3.20 (2.30–4.46)	2.04 (1.51–2.76)	3.32 (2.29–4.82)
Parents’ nationality
At least one foreign parent	1	1	1	1
Both Italians	1.10 (0.86–1.41)	1.46 (1.13–1.89)	1.22 (0.95–1.57)	1.44 (1.10–1.90)
Family structure
One-parent family	1	1	1	1
Two-parent family	1.28 (1.04–1.56)	0.93 (0.76–1.15)	1.17 (0.96–1.43)	1.03 (0.83–1.27)
Family-perceived wealth
Good/quite good	1
Critical/somewhat critical	0.92 (0.79–1.07)	1.18 (1.01–1.37)	1.22 (1.05–1.42)	1.21 (1.03–1.41)
Availability of outdoor spaces
Yes	1
No	1.08 (0.92–1.27)	2.23 (1.90–2.62)	1.77 (1.51–2.09)	2.14 (1.82–2.51)
Home size
> 150 m2	1
91–150 m2	0.96 (0.78–1.19)	1.62 (1.30–2.03)	0.94 (0.76–1.16)	1.32 (1.10–1.82)
<=90 m2	0.91 (0.72–1.14)	1.59 (1.25–2.02)	0.97 (0.77–1.22)	1.41 (1.05–1.68)

*Statistically significant results are in bold

 *Adjusted Odds Ratio for all variables list in the Table 4

***The highest educational level between the two parents

Figure 1 provides a visual summary of the regression models, facilitating interpretation of the strength and direction of associations between sociodemographic factors and changes in children’s behaviours.

Fig. 1.

Forest plot for the logistic regression models (1, 2, 3, 4)

The Hosmer–Lemeshow test results indicated no significant lack of fit for any of the models (p > 0.05), suggesting that the predicted probabilities are consistent with the observed outcomes.

MODEL 1 yielded a χ²(8) = 14.27 with a p-value of 0.075, which is close to the 0.05 threshold, indicating a marginal but acceptable fit. In contrast, MODEL 2, 3, and 4 showed stronger evidence of good fit, with higher p-values (0.564, 0.832, and 0.449, respectively). These results support the adequacy of the models in capturing the relationship between predictors and the outcome.

The findings from the multivariate analysis were consistent with those of the univariate analyses, although in some instances the adjusted odds ratios did not reach statistical significance.

In detail, a higher increase in the time spent studying at home (more than 3 h per day, including hours of online schooling) (MODEL 1) was associated with the parents’ educational level (medium, OR: 1.79; 95% CI: 1.30–2.44; high, OR: 2.55; 95% CI: 1.86–3.51). A statistically significant odds ratio increase was also found for children living in a two-parent family (OR: 1.28; 95% CI: 1.04–1.56).

For the time spent active playing, the logistic regression model (MODEL 2) confirmed each significant result of the univariate analysis. Statistically significant odds ratios for decreases in the time spent playing actively/energetically outside school hours on weekdays were observed among children from southern regions in comparison with those living in northern regions (OR: 1.36; 95% CI: 1.14–1.62). Families with at least one parent with a medium or high educational level were more likely to have decreases in children active playing during the pandemic (medium, OR: 2.46; 95% CI: 1.78–2.41; high, OR: 3.20; 95% CI: 2.30–4.46). Furthermore, major changes in the time spent in active play were associated with the nationality of the parents and family-perceived wealth with a higher decrease in families where both parents were Italian (OR: 1.46; 95% CI: 1.13–1.89) and those reporting to hardly pass the month with their earnings (OR: 1.18; 95% CI: 1.01–1.37). As previously shown in the univariate analyses, small/medium home (91–150 m², OR: 1.62; 95% CI: 1.30–2.03; <=90 m², OR: 1.59; 95% CI: 1.25–2.02) and those without outdoor spaces (no outdoor spaces, OR: 2.23; 95% CI: 1.90–2.62) had statistically significant effects on the decrease in the children’s active play.

Living in the centre of Italy had a statistically significant effect on the increase in screen time activity (OR: 1.26; 95% CI: 1.05–1.50), confirming the results of univariate analysis (MODEL 3). Families with at least one parent with a medium or high educational level (medium, OR: 1.61; 95% CI: 1.22–2.17; high, OR: 2.04; 95% CI: 1.51–2.76) and those reporting to hardly pass the month with their earnings (slightly and very hard, OR: 1.22; 95% CI: 1.05–1.42) were more likely to have increases in children’s screen activities. Moreover, a higher increase in the time spent watching TV, playing videogames or using social media was associated with no availability of home outdoor spaces (no outdoor spaces, OR: 1.77; 95% CI: 1.51–2.09).

The logistic regression model on the increase in sedentary lifestyle (decrease in actively/energetically playing and increase in screen time activities during the COVID-19 pandemic) (MODEL 4) showed that the majority of the adjusted odds ratios were statistically significant except for children’s gender and family structure, confirming the results of the other models. A medium and high parent education level was noticeably associated with a major increase in sedentary behaviours during the pandemic (medium, OR: 2.57; 95% CI: 1.78–3.72; high, OR: 3.32; 95% CI: 2.29–4.82).

Discussion

This study examines the changes between the pre-pandemic COVID period and the pandemic period in 8–9 years-old children’s behaviours living in Italy, in particular time spent studying at home, time spent playing actively/energetically outside school hours and time spent watching TV, playing videogames or using social media for non-educational purposes. The analysis was also stratified by sociodemographic characteristics of children and parents to identify possible different impacts.

The study shows evidence of a negative effect of the COVID-19 pandemic on all considered behaviours for about 40% of the children. In particular, the EPaS-ISS results noticed that the pandemic had a more pronounced effect on children’s use of electronic devices (52.7%, on weekdays) than movement activities (44%, on weekdays).

The results concur with the literature regarding the impact of the pandemic on children’s sedentary behaviours. Rossi et al. (2021) found clear evidence in the COVD-19 literature for decreased physical activity during the lockdown, especially concerning organized sports [31]. Similarly, several reviews published in 2021 reported a decrease in physical activities, especially for organized sports, and an increase in sedentary behaviours and screen time activities during the first year of the pandemic [32–34]. Two systematic reviews published in 2022 confirmed the previous findings. They strengthened the evidence for the reduction of organized physical activity [35] and the increase in daily hours that children spent in screen time activities [36]. Kharel et al.’s review added that children facing stricter lockdowns tended to have a greater impact on these daily behaviours [36].

Changes in the level of daily physical activities during the COVID-19 pandemic were measured through two systematic reviews with meta-analysis. In detail, Neville et al. (2022) found that their level decreased by 20% among children and adolescents, irrespective of pre-pandemic baseline levels, and Pang and colleagues (2023) reported a physical activity reduction of 18.8 min per day for preschoolers, children, and adolescents [37, 38].

To our knowledge, this is the largest Italian study that investigated the impact of the COVID-19 pandemic on children’s sedentary habits, but previous studies on smaller Italian samples were in line with our results [39–41]. Concerning stratification by sociodemographic characteristics of children and parents, our results did not reveal different changes in children’s active play and screen time by sex, suggesting that other characteristics may be more relevant. In the literature, it is not clear how sex influenced the level of physical activity during the pandemic [33, 35]. This lack of clarity might be due to variations linked to age and the specific type of physical activity being studied [42]. On the contrary, our findings showed different impacts of the pandemic on children’s sedentary behaviours depending on parental factors (e.g. parents’ nationality and educational level, geographical area of residence, family structure, family-perceived wealth) as well as home environments (availability of outdoor spaces, home size).

Parental factors

A larger impact was found among children with at least one parent with a medium or high educational level, who reduced time spent in active play and increased screen time and sedentary lifestyle significantly more than children of parents with lower educational level. Higher socioeconomic status is generally linked to better physical activity, particularly in organized sports [24], which were forbidden by governmental law during the pandemic period. Probably, time previously spent on organized sports was not converted into unstructured physical activities such as active play but into sedentary behaviours. In contrast, other authors found inconsistent results on the effect of parental education on young people sedentary habits [32, 33]. One possible explanation is that families with higher education levels were more engaged in structured sports and extracurricular activities, which were largely suspended during the pandemic. Without adequate alternatives, this may have led to an increase in screen time rather than in unstructured play.

Active play, screen time and sedentary lifestyle were also associated with critical family-perceived wealth. A possible reason is that children of anxious parents have a greater reduction in physical activity. Our previous findings showed an increase in the number of families reporting to hardly pass the month with their earnings during the pandemic in comparison to the pre-pandemic period [29]; the deterioration in financial situation of families living in Italy might have increased the level of parents’ fear, anxiety and psychological distress with a negative impact on children’s daily healthy activities such as active play [43]. Moreover, the combination of remote working and heightened academic expectations often observed in highly educated households might have further limited children’s opportunities for unstructured physical activity.

Children of parents resident in northern Italy have increased their daily hours spent studying at home more than families in central and southern Italy. Children living in central and southern Italy have shown a larger increase in sedentary lifestyle compared to children living in the north. Interpreting these findings is difficult due to the variability of the environment within each Italian area, which includes both urban and rural areas. Other studies have found that youths in urban or densely populated areas exhibited larger decreases in physical activity [32, 33, 44].

It should also be noted that regional variation in local policies (e.g., school closures, restrictions on movement) might have modulated children’s behaviours during the pandemic. These differences, though not assessed in detail in the present study, might partly explain geographical disparities observed in physical activity and screen time.

Home environment

Physical activity decreased especially in families with no availability of outdoor spaces. This finding is strongly supported by numerous other studies highlighting that children and youths who had access to outdoor spaces and were consistently allowed to use them during the COVID-19 reported higher levels of physical activity and experienced a lower decline [16, 33, 37, 44, 45]. We found that a home size smaller than 150 square meters was a risk factor for decreasing active play and increasing sedentary lifestyle. Yomoda and Kurita (2021) [32] also found that physical activity was less likely to decrease in children who lived in detached houses, houses with a yard, garage, or larger spaces.

Our findings suggest that governments should consider the adverse effects of restrictive measures on children’s physical activity and take action to promote high levels of physical activity. Maintaining adequate levels of physical activity helps prevent psychological and behaviours [38] and has been positively linked to the quality of life of children and adolescents during periods of physical distancing [46].

In Italy, most of the children do not meet the WHO guidelines on physical activity (at least an average of 60 daily minutes of moderate-to-vigorous intensity physical activity and at least three times per week of vigorous-intensity physical activity), even before the pandemic [5]. Data from the sixth round of the surveillance system “OKkio alla SALUTE” [24], which was carried out in 2019, showed that only 32% of children had a moderate-to-vigorous intensity physical activity level (at least an average of 60 min per 5–7 week days) in 2019 in Italy. For vigorous-intensity physical activity (practising sports), around 30% of children in Italy reported a level of vigorous-intensity physical activity that agreed with the WHO recommendations. Furthermore, the WHO recommends limiting sedentary behaviours, such as screen-based entertainment and digital communication by television, computer and mobile phone. Data from the 2019 “OKkio alla SALUTE” [24] highlighted that around 45% of Italian children spent more than two hours a day watching TV or using electronic devices. A key question is if behaviours formed during the COVID-19 pandemic continued later [47]. The comparison between OKkio alla SALUTE 2019 [24] and OKkio alla SALUTE 2023 [48] shows that the percentage of children practicing sports at least three hours a week and the percentage of those spending at least one hour a day on actively/vigorously playing have not changed a year following the pandemic period. Only a slight decrease in time spent on actively/vigorously playing and a little increase in practising sports were observed. A recent study investigated differences in children physical activity behaviours before and after the COVID-19 pandemic; the authors observed a greater engagement in structured sport activities, and a lower level of unstructured and spontaneous activities, like active play, after the pandemic [49]. Concerning children’s sedentary behaviours such as screen-time activities, data from the 2023 “OKkio alla SALUTE” showed that the percentage of children watching TV or using electronic devices more than two hours per day was comparable with those collected when 2019 “OKkio alla SALUTE” round, before the COVID-19 pandemic [47].

The EPaS-ISS study, at the national level, allowed to detect changes in children’s behaviours during the pandemic and better understand health determinants collected by “OKkio alla SALUTE” surveillance system in 2019 (pre-COVID-19) and in 2023 (post-COVID-19). Furthermore, at the international level, the EPaS-ISS enabled to compare the Italian results with those from other countries of the WHO European region.

Public health implications

Changes in perceived level of children’ inactivity during the pandemic were observed by the EPaS-ISS study. Furthermore, low levels of active play and sport activities along with high levels of sedentary lifestyle such as screen time among Italian children was observed by the Italian COSI “OKkio alla SALUTE” surveillance system also before the pandemic.

The produced knowledge emphasizes the need to improve strategies at school and community level to enhance health-related behaviours in children, such as engaging in physical activity and activities not involving screens. In particular, the opportunities for physical activity should be improved by school curricular and school-based extracurricular activities, as well as community initiatives, accessible to all families. Moreover, instructive screen mediation actions should be adopted at school and family level with the aim of directing children towards healthier choices including more outdoor movement, sport activities as well as peer relationships [50].

Building on these considerations, we suggest the following evidence-based recommendations: expanding school-based extracurricular physical activity programs, especially in socioeconomically disadvantaged areas; improving access to public outdoor spaces through urban planning; and providing families with resources to better manage screen device use and improve their media literacy. Collaboration across education, health, and urban development sectors is essential to address these behavioural inequalities [23, 51].

Strengths and limitations

This study has strengths and limitations. The main strength is the availability of a large national sample from almost all Italian regions to investigate the impact of the COVID-19 pandemic on time that children spent studying at home, playing actively/energetically outside school hours and watching TV, playing videogames or using social media for non-educational purposes. The use of standardized data collection procedures based on the international COSI, as well as the support of regional coordinators and local health unit personnel involved in the “OKkio alla SALUTE” surveillance system were other strengths of this study.

However, this EPaS-ISS also has some limitations. One of the main limitations was requiring respondents to recall their children’s physical activity habits prior to the pandemic or to talk about any perceived changes. Surveys that asked respondents to recall the time before the pandemic began may be affected by recall bias. The EPaS-ISS results are based on parent reports, because young children (8–9 years old) are still limited in their abilities to accurately describe changes in movement and screen-time activities during the COVID-19 pandemic. In our study, the children’s evaluation is based on opinions of their parents, which might not accurately reflect the actual behaviours of the child. When respondents declared that the level of movement behaviors of their children was “less” or “much less” compared to that before the pandemic, it is difficult to accurately evaluate the actual amount of physical activity children engaged in. This might reflect a level of subjectivity in the responses. Furthermore, the “COVID-19 period” (from February/March 2020 to April 2022) was characterized by different containment measures e.g. school closures, physical distancing, lockdown and quarantine at home; this made more difficult to place the events over time and describe in deep the possible changes occurred during the COVID-19 pandemic.

The possibility of selection bias due to low response rate (46.6%) was another limitation of this study. To explore the possible presence of selection bias, data on the educational level obtained from the EPaS-ISS study were compared with those from the “OKkio alla SALUTE” surveillance system. In 2019, “OKkio alla SALUTE” achieved a response rate of 95%. The comparison revealed that the EPaS-ISS study involved a smaller proportion of parents with a lower educational level. Concerning “OKkio alla SALUTE” surveillance system, the information was collected through a printed questionnaire addressed to parents, whereas a web questionnaire was used in the framework of EpaS-ISS. Being an Internet user, i.e., having basic technical skills such as being able to use computers, tablets or smartphones, was required to take part in this investigation and could have been a source of selection bias in the EPaS-ISS. Consequently, it is possible that more vulnerable families were less likely to participate in this study; therefore, this may have led to lower representation from these groups or even an underestimation of inequalities.

In addition, the differences in children’s changes observed by geographical area might also have been affected by diverse local policies, depending on the transmission rates of the SARS-CoV-2 virus in each Italian region and autonomous provinces, which could not be assessed through the EPaS-ISS study.

Moreover, the high number of missing responses for questions concerning parents’ occupation and their work organization during the COVID-19 period [45] did not allow this sociodemographic information to be included in the statistical analyses.

Finally, being the aim of this study the identification of independent predictors of changes in children’s behaviours during the pandemic, the interactions between variables were not addressed.

Conclusions

The EPaS-ISS study showed that the COVID-19 pandemic worsened the trend of inactivity among children living in Italy, which was worrying even before the pandemic. The risks of low physical activity levels and high time spent in screen time activities among children include short-term (overweight, depression, and reduced well-being) and long-term impacts on health. Measures for maintaining adequate levels of physical activity and reducing screen time among children should be adopted in case of a future pandemic or another unexpected life event. The monitoring of these behaviours among young people before (2019) and after the pandemic (2023) through a routine surveillance system like the Italian OKkio alla SALUTE survey has revealed minimal changes in the amount of time spent engaging in movement activities and sports as well as using electronic devices. It will be interesting to assess if these changes in perceived level of inactivity remain and influence childhood health. In this regard, longitudinal studies focusing on children’s physical activity and screen-based behaviours should be undertaken to assess the long-lasting effects of the lockdown measures. Moreover, further research should review the policies and actions adopted by other countries to promote positive changes in lifestyles of children and their families. Finally, future studies on children’s health behaviours using data from the EPaS-ISS study might be focused on socio-environmental interactions using advanced predictive techniques to model non-linear relationships.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

M.B. (Marta Buoncristiano), S.C., P.N., A.S. conceptualized and designed the study; M.B. (Marta Buoncristiano), M.G. analysed the data; S.B., S.C., M.G., P.N. wrote the first draft; A.S., M.B. (Marta Buoncristiano) contributed to the interpretation of data and reviewed the manuscript; S.A., S.B., S.C., M.B. (Marta Buoncristiano), M.B. (Mauro Bucciarelli), M.G., P.N., critically reviewed the manuscript and approved the final version. All authors have read and agreed to the published version of the manuscript.

Funding

This research was promoted and funded by the ISS (Bando Ricerca Indipendente 2020–2022, Project code ISS20-4b59de447cf1) in the framework of the Project “Effects of COVID-19 pandemic on health behaviour and lifestyle of children and their families living in Italy”.

Data availability

The data presented in this study are available in accordance with the ISS data access policy. Requests should be directed to silvia.ciardullo@iss.it, member of the National Centre for Disease Prevention and Health Promotion, ISS.

Declarations

Ethics approval and consent to participate

Documents concerning the project, i.e., study protocol, information note with the description of the aim of the study and the consent participate, privacy policy for participation in the study and consent to the processing of personal data and questionnaire for parents were evaluated by the Data Protection Officer and Ethics Committee of the ISS. All documents were formally approved by the Ethics Committee on 23 December 2021 (0045150 Class: PRE BIO CE 01.00). Informed consent for study participation and informed consent for the processing of personal data were obtained from all participants involved in the research.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.