Evolution of changes in physical activity over lockdown time: Physical activity datasets of four independent adult sample groups corresponding to each of the last four of the six COVID-19 lockdown weeks in Greece.

There is a worldwide interest in how lockdown affects physical activity (PA) during the COVID-19 pandemic. Although it has been shown that the mandated stay-at-home restrictions and self-isolation measures applied in different countries were accosiated with a reduction in physical exercise and activity, such results derive from studying only specific periods of lockdown. However, in order for this hypothesis to be tested, consecutive lockdown periods need to be examined separately. In this study we focus on PA change in Greek adults over time, during each of the last four weeks of lockdown in Greece. The web-based Active-Q questionnaire (see Supplementary file 1_Active-Q) was used to collect data prior to the COVID-19 crisis (PRE condition) and during lockdown measures (POST condition). The period of data collection (5 April to 3 May 2020) was divided into four phases (Ph-I, Ph-II, Ph-III, Ph-V), corresponding to the 3rd, 4th, 5th and 6th lockdown week respectively (out of a six-week total lockdown). There were four independent groups of respondents (G-I, G-II, G-III, G-V) who reported their age, weight, height and usual PA habits. Energy expenditure (EE) was calculated (MET-min/week; see Supplementary file 2_Data) in four main different domains (daily occupation activities, means of transportation to and from daily occupation, leisure time and regular sporting activities; see Supplementary file 3_Corresponding MET values). Each group's dataset corresponded to one of the aforementioned phases (G-I to Ph-I, G-II to Ph-II, and so on). Overall PA change (from PRE to POST condition) ranged from -21.50% in G-I (Ph-I) to -5.03 in G-V (Ph-V); PA change in male subgroups ranged from -26.10% in Ph-I to -13.64 in Ph-V; in female subgroups it ranged from -17.42% in Ph-I to -1.39 in Ph-V. Although the decline in overall PA is evident in all groups during each lockdown phase (p<0.05), the combination of our data demonstrates that towards the end of lockdown this decline showed a gradual decreasing tendency.

Specifications Table

Value of the Data

These data show the magnitude of the daily PA routine disruption in a timely manner (during four consecutive weeks) since social distancing and lockdown were mandated in Greece.

We believe our data will help health authorities’ policy makers to develop specific guidelines for applied lockdown exercise or PA prescription and to create an efficient action plan against inactivity, overcoming COVID-19 lockdown barriers.

In a world of rising physical inactivity, our data will hopefully help determine public health priorities and health services plans, potentially dissuading people from adopting a sedentary lifestyle and rebooting PA in everyday life.

Since ethnicity, socioeconomic status, culture, educational level, religion, environment, health and psychological issues, sleeping/smoking/nutritional habits or other factors affect the population's physical activity, these data may generate additional hypotheses for upcoming research.

Informing the community of the PA decline during lockdown may have short- or long-term implications on the evolution rate of cardiovascular diseases, cancer, chronic respiratory diseases and diabetes, as well as on public health care systems in general.

Data Description

Given that the COVID-19 lockdown applied in each country has had different characteristics and duration, and that ethnicity, culture, environmental or other factors may separately or in combination affect physical activity (PA) in different populations, we focus on PA change in Greek adults during the last four weeks of lockdown in Greece. The Active-Q questionnaire (see Supplementary file 1_Active-Q) was used for data collection [in PRE condition (prior to the COVID-19 crisis) and POST condition (during lockdown measures)] [1], [2], [3]. Respondents stated their age, weight, height, and usual PA habits. Energy expenditure (EE) was calculated (MET-min/week) in four main different domains (daily occupation activities, means of transportation to and from daily occupation, leisure time and regular sporting activities). Data were collected from a total of 5206 individuals (28% males; 72% females) who participated in the survey (age 38.07±0.18 yr). This sample consisted of four independent groups (G-I, G-II, G-III, G-V). Each group's dataset corresponded to one of the last four weeks (phases: Ph-I, Ph-II, Ph-III, Ph-V) of lockdown (see Supplementary file 2_Data). Thus, reported data are presented separately (week by week). Gender variables were reported as frequency and/or relative frequency (%), numeric variables as mean±SE.

Anthropometric characteristics, frequency and relative frequency of groups and gender subgroups are presented in Tables 1, 2. Energy expenditure of activities (separately) and overall (as a sum of all activities) PRE and POST condition in all groups is presented in Table 3. Energy expenditure of leisure activities increased significantly in POST condition in all groups (p<0.05) by 17.53%, 21.72%, 16.07% and 27.41% respectively (Table 3). However, this increase was accompanied by a significant decrease in daily occupation, transportation and sporting activities in all groups and most subgroups (p<0.05; Table 3). As a result, overall PA in all study groups was significantly reduced in POST condition (p<0.05), by 21.50%, 13.93%, 10.63% and 5.03% respectively (Table 3). Overall PA change (from PRE to POST condition) ranged from −21.50% in G-I (Ph-I) to −5.03 in G-V (Ph-V); in male subgroups it ranged from −26.10% in Ph-I to −13.64 in Ph-V; in females subgroups it ranged from −17.42% in Ph-I to −1.39 in Ph-V, (for metadata, see Supplementary file 2_Data, in the sheet under the name Fig. 1).

Table 1

Anthropometric characteristicsa of groups (G-I, G-II, G-III, G-V) and gender subgroups respondents.

	G-I, (N = 1015)	G-II, (N = 2223)	G-III, (N = 1085)	G-V, (N = 883)
Age (yr)	40.33 ± 0.41	34.11 ± 0.29	40.6 ± 0.36	42.31 ± 0.35
Males	41.42 ± 0.66	42.44 ± 0.61	44.92 ± 0.61	46.27 ± 0.69
Females	39.52 ± 0.51	31.68 ± 0.31	38.96 ± 0.42	40.97 ± 0.39
Height (cm)	172.26 ± 0.3	169.39 ± 0.19	169.39 ± 0.25	169.08 ± 0.3
Males	180.31 ± 0.36	179.81 ± 0.34	178.63 ± 0.38	179.22 ± 0.39
Females	166.29 ± 0.25	166.35 ± 0.16	165.88 ± 0.21	165.65 ± 0.27
Weight (kg)	74.65 ± 0.54	69.62 ± 0.34	70.77 ± 0.47	72.64 ± 0.55
Males	85.91 ± 0.7	86.09 ± 0.67	84.88 ± 0.8	87.92 ± 0.92
Females	66.31 ± 0.57	64.81 ± 0.31	65.4 ± 0.45	67.48 ± 0.54
BMI (kg/m2)	25.01 ± 0.15	24.14 ± 0.1	24.54 ± 0.14	25.29 ± 0.16
Males	26.39 ± 0.2	26.59 ± 0.18	26.57 ± 0.23	27.35 ± 0.26
Females	23.98 ± 0.2	23.43 ± 0.11	23.77 ± 0.16	24.59 ± 0.18

Data are presented as mean ± SE. Abbreviation: BMI=body mass index.

Table 2

Frequency and relative frequency (%) of gender subgroups respondents by groups (G-I: N = 1015, G-II: N = 2223, G-III: N = 1085, G-V: N = 883).

	G-I	G-II	G-III	G-V
Males	432 (42.56)	503 (22.63)	299 (27.56)	223 (25.25)
Females	583 (57.44)	1720 (77.37)	786 (72.44)	660 (74.75)

Table 3

Physical activity dataa expressed in energy expenditure values, PRE and POST condition: a) by Active-Q four domainsb and overallc of all groups’ respondents (G-I, G-II, G-III, G-V) corresponding to each of the last four weeks (phases) of lockdown (pH-I, pH-II, pH-III, pH-V) and b) by gender subgroups.

Ph-I	PRE condition (MET-min/week)	POST condition (MET-min/week)
G-I, (N = 1015)
Daily occupation*	4736.30 ± 124.08	1945.34 ± 90.32
Transportation*	1309.86 ± 38.00	714.21 ± 30.88
Leisure time activities*	6241.27 ± 197.67	7335.26 ± 196.68
Sporting activities*	3511.92 ± 182.79	2406.97 ± 184.54
Overall*	15,799.35 ± 345.60	12,401.78 ± 304.26
G-I, males (N = 432)
Daily occupation (males)*	5389.69 ± 207.55	2310.97 ± 157.29
Transportation (males)*	1533.87 ± 63.91	885.12 ± 56.04
Leisure time activities (males)*	5844.67 ± 291.60	6962.63 ± 324.27
Sporting activities (males)*	4711.57 ± 347.93	2759.22 ± 223.91
Overall*	17,479.80 ± 559.02	12,917.95 ± 479.92
G-I, females (N = 583)
Daily occupation (females)*	4252.13 ± 148.71	1674.41 ± 104.26
Transportation (females)*	1143.87 ± 45.03	587.56 ± 33.22
Leisure time activities (females)*	6535.15 ± 267.42	7611.38 ± 243.56
Sporting activities (females)	2622.99 ± 178.12	2145.96 ± 274.78
Overall*	14,554.14 ± 429.60	12,019.31 ± 392.19
Ph-II	PRE condition (MET-min/week)	POST condition (MET-min/week)
G-II, (N = 2223)
Daily occupation*	4215.07 ± 78.41	1968.05 ± 58.68
Transportation*	1218.20 ± 23.71	697.59 ± 19.92
Leisure time activities*	6097.88 ± 127.67	7422.19 ± 132.92
Sporting activities*	2763.87 ± 151.89	2215.42 ± 134.19
Overall*	14,295.02 ± 248.35	12,303.25 ± 234.23
G-II, males (N = 503)
Daily occupation (males)*	5177.27 ± 182.93	2711.06 ± 153.70
Transportation (males)*	1314.24 ± 50.73	841.68 ± 45.25
Leisure time activities (males)*	6062.08 ± 289.05	7198.64 ± 282.94
Sporting activities (males)*	3849.72 ± 529.94	2809.8 ± 509.74
Overall*	16,403.31 ± 695.56	13,561.17 ± 722.59
G-II, females (N = 1720)
Daily occupation (females)*	3933.68 ± 84.90	1750.77 ± 60.11
Transportation (females)*	1190.11 ± 26.79	655.45 ± 21.99
Leisure time activities (females)*	6108.35 ± 141.75	7487.56 ± 150.56
Sporting activities (females)*	2446.32 ± 119.58	2041.59 ± 88.39
Overall*	13,678.46 ± 246.46	11,935.38 ± 216.14
Ph-III	PRE condition (MET-min/week)	POST condition (MET-min/week)
G-III, (N = 1085)
Daily occupation*	4284.91 ± 106.59	2349.61 ± 92.47
Transportation*	1254.93 ± 34.68	852.65 ± 31.60
Leisure time activities*	7007.04 ± 212.80	8133.13 ± 205.16
Sporting activities*	2328.27 ± 144.66	1901.51 ± 99.98
Overall*	14,875.14 ± 328.38	13,236.89 ± 284.78
G-III, males (N = 299)
Daily occupation (males)*	4806.32 ± 217.88	2676.81 ± 181.48
Transportation (males)*	1477.38 ± 77.31	996.89 ± 70.51
Leisure time activities (males)	7075.62 ± 458.93	7247.03 ± 411.29
Sporting activities (males)*	3404.49 ± 278.21	2823.46 ± 248.08
Overall*	16,763.82 ± 722.40	13,744.19 ± 632.52
G-III, females (N = 786)
Daily occupation (females)*	4086.56 ± 120.91	2225.13 ± 107.11
Transportation (females)*	1170.31 ± 37.37	797.79 ± 34.23
Leisure time activities (females)*	6980.94 ± 236.44	8470.21 ± 235.12
Sporting activities (females)*	1918.87 ± 167.14	1550.79 ± 97.97
Overall*	14,156.68 ± 357.50	13,043.91 ± 310.87
Ph-V	PRE condition (MET-min/week)	POST condition (MET-min/week)
G-V, (N = 883)
Daily occupation*	4328.27 ± 123.68	2174.39 ± 98.25
Transportation*	1149.50 ± 36.35	812.57 ± 34.21
Leisure time activities*	6767.79 ± 206.00	8622.59 ± 212.36
Sporting activities	1820.42 ± 141.94	1749.32 ± 171.66
Overall*	14,065.98 ± 335.14	13,358.87 ± 326.99
G-V, males (N = 223)
Daily occupation (males)*	5477.33 ± 294.99	2699.65 ± 227.52
Transportation (males)*	1370.91 ± 84.69	1008.81 ± 79.34
Leisure time activities (males)	7248.37 ± 511.78	8009.67 ± 479.28
Sporting activities (males)	2437.01 ± 350.79	2561.05 ± 338.68
Overall*	16,533.62 ± 912.29	14,279.18 ± 756.51
G-V, females (N = 660)
Daily occupation (females)*	3940.03 ± 128.76	1996.91 ± 105.84
Transportation (females)*	1074.70 ± 38.94	746.27 ± 36.77
Leisure time activities (females)*	6605.41 ± 214.51	8829.68 ± 233.13
Sporting activities (females)	1612.08 ± 147.68	1475.05 ± 198.13
Overall	13,232.22 ± 319.70	13,047.91 ± 354.58

Data are presented as mean ± SE.

Daily occupation, transportation, leisure time, and sporting activities.

Overall= daily occupation + transportation + leisure time activities + sporting activities.

p<0.05, significant difference between PRE and POST conditions. Abbreviation: MET=metabolic equivalent task (=3.5mLO2•kg−1•min−1).

Respondents’ PA estimates, expressed in energy expenditure values, POST condition by Active-Q four domains and overall, adjusted for the PRE condition covariate values, in each of the four phases are presented in Table 4, Table 5, Table 6. Although the decline in overall PA in all groups is evident at each phase of lockdown (p<0.05; Table 3), it appears that toward the end of lockdown this decline showed a gradual decreasing tendency (Table 3–6).

Table 4

Respondents’ PA estimatesa (expressed in energy expenditure values) POST condition by Active-Q four domainsb and overallc, adjusted for the PRE condition covariate values, in each one of the last four weeks (phases) of lockdown (Ph-I: N = 1015, Ph-II: N = 2223, Ph-III: N = 1085, Ph-V: N = 883 respectively).

	PA estimates POST condition (MET-min/week)
Daily occupation1
Ph-I	1823.58 ± 82.50
Ph-II	2010.78 ± 55.71
Ph-III*†	2370.29 ± 79.72
Ph-V*	2181.38 ± 88.36
Transportation2
Ph-I	694.97 ± 29.59
Ph-II	701.02 ± 19.98
Ph-III*†	847.00 ± 28.60
Ph-V*†	832.99 ± 31.72
Leisure time activities3
Ph-I	7429.58 ± 173.80
Ph-II	7588.55 ± 117.50
Ph-III	7842.62 ± 168.23
Ph-V*†	8452.31 ± 186.36
Sporting activities4
Ph-I	1959.18 ± 142.45
Ph-II	2160.32 ± 96.08
Ph-III	2075.08 ± 137.56
Ph-V	2189.49 ± 152.68
Overall5
Ph-I	11,824.31 ± 267.50
Ph-II	12,495.24 ± 180.61
Ph-III*	13,132.15 ± 258.47
Ph-V*†	13,668.01 ± 286.58

Estimates are presented as mean ± SE.

Daily occupation, transportation, leisure time, and sporting activities.

Overall= daily occupation + transportation + leisure time activities + sporting activities.

and †: p<0.05, significant difference with Ph-I and Ph-II respectively.

Covariates appearing in the model are evaluated at the following PRE values respectively: 1 4350.44, 21,232.07, 3 6428.94, 4 2658.91, 5 14,670.37. Abbreviations: MET=metabolic equivalent task (=3.5mLO2•kg−1•min−1); PA=Physical activity.

Table 5

Male respondents’ PA estimatesa (expressed in energy expenditure values) POST condition by Active-Q four domainsb and overallc, adjusted for the PRE condition covariate values, in each of the last four weeks (Phases) of lockdown (Ph-I: N = 432, Ph-II: N = 503, Ph-III: N = 299, Ph-V: N = 223 respectively).

	PA estimates POST condition (MET-min/week)
Daily occupation1
Ph-I	2248.61 ± 144.31
Ph-II	2722.44 ± 133.70
Ph-III	2816.97 ± 173.58
Ph-V	2606.86 ± 200.86
Transportation2
Ph-I	847.27 ± 50.49
Ph-II	877.82 ± 46.80
Ph-III	978.07 ± 60.63
Ph-V	1025.86 ± 70.21
Leisure time activities3
Ph-I	7247.01 ± 275.06
Ph-II	7369.05 ± 254.77
Ph-III	6886.16 ± 330.66
Ph-V	7558.25 ± 382.94
Sporting activities4
Ph-I	2084.03 ± 211.67
Ph-II	2771.35 ± 195.83
Ph-III*	3113.94 ± 254.05
Ph-V*	3566.31 ± 294.68
Overall5
Ph-I	12,471.21 ± 432.31
Ph-II	13,839.37 ± 400.56
Ph-III	13,779.61 ± 519.45
Ph-V*	14,469.62 ± 601.51

Estimates are presented as mean ± SE.

Daily occupation, transportation, leisure time, and sporting activities.

Overall= daily occupation + transportation + leisure time activities + sporting activities.

: p<0.05, significant difference with Ph-I.

Covariates appearing in the model are evaluated at the following PRE values respectively: 1 5210.05, 2 1421.51, 3 6387.18, 4 3797.67, 516,816.42. Abbreviations: MET=metabolic equivalent task (=3.5mLO2•kg−1•min−1); PA=Physical activity.

Table 6

Female respondents’ PA estimatesa (expressed in energy expenditure values) POST condition by Active-Q four domainsb and overallc, adjusted for the PRE condition covariate values, in each of the last four weeks (Phases) of lockdown (Ph-I: N = 583, Ph-II: N = 1720, Ph-III: N = 786, Ph-V: N = 660 respectively).

	PA estimates POST condition (MET-min/week)
Daily occupation1
Ph-I	1606.47 ± 101.75
Ph-II	1774.60 ± 59.22
Ph-III*†	2204.91 ± 87.60
Ph-V*	2018.91 ± 95.60
Transportation2
Ph-I	590.32 ± 36.82
Ph-II	649.46 ± 21.44
Ph-III*†	795.54 ± 31.71
Ph-V*†	762.13 ± 34.63
Leisure time activities3
Ph-I	7567.07 ± 225.91
Ph-II	7653.42 ± 131.61
Ph-III	8206.37 ± 194.72
Ph-V*†	8750.78 ± 212.33
Sporting activities4
Ph-I	2048.60 ± 175.50
Ph-II	1986.54 ± 102.17
Ph-III	1622.01 ± 151.12
Ph-V	1619.72 ± 165.09
Overall5
Ph-I	11,735.82 ± 336.95
Ph-II	11,997.73 ± 196.10
Ph-III*	12,917.40 ± 290.10
Ph-V*†	13,286.51 ± 316.66

Estimates are presented as mean ± SE.

Daily occupation, transportation, leisure time, and sporting activities.

Overall= daily occupation + transportation + leisure time activities + sporting activities.

and †: p<0.05, significant difference with Ph-I and Ph-II respectively.

Covariates appearing in the model are evaluated at the following PRE values respectively: 1 4016.37, 21158.45, 3 6445.17, 4 2216.35, 5 13,836.34. Abbreviations: MET=metabolic equivalent task (=3.5mLO2•kg−1•min−1); PA=Physical activity.

Experimental Design, Materials and Methods

The interactive web-based Greek revised version of Active-Q for adults’ questionnaire (Active-Q) was used to access PA on weekly basis [1,2]. The aforementioned PA self-reported questionnaire has been previously validated, tested for reliability and adequately described elsewhere [1], [2], [3]. For the purpose of data collection, individuals were openly invited via social media hosted links, emails and public advertisements to provide their personal PA information.

Via the questionnaire all respondents reported their age (yr), weight (kg), height (cm) and usual PA habits in four main different domains [daily occupation activities (i.e. sitting, standing, walking, heavy manual labor, etc.), means of transportation to and from daily occupation (i.e. walking, cycling, by car, by public means, etc.), leisure time (i.e. housework, gardening, watching television, using the computer, walking, etc.), and regular sporting activities (i.e. aerobics, weight lifting, jogging, team ball games, martial arts, tennis, swimming, etc.)]. The Active-Q comprised 9 to 46 items (questions), dependent on previous answers and follow-up patterns [1]. Items were aimed at assessing, by predefined answers, the frequency and duration of PA habits during a participant's weekly period. For instance, the Active-Q opened with a screening question assessing employment status (Yes/No); consequently, if a participant reported that they did not work, the upcoming items in relation to PA at work were omitted. Similarly, the initial items relating to means of transportation, leisure time and sporting activities were screening questions listing all activities included in each domain. Therefore, answers relating to frequency and duration were only required by those respondents whose relevant activities had been determined by the introductory screening questions. In this manner the cumulative duration of any PA stated in each domain was recorded and the EE of PA was calculated based on the corresponding metabolic equivalent tasks (METs) value, according to the 2011 Compendium of Physical Activities (e.g., PA of 5MET x 20 min/day x 3times per week =300MET-min/week [1,4]; see Supplementary file 3_Corresponding MET values).

Concisely, all participants were requested to fill out the Active-Q in a double fashion administration in the same day (the Active-Q being referred to in this article as a double fashion administration questionnaire). In their first admission (PRE condition), respondents recalled and stated any PA under normal life conditions (i.e. during the first two weeks of March 2020), prior to the COVID-19 presence in Greece. In the second admission (POST condition), respondents stated any PA post COVID-19 Greece outbreak and under lockdown conditions (i.e. from 5 April to 3 May 2020). Data collection was divided into four phases (Ph-I, Ph-II, Ph-III, Ph-V), each one corresponding to the 3rd, 4th, 5th and 6th week respectively, out of a six-week total lockdown (22 March to 3 May 2020). Accordingly, in each of the four phases of the survey, a different and independent sample group of participants (G-I, G-II, G-III, G-V) answered the questionnaire, stating their PA habits prior to COVID-19 (PRE condition) and only for the corresponding lockdown week (POST condition).

Paired t–tests were used for comparing PA PRE and POST conditions in the same sample group or gender subgroup. One way ANCOVA was performed to measure any significant differences in POST condition of PA-adjusted means after checking the PA PRE condition between independent groups and subgroups. Bonferroni pairwise comparisons (post hoc analysis) were also used to determine specific group or subgroup significant difference in cases where any was found. The level of statistical significance was set a priori at p ≤ 0.05. Statistical analyses were performed with SPSS for windows (v23, Armonk, NY: IBM Corp., USA).

Ethics Statement

Ethical approval was obtained by the National and Kapodistrian University of Athens review board. All participants provided their written consent in accordance with The Code of Ethics of the World Medical Association (Declaration of Helsinki) for experiments involving humans.

CRediT authorship contribution statement

Dimitrios I. Bourdas: Conceptualization, Methodology, Formal analysis, Investigation, Resources, Writing - original draft, Writing - review & editing, Supervision, Project administration. Emmanouil D. Zacharakis: Conceptualization, Writing - review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships which have, or could be perceived to have, influenced the work reported in this article.

Subject	Public Health and Health Policy
Specific subject area	Physical activity and lockdown
Type of data	TableFigure
How data were acquired	Survey.The interactive web-based Greek revised version of Active-Q for adults questionnaire was used to access physical activity on weekly basis (see the supplementary material file1)
Data format	Anthropometric data are in raw format and body mass index was calculated. Physical activity data have been analysed and energy expenditure (EE) was calculated (MET-min/week).
Parameters for data collection	Inclusion criteria for participation in the survey: ≥18 yr, Greece residency, Internet access.Exclusion criteria: participation in a strict weight loss control program, any form of illness, pregnancy or childbirth in the previous year.
Description of data collection	Data were collected in four consecutive weeks (5th of April till 3th of May 2020), each week by a different independent volunteer sample group, through an Internet survey source providing an interactive web-based questionnaire. An Excel file with the aforementioned data has been uploaded (see supplementary material file 2).
Data source location	Institution: National and Kapodistrian University of AthensRegion: EuropeCountry: Greece
Data accessibility	Data and metadata are hosted with the article.