Current study

The present study examines the trajectory of MVPA from April 2020 to January 2021 for young adults (ages 18 to 29) and investigates differences in the mean level and rate of change for key demographic and health-related variables. Data were drawn from the Understanding Society COVID-19 Study. Based on previous research conducted during and shortly after the first lockdown, we expect; (a) MVPA outcomes to model restrictive measures, improving during periods of eased restrictions and declining in line with the three national lockdowns [3]; (b) worse MVPA outcomes for women, ethnic minorities, individuals with a lower income, individuals living alone, individuals with a pre-existing physical or mental health diagnoses, individuals who report greater cigarette use, and individuals with limited access to outdoor space [4, 5, 8, 10].

Participants and data source

Secondary data were derived from the Understanding Society COVID-19 Survey, a nationally representative, longitudinal panel study of UK households participating in the Understanding Society UK Household Longitudinal Study (UKHLS). The Understanding Society COVID-19 Survey data are available through the UK Data Service (https://ukdataservice.ac.uk/). The data can be downloaded, ordered or analysed online by registering and accepting their End User Licence. The overall method for gaining consent for participation is oral (see https://www.understandingsociety.ac.uk/documentation/mainstage/consents, for more information about consent). Questionnaires were sent monthly between April 2020 and June 2020, bi-monthly between September 2020 and March 2021, and finally in September 2021. Questionnaires for waves 1 to 4 were sent to participants who had responded to waves 8 or 9 of the UKHLS. Questionnaires for waves 5 to 8 were sent to participants who had recorded at least one partially complete response for any of the initial four COVID-19 survey waves. Wave 9 questionnaires were distributed to participants invited to wave 8. For each completed survey, participants were awarded £2. A final £10 payment was awarded for completion of the wave 9 survey, marking the end of the study. Ethical approval was granted by the University of Essex Ethics Committee for the primary data collection (ETH1920-1271). The present secondary data analysis was exempt from ethical approval.

The current study examined data from 18-29-year-old-adults from Waves 1 (April 2020), 5 (September 2020), and 7 (January 2021) as these were the only waves that assessed MVPA. Participants who did not provide responses for covariates collected at Wave 1 and did not provide two or more responses for the MVPA outcome were excluded from the final analytic sample due to the requirements of growth curve modelling (GCM).

Measures

Table 1 presents the wave of measurement, ranges, means and standard deviations of measures used for the MVPA model.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Characteristics of measures for the model predicting MVPA.

https://doi.org/10.1371/journal.pone.0289416.t001

Gender was coded as a dichotomous variable where the value “1” was assigned to male participants and “0” to female participants.

Ethnicity was dummy coded into five ethnic categories including, White, Mixed, Asian, Black, and Other.

Household Income was coded as a 15-interval categorical variable where each interval represented a £5000 increase in annual income (e.g., 1 = “£0–£5000”; 15 = “£70,001 and greater”).

Total Household Composition was a count of five variables indicating the number of household members in the age ranges of: “0–4”; “5–15”; “16–18”; “19–69”; “70 or older”. A higher score indicated a greater number of household members.

Living With Partner was a dichotomous variable drawn from the question “are you currently living with a partner?” where “yes” was assigned the value “1” and “no” the value “0”.

Access to Outdoor Space was measured using five dichotomous variables indicating whether the participant has access to: a private garden; shared garden; balcony, rooftop garden or terrace; other outdoor space; or no access. For each variable, “yes” responses were coded with the value “1” and “no” responses with the value “0”.

Mental Health Condition was a single question asked whether the participant had a pre-existing emotional, nervous, or psychiatric problem. “yes” responses were coded with the value “1” and “no” responses with the value “0”.

Physical Health Condition was a count of 20 items asking whether participants had a specific pre-existing physical health condition. A range of conditions was covered including asthma, diabetes, cancer or malignancy, epilepsy, H.I.V. and other medical conditions.

Cigarette Use was a continuous variable derived from the question “Approximately how many cigarettes a day do you usually smoke, including those you roll yourself?” Responses were coded as: 0 = “none”; 1 = “1–5”; 2 = “6–10”; 3 = “11–15”; 4 = “16–20”, 5 = “20 or more”.

Moderate-Vigorous Physical Activity (MVPA) was a continuous variable represented in waves 1, 5, and 7, indicating the total hours spent engaging in moderate and vigorous physical activities during the week prior to the survey completion. Measured using the International Physical Activity Questionnaire, moderate activities denoted those that “make you breathe somewhat harder than normal”. Vigorous activities denoted those that “make you breathe much harder than normal”. For each intensity, a total time variable was calculated by multiplying participants’ estimates for the hours spent engaging in PA on a typical day by their estimate for the number of days spent engaging in PA over the previous 7 days. The total variables for moderate and vigorous PA were then summed to calculate the combined MVPA total. Totalled estimates of 16 hours or more (n = 237) for each intensity were removed from the data set to account for outliers and overestimation [14]. Use of the MVPA weekly total permitted comparison against the 75–150-minute baseline recommended by the UK Physical Activity Guidelines to attain mental and physical health benefits [2].

Statistical analysis

GCM was utilised to explore the trajectory of MVPA (April 2020-January 2021). GCM was selected for the present data set as it accounts for missing data through maximum likelihood estimation and does not require equal time interval spacing across measurement waves [15]. All statistical analyses were conducted using SPSS 27.

To assess the linear rate of change across each assessment wave, a time variable was devised reflecting the number of months passed since wave 1: Wave 1 = 0, Wave 5 = 5, and Wave 7 = 9. To assess the variable rate of change across each assessment wave, a quadratic time variable was devised by squaring the time variable at each wave. The model was constructed by incorporating the linear and quadratic time components, and significant covariates. Interaction terms between the covariates and time components were tested to assess their association with the rate of change in MVPA outcomes.

Further strengths and limitations

Several limitations should be considered prior to the practical application of our findings. Firstly, covariate data that was likely to change during the pandemic, such as income and household composition, were obtained from the first or second waves of data collection. Thus, the dynamic moderation of MVPA outcomes by these predictors was not captured [13]. Moreover, several covariates which may be associated with MVPA outcomes during COVID-19, such as employment status and the provision of childcare [4, 10, 18], were not included in analyses. Further longitudinal research employing a wider range of covariates is warranted to corroborate the findings of the present study and to explore other potential predictors of MVPA during the pandemic.

Secondly, males and BAME groups were under-represented in the sample. Therefore, the full extent of MVPA variation within these demographic groups may have been under-represented. Furthermore, a significant number of participants (n = 1489) were excluded from the final analytic sample due to the requirements of GCM. Comparative analyses between these groups revealed several differences in participant characteristics, which, if represented, may have altered the results. Despite the data being drawn from a nationally representative data set, these limitations highlight biases in the present sample and emphasize the importance of further longitudinal research with larger and more diverse samples.

Lastly, elements of the questionnaire content and distribution limited the present findings. For instance, MVPA outcomes were collected via self-report. As participants are evidenced to answer questions relating to personal health and well-being in a socially desirable manner [26], MVPA may have been over-reported by the current sample. That said, the IPAQ research committee’s recommended cut-off of 16 hours or more for totalled weekly exercise estimates was employed for our MVPA data to mitigate over-reporting [14]. Additional research efforts may consider corroborating the present findings via more objective means, such as exercise data tracking technology.

Conclusions

Despite these limitations, the current study provides novel insight into the trajectory and associated risk factors of MVPA for young UK adults during COVID-19. The findings suggest that further efforts should be made to incorporate physical activity promotion initiatives, such as the “Join the Movement” and “This Girl Can” campaigns set up by Sport England within academic and professional workplaces. Resources can also be directed towards more proximate modes of promoting physical activity, such as increasing accessibility to exercise outlets within economically deprived communities, investing in activity-friendly routes for commuting or leisure, or hosting subsidized community sport classes among other strategies. Overall, the findings contribute to the growing body of COVID-19 literature, providing direction for further research and aiding the development of immediate and future interventions.