Main body
Similar to other studies that assessed a range of lifestyle habits during lockdown [
19,
20,
30], our study showed consistent findings in regard to the adverse effects and the interlinked association of sleep, stress and social support. For example, the multi-country study by Ammar et al. of 1047 adults across 3 continents showed a positive correlation of mental health with higher social support and a negative correlation with poor sleep [
30]. Similarly, the Spanish study of 1254 adults showed that higher social support, stress management and higher outdoor exposures were the most important factors associated with better health behaviours [
19]. There was also agreement between our findings and those of the other studies in regard to the higher food consumption and sedentary time during lockdown [
19,
20,
30]. However, there was variation in the findings in regards to physical activity changes which could be explained by differences in restriction measures and population baseline physical activity levels [
20,
30].
Regarding our findings in each individual pillar and starting with diet during lockdown, participants in our study were affected in various ways. Firstly, diet quality, as reported by the MEDAS score, was moderate. Whilst diet quality seemed to improve in some areas but not in others during lockdown, there was an overall improved MD adherence score amongst participants who were fasting, as per the Greek Orthodox religion tradition. This is not surprising since fasting is a plant-based diet, thus closer to the original MD [
31]. Agreeing with our study, adherence to MD during lockdown was moderate in an Italian study with 3533 participants aged 12–86 years. This study reported an increase in the sense of hunger and appetite as well as perceived weight gain in almost half of the participants [
32]. Similarly, in another Italian study, half of participants reported higher food consumption as a result of eating more “comfort food” (sweets and salty snacks) but also fruits [
11], whereas a study from Poland reported increased snacking between meals especially amongst the obese [
33]. The ECLB-COVID19 International Online Survey also showed an increase in the number of meals and snacking during confinement and a higher unhealthy diet score [
30]. However, the Spanish COVIDiet study, which also assessed adherence to MD before and after lockdown, adherence to the MD increased significantly from 6.53 +/− 2 to 7.34 +/− 1.93. COVIDiet participants with higher MD adherence decreased intake of sweet/carbonated beverages, red meat and pastries by 16–18%, yet increased fruit and vegetable intake by around 12% [
34]. Similar to our findings, COVIDiet participants with postgraduate education had higher MD adherence.
Pertaining exercise, the average weekly activity score per participant in our population did not change during lockdown. These findings are in contrast to a recent systematic review on physical activity and sedentary behaviour during COVID19 lockdown in healthy adults which included 44 studies that in their majority did not use a validated PA measurement tool and suggested a reduction in physical activity levels of individuals in lockdown in most countries [
12]. For example, an online survey of 1471 adults in Australia reported a negative change in the physical activity of almost half of the participants [
9] whereas a study in Italy showed a significant decrease in the weekly MET-min score across all activity categories in 2524 adults [
35]. However, just over 75% of the Italian participants had moderate or high physical activity levels before lockdown and the negative impact of lockdown was mostly seen in these individuals Similarly, the systematic review also showed that people with higher pre-lockdown physical levels were more likely to have a larger decrease in PA levels during lockdown [
12]. This is in agreement with our findings that showed a decrease in physical activity levels only in those with moderate and vigorous pre-lockdown physical activity levels. In our study however, half of our participants had low physical activity levels. Individuals classified as low active before lockdown showed a significant increase in their physical activity levels in lockdown as demonstrated by the study of Rodrigo et al. in 1155 adults in Spain [
13]. This can explain the fact that in our population, the number of participants who spent time walking increased during lockdown Walking is usually a preferred exercise amongst less active individuals [
36]. Moreover, in Cyprus during spring, walking was likely a well-suited outdoor activity for families and seniors. This increase in energy expenditure in walking however was negated by the decrease in moderate and vigorous activity, thus explaining the overall picture of no change in physical activity in our population. Finally, and unsurprisingly, staying at home with a “once a day” allowance to go out led to an increase in the time participants spent sitting and in other sedentary activities, something evidenced by other studies [
9,
35] and a recent systematic review on physical activity and sedentary behaviour during COVID-19 lockdown [
12].
Sleep, stress, and social support are important interrelated factors in lifestyle medicine [
21]. During the COVID-19 lockdown, significant associations were reported between them in studies that evaluated stress and sleep [
16] and social support and stress [
37]. To our knowledge, few other studies to-date have evaluated social support, sleep, and stress [
19,
38]. A smaller study (
n = 170) in China evaluated persons under self-isolation [
38] and showed that low levels of social capital were associated with increased stress, which in turn reduced sleep quality.
In our study, social support decreased during COVID-19 lockdown, which differs from results seen in studies in the US [
39] and Egypt [
37], where social support increased. This difference may be driven by factors such as timing of the study and degree of lockdown measures as well as societal and cultural differences. Decreased social support in our study was associated with increased perceived stress (
r = − 0.3742,
p < 0.01), related to findings of other studies showing the adverse effects of loneliness and lack of social support on stress and mental health during the COVID-19 pandemic [
39‐
41]. Additionally, our study confirms other findings during the COVID-19 pandemic that higher perceived stress is associated with lower sleep quality [
16] and that the proportion of those with poor sleep quality increased [
42]. A recent systematic review and meta-analysis reported a global pooled prevalence rate of sleep problems of 35% [
14]. In our population and although global sleep quality significantly changed in our participants during lockdown (global PSQI score: 4 before vs. 5 during lockdown,
p < 0.01), it is noteworthy that both before and during lockdown our respondents, overall, had “good” sleep quality (global PSQI score ≤ 5).
Given the association between stress, anxiety and substance use [
43], smoking and alcohol consumption frequency and/or intensity during lockdown were expected to increase in some people due to higher stress levels and decrease in others who smoke or drink socially. Findings from our study confirm the above; 43.8% of smokers increased and 28.1% decreased the daily number of cigarettes smoked during the lockdown. Similarly, the overall frequency of alcohol consumption increased in 11.5% and decreased in 26% of participants, while the number of drinks consumed showed a similar pattern.
Regarding smoking, similar findings have been found in a study conducted during the COVID-19 lockdown in the US, where approximately a quarter of participants reduced smoking and a third increased their motivation to quit, whilst 30% increased their smoking [
44]. A similar survey conducted across five countries (Italy, India, South Africa, UK, and US) including 6800 smokers under a variety of lockdown measures, found that e-cigarette consumption marginally increased during lockdown [
45]. The latter study also revealed that in-home smoking increased in Italy and India among exclusive tobacco cigarette smokers. Both studies note that smoking behaviour of participants was also affected by the perception of increased risk of infection or higher COVID-19 disease severity [
44,
45]. Although we did not assess perceptions of infection related to smoking, it is very likely that our participants who reduced or quit smoking during lockdown had similar concerns or that the strict Cyprus lockdown measures prevented social smoking.
Concerning alcohol, there are conflicting findings in the literature. Our study findings are in line with an Italian survey reporting a 36.8% reduction in alcohol intake, probably due to reduced social drinking [
11]. Conversely, a study conducted in Poland reported an increase in alcohol consumption in approximately 14% of participants, although more pronounced in alcohol addicts [
33]. Similarly, UK evidence on drinking habits during COVID-19 lockdown [
46] saw elevation in the proportion of risky drinkers. This is in contrast to our findings, showing a much higher decrease than increase (11.6% vs. 3.1%) in high-risk drinking (≥6 alcoholic drinks on one occasion) during lockdown. Similarly, the ECLB-COVID19 International Online Survey also showed a reduction in binge drinking at a global level [
30]. Similar to our findings, in the UK study the proportion of people drinking less during lockdown was similar or exceeded the proportion of those drinking more [
46]. Furthermore, a survey conducted by the charity Alcohol Change UK [
47] also revealed that one in five participants drunk as a response to stress or anxiety during the lockdown and more than one in three acted to manage their drinking, with 7% stopping altogether. Of note, in our study the proportion of people reporting never drinking increased (36.2% vs. 22.3%).