In this paper, we examined the correlates and adaptive function of anxiety during the first wave of the COVID-19 pandemic in the UK and Turkey. As predicted, more future-oriented participants had higher levels of generalized and pandemic related anxieties, however the effect sizes were small, and intolerance of uncertainty had the largest effect on anxiety levels. Participants with elevated pandemic-related anxiety levels engaged in risk avoidance behaviour more frequently, stressing the adaptive role of anxiety in reducing mortality risk. Participants with increased mindful attention awareness had lower levels of both generalized and pandemic related anxiety. Several other behavioural and demographic factors, such as age, following the news, financial satisfaction and trust in others were associated with pandemic related anxiety levels and risk avoidance behaviour. Overall and pandemic driven anxiety levels were higher among Turkish participants whose risk avoidance behaviour scores were also higher compared to the participants in the UK. Below, we discuss each of these findings.
Future orientation and anxiety
We hypothesized that “too much” future orientation in modern societies may be contributing to the recent rise in anxiety disorders, since anxiety at it is core is an emotional response triggered in anticipation of possible future outcomes. Previous research on anxiety showed that anxious individuals exhibit a cognitive bias that they are more likely to attribute negative outcomes to uncertain situations 26,27 and find it hard to tolerate or accept uncertainty 28,29. Moreover, self-labelled worriers are primarily concerned about the uncertain future 13. Our findings support these observations suggesting that it is not future-oriented thinking per se but intolerance of future uncertainties that contribute to increased anxiety response. Since pandemics such as the COVID-19 bring about many future uncertainties, those who are less tolerant of uncertainty exhibit the highest anxiety response.
To further examine the association between individual time perspective and anxiety levels during a pandemic we measured participant’s awareness of the present moment experience (i.e. mindfulness). We found that participants with higher mindful attention awareness scores had lower levels of both generalized and pandemic anxiety. Our results reinforce previous findings on the positive effect of the cognitive therapies that involve mindfulness techniques, such as mindfulness based cognitive therapy 30 and acceptance and commitment therapy ACT 31, on reducing anxiety 16,32. Training the mind to focus on the present-moment experience seems to alleviate anxiety, possibly by taking one’s focus away from the future and the excessive thoughts of potential negative outcomes. We will discuss the benefits of anxiety response during a pandemic below, but it is important to note here that although mindfulness decreased the anxiety response, it did not make people less risk averse. On the contrary, mindfulness was positively associated with risk avoidance behaviour, suggesting that mindfulness-based therapies can alleviate anxiety without interfering with the benefits of a healthy anxiety response.
Adaptive function of anxiety during a pandemic
Only a few studies have demonstrated the benefits of anxiety 22,23, however an emphasis on “diagonal psychology” (i.e. the benefits of negative states and disadvantages of positive states) can help with better clinical decisions on when to act on emotional states and when a response can be considered normal 6. By focusing on the levels of anxiety during a pandemic, we were able to measure the effect of anxiety response on the avoidance of infection risk. Signal detection theory allows us to make predictions on the situations in which the anxiety response will be evoked, and which individuals will be more likely to fire this response 2. One such prediction is that the probability of dangerous events occurring in the environment should increase anxiety response 2. Supporting this prediction, our results showed that the mean GAD-7 scores during the first wave of the COVID-19 pandemic in both countries (Turkey: 8, UK: 6) were higher than the cut point of 5 for the mild levels of anxiety 33. Another prediction is that individuals with lower thresholds for threat detection should experience more anxiety 2. In line with this prediction, participants in both countries who reported higher levels of perceived risk of catching the novel coronavirus had higher pandemic related anxiety levels.
We found a strong correlation between an individual’s overall anxiety level (measured as generalized anxiety) and their level of pandemic-related worries, such as feeling stressed about leaving their house or being worried about their/their family’s health. Since individuals with high anxiety are predicted to have lower threat detection threshold, their pandemic related anxiety levels were also expected to be higher. Furthermore, the anxiety subtypes (e.g. various anxiety disorders) can be considered as partially differentiated responses of a general anxiety response adapted to different threat situations 21. For example, while social threats may trigger an anxiety response that evokes submissive behaviour, an encounter with a predator may trigger a response promoting freezing behaviour 21. Following this, pandemic driven anxieties are expected to trigger avoidance behaviours to protect against infectious agents. Indeed, our findings showed that risk avoidance behaviours such as keeping a distance from other people, avoiding crowded places and staying at home, were higher among the participants who reported higher levels of pandemic related anxiety. Elsewhere, we also showed that participants with higher pandemic related anxieties were more likely to vaccinate against COVID-19 24, which supports the adaptive role of anxiety in reducing mortality risk.
The benefit of anxiety in specific situations, such as a pandemic, brings about the question of what level of anxiety can be considered normal, especially given that the anxiety response is context dependent. In clinical psychology a condition is thought to be pathological if it is impairing the quality of life of an individual. Nevertheless, it is important to acknowledge that anxiety is on a spectrum and the clinical anxiety share the continuous spectrum with the normal anxiety response 2. An evolutionary perspective suggests that if a biological system is not producing the effects that it was selected for and is leading to harm, then it is not functioning normally, and can be considered a disorder 34. In the case of anxiety, a decision on whether to intervene with a therapeutic method can be based on asking whether the individual is avoiding situations and activities that are harmless or even beneficial, as avoidance is the core component of generalized anxiety 32. Moreover, although cognitive therapies and administration of drugs have been the main therapeutic methods, an evolutionary perspective reminds us of the importance of the situations (environments) that trigger negative mood states at the first place 19. Although it may not be possible to alter certain situations, such as an ongoing pandemic, more focus should be given on changing structural inequalities that put certain groups under more risk 35, leading to anxiety.
Age and anxiety
Although elderly people have been among the most vulnerable group during the COVID-19 pandemic, our findings showed that both generalized and pandemic-related anxiety levels declined with age in both countries (supplementary Fig. S1). Our results support previous findings on the prevalence of generalized anxiety disorder by age in the UK, where the prevalence start declining by age 55, with a sharper decline after 65 36. Moreover, a recent survey found a negative relationship between GAD-7 scores and age in Cyprus during the COVID-19 pandemic 37. It is possible that as people get older, they may get more experienced at coping with uncertainties, which render them less susceptible to anxiety. Indeed, we found a strong significant negative correlation between age and uncertainty intolerance in both countries (supplementary Table S3 and S4). It is also possible that there are fewer uncertainties associated with long-term future goals (e.g., getting promoted, completing a degree etc.) for the elderly. Likewise, financial satisfaction score was positively correlated with age in both countries (supplementary Table S3 and S4). Moreover, we found a strong positive correlation between the MAAS score and age, suggesting that mindful attention awareness is higher among the elderly (supplementary Table S3 and S3). These findings suggest that despite high mortality risk for the elderly during the COVID-19 pandemic, their increased tolerance of uncertainty and mindfulness result in decreased anxiety levels. Interestingly, being less anxious did not result in less risk avoidant behaviour in the older population, on the contrary engagement in risk avoidance behaviour increased with age (supplementary Fig. S1).
Trust in others, anxiety and risk avoidance behaviour during a pandemic
We found that trust in others in following social distancing rules decreased pandemic related anxiety but also decreased risk avoidance behaviour, especially in the UK (effect sizes were larger in the UK sample). This suggests that trust in others do not always act positively to contain the spread during pandemics. A recent multi-national study, for example, has found that social trust was associated with more COVID-19 deaths 38. Studies also showed that shared group membership and increased trust in others result in lower risk perception and higher risk-taking behaviours 39. Therefore, trust in others during a pandemic can act as a double edge knife, increasing pro-social behaviours such as mutual-aid and cooperation on one hand 40, and promoting risky health behaviours on the other 41.
We found differences in the overall emotional and behavioural response to the pandemic between the UK and Turkey. For example, both generalized anxiety levels and pandemic related anxiety levels were higher among Turkish participants. There was significant difference in the mean intolerance of uncertainty score between the two countries (on a scale of 1–5, MTurkey = 3.49 vs. MUK = 2.94), which was probably the main driver behind the higher anxiety scores in Turkey. Another factor that potentially contributed to the elevated levels of pandemic anxieties in Turkey was financial satisfaction. The average level of financial satisfaction, on a scale of 0-100, was 48 for Turkish participants, and 67 for the participants in the UK. Moreover, we found that perceived life expectancy was much lower among the Turkish participants, which predicts that perceived mortality risk and hence anxiety levels are higher. Finally, Turkish participants reported a lower level of trust in others in following the social distancing rules compared to the participants in the UK, which might have contributed to the increased pandemic anxieties in Turkey.
Engagement in risk avoidance behaviour during the COVID-19 pandemic was also significantly higher among the Turkish participants. It is possible that elevated anxiety levels in Turkey rendered people to be more careful in maintaining a social distance, wearing masks outside and avoiding crowded places. Another factor explaining the reported differences in following government guidelines (social distancing, staying home, wearing a mask, frequent hand washing and avoiding crowded places and social gatherings) may be the cultural differences in collectivist attitude (individualism score for Turkey was 37, as opposed to 89 for the UK) 42. Collectivist cultures may be more receptive to government guidelines and interventions, such as contact tracing, during an epidemic compared to individualistic cultures 43. Indeed, in a recent study levels of collectivism was associated with higher intentions to engage in social distancing behaviours during the COVID-19 pandemic 44. Our findings showed that there was an especially large difference in mask-wearing behaviour between the two countries. The majority of Turkish participants reported wearing a mask when outside at the start of the pandemic, whereas the score for UK participants was much lower (supplementary Table S1). This may also be due to the differences in government policies on mask-wearing between the two countries at the start of the pandemic.