*3.3. Regression Analysis*

We hypothesized that PES and ERQ evaluated prior to the pandemic could predict mental and physical symptoms during the COVID-19 lockdown. To test this hypothesis, we conducted a linear regression analysis using GSI (SCL-90R) scores as dependent variables and pre-pandemic ANPS and ERQ scores, age, and gender as independent variables.

The model explains 51% of the GSI scores (*R*<sup>2</sup> = 0.51; adjusted *R*<sup>2</sup> = 0.38; *F* = 3.85; *p* = 0.001), thus indicating an adequate fit of the model tested. The independent variables that showed a significant effect were: expressive-suppression ERQ (β = 0.45; *t* = 2.71; *p* < 0.001) and SADNESS ANPS (β = 0.71; *t* = 3.65; *p* < 0.001). Age and gender did not show any statistically significant results.

#### **4. Discussion**

Consistent with previous studies regarding the impact of the COVID-19 pandemic on psychological well-being [4,5,22,23], the presence of psychopathological symptoms was found in the sample we examined. More specifically, participants displayed higher symptoms of depression, anxiety and obsessive compulsiveness. Italy was one of the first countries to be significantly affected by COVID-19 and social isolation measures were immediately enforced there. Social distancing and government-imposed lockdowns have effectively kept many people in their homes. Italy is among the countries with higher death rates from COVID-19, and with a higher average age than the rest of the world, especially during the first stage of pandemic [24]. The high number of deaths that hit the country exposed all citizens to great stress and concern for their own safety. At that time, our survey focused on detecting the psychological and physical effects of the situation, as well as understanding the modalities of emotional regulation of those who were experiencing the lockdown. The opportunity to be able to contact those who had participated in one of our research projects before the outbreak of the pandemic allowed us to verify the balance/imbalance of the PES and of the cognitive-reappraisal and expression–suppression emotional regulation strategies investigated. Elderly people were the most affected by COVID-19, and, therefore, we investigated whether age had an incidence with the variables explored and with psychopathological symptoms in the lockdown phase. The results showed that seniority is negatively related to FEAR in both pre- and post- evaluation, and also negatively related to CARE systems in pre-pandemic evaluation, which is not a significant relationship to psychological symptoms. Therefore, although the elderly is the most affected population, age does not emerge as a risk factor for psychological stress. The relationship with fear as a primary emotional system and young age is in line with the data showing that young people have suffered greatly during the pandemic, accentuated by perhaps feeling more exposed [25,26].

The results showed that COVID-19 lockdown rules had an impact on Emotional Regulation and displayed a re-balancing effect on PES. Further, during the lockdown, there was a decrease of ERQ cognitive reappraisal. Cognitive-Reappraisal (CR) generally has a buffer function which aids in the prevention of psychopathology [27]. CR is a flexible and adaptive function of emotion regulation, but the results displayed a decrease in it due to the traumatic impact of COVID-19. However, in regard to PES, we found an increase in the PLAY and CARE systems, which served as protective elements against danger. Physical PLAY is the most complex basic social emotion and persists after neo decortication [11]; CARE, or the maternal nurturance system, includes nurturance and social bonding and suggests that there is an intimate evolutionary relationship to maternal motivations. These results are consistent to studies which have shown that lower scores in the PLAY and CARE systems are linked to depression [28]. In analyzing ERQ and PES results, it could be said that the decrease in CR showed a reduction in the ability to regulate the positive reinterpretation of situations [29,30]. In other words, the results showed how people's emotions manifest during traumatic events and how this, in turn, activated a rebalancing of their positive systems in attempt to cope with the stressful situation.

The second hypothesis tested whether specific PES and ERQ factors evaluated before the pandemic predicted psychopathological symptoms. A regression model showed that pre-pandemic variables, such as SADNESS and Expressive–Suppression were a risk factor and predicted higher psychopathological symptoms of GSI during lockdown. These results are consistent with recent results obtained during the COVID-19 pandemic, which showed the protective function of the ability to cope with anxiety and stress through cognitive reappraisal. On the other hand, expressive suppression served as a risk factor for the development of psychopathological symptoms [31,32]. This result is consistent with previous studies which found that other traumatic events, such as war or abuse, suggested that PES involved SADNESS [33,34]. Several developmental studies have highlighted how the persistence of specific negative emotions in children can predict future psychopathology. It was found that higher levels of sadness in adolescents predicted internalizing symptomatology [35]. Our results confirm a significant link between the dominance of a negative primary emotional system and psychopathological manifestation.

All of these findings need to be interpreted in the light of some limitations. First, the sample size of this study is relatively small. Furthermore, the sample was composed of volunteers available to be re-selected from a previous study, which may have introduced a selection bias. A further limitation is linked to having investigated only two macro strategies for regulating emotions, instead of explaining more specific modalities that are useful in the management of psychopathology [36].
