In summary, the most important findings in this study were that participants who perceived that their mental health had been negatively affected by the covid-19 pandemic had more symptoms of anxiety, depression/dejection, and stress. When analyzing risk factors for experiencing a decline in mental health during the pandemic, of these, only anxiety fell out as a risk factor, as well as a negative change in social life, a change in physical activity, experiencing a bad family situation, and female gender. Anxiety is common in the older population (16, 17), and during the Covid − 19 pandemic an increase was likely to occur as this age group was especially vulnerable to the virus, more socially isolated, and the special situation of a pandemic is associated with uncertainty regarding both the present and the future. Studies performed during the Covid − 19 pandemic confirmed increased anxiety among the older adults (10, 11, 18). Our result reflects what other research have found and suggests that anxiety is of major concern when evaluating and predicting mental health in the older population during pandemic-like conditions.
The participants in this study reported that the factors affecting them the most during the Covid-19 pandemic were social distancing, uncertainty, and fear of a close relative, or themselves, catching Covid-19. Comparable results are found in previous studies, but few asked their participants such a direct question, and, thus, their results are concluded more indirectly (10, 18, 19). However, to our surprise, the majority of our study participants did not perceive that the Covid − 19 pandemic had affected their mental health. This contrasts with previous studies. In two reviews from 2022 (10, 18) it is concluded that older adults are prone to suffer mental health issues due to the Covid 19 pandemic, e.g: loneliness, depression, anxiety, and stress. One major reason for this dissimilarity may be due to a large majority of our participants living with a partner, and, thus, were not completely socially isolated, which might also explain why they experienced a low degree of loneliness and perceived their mental health as good. In line with this hypothesis, a previous Swedish study found an association between decreased mental wellbeing and social isolation among older adults (11). Another factor to consider is that Sweden did not apply lock downs, as most other nations, and therefor loneliness and social isolation might be less of an influence on our population.
In Sweden it is estimated that approximately 300,000 people are socially isolated, pre-pandemic, which is defined as a person living alone, meeting relatives, friends, or acquaintances twice a month or less. After retirement social isolation increases, and in the age group 75–84 years every 10th person is socially isolated (20). It is known that the pandemic increased social isolation in this age group and several studies show that this affected their mental health (10, 11, 18). In accordance with this our study found that those who experienced a deteriorating social life during the pandemic also suffered from declining mental health. Interestingly though, as many as 80% reported no change in their social life, and 6% reported a positive change, few lived alone, and they were doing well in general. A possible explanation could be that these participants adapted well to the pandemic changes. Supporting this are the findings of Özdemir and Çelen (21) who saw that increased stress among old adults was associated to social isolation and chronic disease, and Sardella et al (22) whose results show that living with someone might be a factor of resilience. Studies also show that this age group have good psychological coping and adaptability, and that activities such as social media use, communication with others e.g. neighbors, seeking social support, and keeping themselves busy during the pandemic were successful protective measures (23, 24). We did not ask specifically about our participants’ social media use or communications with others during the pandemic, but in Sweden internet access and use is high, and in the ages 70 and older up to 75% use social media daily, thus, it is possible that this has acted as a favorable factor in our studied population (25).
Previous study results show that, in general, older women suffer from more anxiety, fear, worry, depression, and depressive symptoms than older men (2, 17). In this study we found a similar pattern with women being more negatively affected mentally by the pandemic than men. These findings were expected and are consistent with other research during the pandemic, e.g. C Reppas-Rindlisbacher et al who found that women were twice as likely to report depressive symptoms than men (26), and F Hou et al reporting that women were more affected by anxiety and stress (27). Uniformly, there are findings suggesting that older women might be more vulnerable in facing the Covid − 19 pandemic than older men (22).
According to WHO, 14% of the world’s population aged 60 and over suffer from at least one mental health disorder, depression and anxiety being the most common, and more than a quarter of all deaths by suicide occur in this age group (16). When comparing GDS 20, HADS-D and HADS-A to self-reported negative effect on mental health by the Covid − 19 pandemic in this study, symptoms of anxiety and stress in HADS-A and PSS 10 fall out as strong predictors for declined self-perceived mental health due to the pandemic (p < 0.001 and p = 0.026 respectively). Depression also falls out as a strong predictor, however, only in GDS 20 (p < 0.001) and not in HADS-D (p = 0.053), which was somewhat surprising. But, in HADS-D a large majority scored as “dejection”, in fact the exact same percentage (76%) as those who scored “likely depression” in GDS 20. This discrepancy between the two screening scales is in line with the findings of G. Campbell et al. (3), and is probably because GDS 20 is designed especially for the older population whereas HADS is constructed for the adult population in general.
In contrast to other studies (28, 29), like for depression, stress does not fall out as a risk factor. A reason for this might be, as discussed earlier, that most of our population lived in a relationship and few felt lonely, which could have increased their psychological resilience. One could also argue that stress levels might be lower later in the pandemic as fewer became severely ill after vaccinations started, restrictions lessened, and both the present and the future were less uncertain. However, our participants did state that social distancing and uncertainty were two of the factors that affected them the most, but apparently only a minority were so affected that they experienced decreased mental health.
A major strength in this study is the use of reliable and well-validated screening scales that together capture different aspects of mental illness. In addition, there is good agreement between the self-rating scales, which further strengthens our result. Two different scales for depression were used, which we consider to be a strength. For anxiety only HADS was used. This could be considered a limitation. However, HADS is well renowned and used frequently for investigating anxiety, also in older people. Furthermore, our results regarding anxiety are very rigid and show consistent statistical significance in the different analyses, thus, we consider our results reliable and unlikely to change direction using another screening tool. Physical activity is known as an important factor for maintaining good mental health, and the same has been shown for older people during the pandemic (30, 31). The results of this study indicate that a change in physical activity is a possible strong risk factor for declining mental health (p = 0.006, OR 5.76), however, the direction of change is unknown due to how the question was formed. This, of course, is a limitation. Regarding that our analysis found a correlation between change in physical activity and a decline in social life (r = 0.156, p = 0.012) indicates that the change is negative (i.e. reduced physical activity), but we cannot know for sure. Another limitation is the possible impact of recall bias affecting the result, a well-known phenomenon to take into consideration when using self-reported data, but, in general self-reports are considered reliable and well established (32). Our focus on a single question, “Has your mental health been affected negatively by the Covid − 19 pandemic?”, for much of our analyses is also a limitation that could make our results more vulnerable. Not least since the sub-population answering this question “Yes” was small (n = 24), which indicates carefulness when analyzing the results. Those who answered “Do not know” (n = 30) were considered to be included into this sub-population as sub-analyses showed that their mental health was nearly identical to those who answered “Yes”. This would have more than doubled the sub population size, and, thus, strengthened the analyzes. However, they were excluded since it might be that these participants had equally poor mental health before the pandemic, and, thus, their screening scores would not reflect a decline during to the pandemic. A strength to consider, though, is that the base population of the HOLD-study is a robust size, N = 260, and recruited from the general primary care population with few exceptions, which indicates that it is a good representation of the age group.