Older People’s Experiences of the Initial Outbreak of the COVID-19 Pandemic in Europe and Highlights for Türkiye

Full Text

After the outbreak of the COVID-19 pandemic at the beginning of 2020, severe economic, social, and health-related issues affected people worldwide. Governments implemented various measures, such as social distancing and lockdowns, to prevent the virus from infecting more people. Older adults, especially those suffering from chronic diseases, were the main target group of these practices. Scholars in Türkiye examined an increase in ageist attitudes during the pandemic (Arun, 2020), the resources that older adults used to gain information about the pandemic, and how older adults overcame loneliness using digital technology (Binark et al., 2020).

Studies based on the Survey of Health, Ageing, and Retirement in Europe (SHARE) COVID-19 data have examined issues that aging scholars are familiar with, such as economic vulnerability and unmet healthcare needs (Arnault et al., 2021), factors affecting the active life (Angelova, 2021), depression and loneliness (Atzendorf & Gruber, 2021). Several studies are unique to the time of the pandemic and practices that emerged from it, such as willingness to be vaccinated (Bergman et al., 2022), precautionary health behaviors (Biro et al., 2021), and the mental health effects of lockdown (Garcia-Prado et al., 2022).

Instead of focusing on a certain aspect of the effect of the COVID-19 pandemic on older people’s lives, we aimed to evaluate how the population aged 65 and above in Europe experienced the initial outbreak in general. Hereby, the objective of this study was to examine 1) housing and independent living, 2) social networks, 3) economic consequences, and 4) support. We aimed to use the findings of this descriptive study to contribute to the ongoing discussion in Türkiye.

Measures and Methods

Data Source and Collection

SHARE is a multidisciplinary study that provides internationally comparable longitudinal microdata. To investigate the impact of the pandemic on health and socioeconomic conditions, a COVID-19 questionnaire was fielded with the longitudinal respondents of the SHARE (SHARE, 2022, p. 41). The SHARE data collection is based on computer-assisted personal interviewing since 2004. However, for the SHARE corona survey, computer-assisted telephone interviews were preferred to collect data to protect respondents and interviewers. There were two rounds of the SHARE corona survey. The first one was conducted immediately after the outbreak (June–July 2020) and the second one after a year (between June and August 2021) to observe developments during the pandemic (SHARE, 2022, p. 6, 7). The first round of the SHARE corona survey data questionnaire was designed to compare pre- and post-pandemic lives, and this study aimed to evaluate the initial experience of the pandemic’s effects.

Study Sample

There were 39,659 respondents aged 65 and over in the sample obtained from 27 European countries: Austria (5.6%), Germany (5.0%), Sweden (3.0%), the Netherlands (1.6%), Spain (4.6%), Italy (7.2%), France (4.0%), Denmark (3.4%), Greece (6.9%), Switzerland (3.9%), Belgium (6.6%), Czech Republic (5.6%), Poland (4.7%), Luxemburg (1.5%), Hungary (2.1%), Portugal (2.2%), Slovenia (6.1%), Estonia (8.4%), Croatia (3.5%), Lithuania (2.0%), Bulgaria (1.4%), South Cyprus (1.6%), Finland (2.4%), Latvia (1.6%), Malta (1.5%), Romania (2.4%), and Slovakia (1.1%)). Furthermore, 43.5% of respondents were male (n = 17,231), and 56.5% of respondents were female (n = 22,419).

Variables

Housing and independent living. Participants were asked whether they had temporarily moved to another place owing to the COVID-19 pandemic or they lived in their usual home and whether they had ever left home since the outbreak. In addition, they were asked how often they a) went shopping, b) visited family members, c) met more than five people, and d) went out for a walk compared with the time before the COVID-19 outbreak (1: “not anymore,” 2: “less often,” 3: “about the same,” and 4: “more often”). We recoded this variable into a binary one—1 indicated “not anymore/less often” and 2 indicated “about the same/more often.”

Social network. We reverse coded the contact frequency with a) own children, b) own parents, c) other relatives, and d) neighbors, friends, or colleagues since the outbreak as 1: “never,” 2: “less often,” 3: “about once a week,” 4: “several times a week,” and 5: “daily.”

Economic consequences. Respondents were asked whether they used their savings, postponed regular bills, and received financial support because of the outbreak. Finally, we recoded households’ ability to make ends meet since the outbreak variable into a binary one—1 indicated “with great difficulty/with some difficulty” and 2 indicated “fairly easily/easily.”

Support. Respondents were asked whether they received help to obtain necessities, helped others obtain necessities, and provided personal care outside the home since the outbreak. Furthermore, they were asked the amount of help they received from a) children, b) parents, c) other relatives, and d) neighbors, friends, or colleagues since the outbreak, which was coded as 1: “less often,” 2: “about the same,” and 3: “more often.”

Analysis

We used frequencies and percentages to summarize categorical variables and the median and interquartile range to summarize Likert-type variables. The Pearson’s chi-square test was employed to determine a relationship between two categorical variables (such as “gender” and “ever left home since the outbreak”). Cramer’s V was given as an effect size of the chi-square test, where 0–0.10 is considered negligible, 0.10–0.20 is weak, 0.20–0.40 is moderate, 0.40–0.60 is relatively strong, 0.60–0.80 is strong, and 0.80–1.00 is very strong. The observed count and expected values were checked for the chi-square test assumption. The Mann–Whitney U test was used to compare differences between two independent groups (such as gender) when the dependent variable was ordinal (such as the contact frequency with social networks) because the chi-square test ignores the ordinality. We used the point-biserial correlation (r_p) as an effect size for the Mann–Whitney U test where 0.01–0.09 is considered negligible, 0.10–0.29 is low, 0.30–0.49 is moderate, 0.50–0.69 is substantial, and 0.70 or higher is very strong. Bar charts were used to show the amount of help received from social networks. SPSS 21 was used, and p < 0.05 was considered statistically significant.

 Findings

We aimed to investigate the housing and independent living among the population aged 65 and above after the initial outbreak.

Table 1

Crosstabulations of Housing and ever Left Home Since Outbreak and Gender

		Male	Female
Usual home or moved temporarily	Usual home (n= 38818)	16894 (98.1)	21924 (97.8)
	Moved temporarily (n= 824)	331 (1.9)	493 (2.2)
Ever left home since outbreak	Yes (n= 31087)	14122 (82.0)	16965 (75.7)
	No (n= 8539)	3101 (18.0)	5438 (24.3)

The cells represent frequencies with percentages in parenthesis; n: sample size.

There was no significant association between gender and whether the respondent moved temporarily since the outbreak or lived at the usual home with a negligible effect size (χ2 (1) = 3.69, p = 0.06, Cramer’s V = 0.010).

There was a significant association between gender and whether the respondent ever left home since the outbreak with a negligible effect size (χ2 (1) = 226.33, p < 0.001, Cramer’s V = 0.076). Based on the odds ratio, the odds of male respondents (82.0%, n = 14,122) ever leaving home since the outbreak was 1.46 times higher than that of female respondents (75.7%, n = 16,965).

Table 2

Crosstabulations of Daily Activities since the Outbreak and Gender

		Male	Female
Went shopping	Not anymore/less often	8987 (64.8)	12766 (76.1)
	About the same/more often	4885 (35.2)	4014 (23.9)
Walking outside	Not anymore/less often	5977 (43.3)	8235 (49.7)
	About the same/more often	7822 (56.7)	8344 (50.3)
Met more than five people	Not anymore/less often	12144 (88.3)	15101 (91.3)
	About the same/more often	1607 (11.7)	1430 (8.7)
Visited family members	Not anymore/less often	11409 (82.9)	14248 (86.1)
	About the same/more often	2352 (17.1)	2304 (13.9)

The cells represent frequencies with percentages in parentheses.

There was a significant association between gender and the respondent going shopping whether “not anymore/less often” or “about the same/more often” since the outbreak with a weak effect size (χ2 (1) = 470.10, p < 0.001, Cramer’s V = 0.124). Based on the odds ratio, the odds of female respondents going shopping “not anymore/less often” was 1.73 times higher than that of male respondents.

There was a significant association between gender and the respondent going out for a walk whether “not anymore/less often” or “about the same/more often” since the outbreak with a negligible effect size (χ2 (1) = 122.22, p < 0.001, Cramer’s V = 0.063). Based on the odds ratio, the odds of female respondents going out for a walk “not anymore/less often” was 1.29 times higher than that of male respondents.

There was a significant association between gender and the respondent meeting more than five people whether “not anymore/less often” or “about the same/more often” since the outbreak with a negligible effect size (χ2 (1) = 76.68, p < 0.001, Cramer’s V = 0.050). Based on the odds ratio, the odds of female respondents meeting more than five people “not anymore/less often” was 1.40 times higher than that of male respondents.

There was a significant association between gender and the respondent visiting family members whether “not anymore/less often” or “about the same/more often” since the outbreak with a negligible effect size (χ2 (1) = 58.15, p < 0.001, Cramer’s V = 0.044). Based on the odds ratio, the odds of female respondents visiting family members “not anymore/less often” was 1.28 times higher than that of male respondents.

Table 3

Contact Frequency with the Social Networks since the Outbreak and Gender

	Male (n= )	Female (n= )	p	Effect Size (rp)
With own children, Md (IQR) Min; Max	3 (2) 1; 5	3 (2) 1; 5	<0.001	-0.03
With own parents Md (IQR) Min; Max	1 (1) 1; 5	1 (1) 1; 5	0.26	− 0.02
With other relatives Md (IQR) Min; Max	2 (1) 1; 5	2 (1) 1; 5	<0.001	-0.03
With neighbors, friends, and colleagues Md (IQR) Min; Max	2 (2) 1; 5	2 (2) 1; 5	<0.001	− 0.05

Md: Median; IQR: interquartile range; n: sample size; p: Mann–Whitney U test p value; r_p: point-biserial correlation.

The results of the Mann–Whitney U test revealed that the contact frequency with own children since the outbreak for female respondents (Md = 3) was significantly higher than that of male respondents (Md = 3) according to mean ranks with a negligible effect size (U = 167113197.50, z = −5.17, p < 0.001, r_p = −0.03).

The results of the Mann–Whitney U test revealed that the contact frequency with own parents since the outbreak for female respondents (Md = 1) did not differ significantly from that of male respondents (Md = 1, U = 2554117.0, z = −1.127, p = 0.26, r_p = −0.02).

The results of the Mann–Whitney U test revealed that the contact frequency with other relatives since the outbreak for female respondents (Md = 2) was significantly higher than that of male respondents (Md = 2) according to mean ranks with a negligible effect size (U = 170695498.00, z = −5.78, p < 0.001, r_p = −0.03).

The results of the Mann–Whitney U test revealed that the contact frequency with neighbors, friends, and colleagues since the outbreak for female respondents (Md = 2) was significantly higher than that of male respondents (Md = 2) according to mean ranks with a negligible effect size (U = 176228772.00, z = −9.62, p < 0.001, r_p = −0.05).

Table 4

Crosstabulation of the Household’s Ability to make Ends Meet since the Outbreak and Gender

		Male (n= 10099)	Female (n= 17276)
Household’s ability to make ends meet since outbreak	with great/some difficulty (n= 9343)	3039 (30.1)	6304 (36.5)
	fairly easily/easily (n= 18032)	7060 (69.9)	10972 (63.5)

The cells represent frequencies with percentages in parentheses; n: sample size.

There was a significant association between gender and the household’s ability to make ends meet since the outbreak whether “with great/some difficulty” or “fairly easily/easily” with a negligible effect size (χ2 (1) = 116.04, p < 0.001, Cramer’s V = 0.065). Based on the odds ratio, the odds of female respondents (36.5%, n = 6,304) having “some degree of difficulties to make ends meet” was 1.34 times higher than that of male respondents (30.1%, n = 3,039).

Table 5

Crosstabulations of Economic Effects of the Outbreak and Gender

		Male	Female
Dipped into savings since outbreak	Yes (n= 1419)	470 (15.5)	949 (15.2)
	No (n= 7868)	2556 (84.5)	5312 (84.8)
Postponed regular bills since outbreak	Yes (n= 667)	261 (8.6)	406 (6.5)
	No (n= 8650)	2770 (91.4)	5880 (93.5)
Respondent received financial support due to outbreak	Yes (n= 1465)	458 (4.6)	1007 (5.9)
	No (n= 25685)	955 (95.4)	16130 (94.1)

The cells represent frequencies with percentages in parentheses; n: sample size.

There was a significant association between gender and whether the respondent postponed bills since the outbreak with a negligible effect size (χ2 (1) = 14.25, p < 0.001, Cramer’s V = 0.039). Based on the odds ratio, the odds of male respondents postponing bills since the outbreak was 1.37 times higher than that of female respondents.

There was not a significant association between gender and whether the respondent used savings since the outbreak with a negligible effect size (χ2 (1) = .22, p = 0.64, Cramer’s V = 0.005).

There was a significant association between gender and whether the respondent received financial support because of the outbreak with a negligible effect size (χ2 (1) = 20.99, p < 0.001, Cramer’s V = 0.028). Based on the odds ratio, the odds of female respondents receiving financial support because of the outbreak was 1.30 times higher than that of male respondents.

Table 6

Crosstabulations of Support since the Outbreak and Gender

		Male	Female
Help received to obtain necessities since the outbreak	Yes (n= 13898)	4756 (27.6)	9142 (40.8)
	No (n= 25703)	12456 (72.4)	13247 (59.2)
Helped others to obtain necessities since the outbreak	Yes (n= 4308)	1922 (11.2)	2386 (10.7)
	No (n= 35290)	15290 (88.8)	20000 (89.3)
Provided personal care outside home since outbreak	Yes (n= 794)	286 (1.7)	508 (2.3)
	No (n= 38809)	16928 (98.3)	21881 (97.7)

The cells represent frequencies with percentages in parentheses; n: sample size.

There was a significant association between gender and whether the respondent received help to obtain necessities since the outbreak with a weak effect size (χ2 (1) = 744.44, p < 0.001, Cramer’s V = 0.137). Based on the odds ratio, the odds of female respondents (40.8%, n = 9,152) receiving help to obtain necessities since the outbreak was 1.81 times higher than that of male respondents (27.6%, n = 4,756).

There was not a significant association between gender and whether the respondent helped others obtain necessities since the outbreak with a negligible effect size (χ2 (1) = 2.59, p = 0.11, Cramer’s V = 0.008).

There was a significant association between gender and whether the respondent provided personal care outside the home since the outbreak with a negligible effect size (χ2 (1) = 18.28, p < 0.001, Cramer’s V = 0.021). Based on the odds ratio, the odds of female respondents (2.3%, n = 508) providing personal care outside the home since the outbreak was 1.37 times higher than that of male respondents (1.7%, n = 286).

 Discussion and Highlights for Türkiye

This study aimed to examine the effect of the first wave of the COVID-19 pandemic in Europe on older people’s lives in terms of 1) housing and independent living, 2) social networks, 3) economic consequences, and 4) support using the SHARE COVID-19 data. There were 39,659 respondents aged 65 and over in the sample obtained from 27 European countries (43.5% male and 56.5% female). The findings illustrate the pandemic’s initial effects on older people’s lives because the data were collected between June and July 2020.

The results revealed that only a small group moved temporarily owing to the outbreak, whereas most people aged 65 and over lived in their usual homes. Considering that daily activities required an independent life, the results showed that the least affected activity was going out for a walk for both genders. For other activities, the results demonstrated that social life (visiting other family members and meeting people) was the most negatively affected. Only a small group participated in social activities as often as they used to before the outbreak. The study’s findings concerning social activities are as expected because the first wave of the pandemic was characterized by obligatory lockdowns and social distancing rules for all age groups. Notably, a group (18.0% of male respondents and 24.3% of women) never left home after the outbreak, which is an essential finding for policymakers, particularly regarding more assistance that this group needed to live a healthy life during a disaster than other age groups. In the case of Türkiye, the results of 2015 Active Ageing Research indicated that 23.6% of the population aged 65 and above reported being dependent on others for daily shopping and going outdoors at least to some degree and 9.9% of them reported being fully dependent (Özmete, 2016, p. 87).

Findings about housing and independent living allow us to examine how aging-in-place practices can be affected by disasters that do not necessarily affect the physical environment; more precisely, those are disasters other than earthquakes and floods. Our findings indicated that European older adults lived in their usual homes. This raises a question about how well prepared local authorities and communities are to help and support older adults in times of a disaster. Community-based organizations are critical partners in local management. In their study, Pendergrast and colleagues (2021, p. 15) showed that local institutional actors played influential roles in supporting older adults’ disaster resilience. According to the latest research on family structure in Türkiye (2022), which shows the results of data obtained from individuals aged 65 and above in Türkiye, the main preference for living arrangements in old age when the respondents become extremely old to take care of themselves is having homecare while living in their own household (31.6%). Only 10.3% of respondents prefer moving to a senior center, whereas 46.0% prefer living with their children (TurkStat Research on Family Structure in Türkiye, 2022). This demonstrates the significance of supporting aging-in-place practices in times of a disaster in the case of Türkiye as well.

In terms of social networks, the results indicated that the contact frequency with children was the highest among social network types, whereas other types were less frequent. In the case of Türkiye, the data of the research on family structure in Türkiye (2022), which was collected in August–November 2021, shows that 56.7% of individuals aged 60 and above in Türkiye are visited by their children several times a week, whereas 1.4% of them are never visited by their children (TurkStat Research on Family Structure in Türkiye, 2022). Herein, we should add for further evaluation that 96.6% of individuals aged 65 and above in Türkiye have children (Şentürk, 2019, p. 275).

In terms of care and support, our results showed that 40.8% women and 27.6% men respondents received help to obtain necessities since the outbreak. Furthermore, the results showed that 2.3% women and 1.7% men respondents provided personal care outside the home. The bar charts in Appendix 1 suggest that the support needs of the population aged 65 and above were mainly addressed by their children for both men and women. Moreover, 23.25% men and 43.59% women respondents who received help from their children obtained this support more often than in the pre-COVID times. The role of children in support reveals a challenge for older adults without children.

In Türkiye, 11.0% of old-aged women who live alone and 3.9% of men required care in 2016 (Arun & Holdsworth, 2019, p. 343). In addition to this fragile group, policymakers should consider the increasing number of nuclear families that do not have any children (Arun & Holdsworth, 2019, p. 352). The change in the family structure in Türkiye shows that the number of childless nuclear families has increased more rapidly than the number of nuclear families with children (Koç, 2019, p. 36). Their care and support represent another challenge for policymakers, particularly during disasters.

One aspect to highlight in Türkiye’s case is the geographical distance, which plays a key role in intergenerational support. Both older people and their offsprings are concerned about its negative effect on support (Özmete, 2017). Thus, Şentürk (2019, p. 303) suggested that policies that maintain housing for individuals aged 65 and above within walking distance from their children and relatives in Türkiye are vital. This idea is based on the living arrangement preferences in old age in Türkiye as well as the role of families in support and care practices.

In times of disaster, social capital is known as a critical source of resilience. Scholars explained the differences among neighborhoods of Kobe—a Japanese city that was hit by a devastating earthquake—in terms of preparedness and speed of response by social capital (Kawachi & Berkman, 2014, p. 290): “In the immediate aftermath of earthquake, local neighborhood associations assisted in rescue operations, helped to evacuate homeless residents to nearby schools, established community kitchens, and organized night watchmen to guard abandoned property.” Furthermore, 71% of people who lost their lives in Hurricane Katrina were over 60 years of age and most died in their homes and communities (Gibosn, 2006, as cited in Pendergrast et al., 2020, p. 1). Along with climate change, disasters have become more frequent and severe. Hence, disaster resilience for all age groups is an urgent public health issue.

It is known that older people experienced a decline in income less often than the younger population because pensions were a stable source of income (Eurofound, 2022, p.1). Our examination of the economic effects of the first wave of the outbreak showed that 36.5% of women and 30.1% of men in our sample had a hard time making ends meet in their households. There were 8.6% of men and 6.5% of women who postponed their bills. Approximately 15% of men and women used their savings after the first wave of the pandemic. Moreover, 5.9% of female and 4.6% of male respondents received financial support because of the outbreak.

Gender differences in the older population are important for policymakers in Türkiye for three reasons. First, considering the poverty rate among the population aged 65and above by sex in Türkiye, more older women suffer from poverty than men. In 2020, the poverty rate for older men was 15.6%, whereas it was 17.6% for older women according to the TurkStat Income and Living Conditions Survey (cited in TurkStat Elderly Statistics, 2021). Second, we also know that older women in Türkiye are less educated than their male peers. In a 2020 statistics, while 24.2% of older women were illiterate and 19.9% of them did not complete school, these percentages for men were 4.7% and 8.7%, respectively (TurkStat, National Education Statistics Database, 2016-2020 cite in TurkStat Elderly Statistics, 2021). Although the disadvantages in education have increasingly diminished every year (the population rate of the older females who were illiterate and those who did not complete school was 31.1% in 2016, and it has decreased gradually), there is still a long way to go until reaching equality in education. One last thing to highlight is the older population who lives alone in Türkiye. In 2021, in 24.1% of total households, there was at least one older person. Considering the proportion of one-person elderly households in total one-person households, it was 32.7%, indicating that a considerable part of the population aged 65 and above lives alone. Moreover, there are sex differences, as 74.9% of one-person elderly households comprise women. Thus, most individuals aged 65 and above who live alone are women (TurkStat Elderly Statistics 2021). This combination of poverty, educational disadvantages, and living alone puts a new perspective on supporting older people in Türkiye in times of a disaster. The aftermath of the COVID-19 pandemic showed that a more resilient future depends on inclusive policy choices that governments make.

Finally, the SHARE’s data collection experience during the time of an outbreak of a disaster demonstrates that a panel study can collect data even in extraordinary times because the SHARE included only the longitudinal respondents to the COVID-19 survey. Panel research that targets the population aged 50 and above in Turkey is important for examining and understanding the needs of this age group in times of disasters. In 2021, the proportion of the population aged 65 and above in Türkiye was 9.7%, 17.3% in the USA, 22.6% in Germany, 20.3% in Bulgaria, 20.1% in Denmark, 22.6% in Finland, 20.8% in France, 20.1% in the Netherlands, 19.9% in Spain, 20.5% in Sweden, 18.9% in Switzerland, 22.8% in Italy, and 28.8% in Japan, according to TurkStat Elderly Statistics 2021 and United States Census Bureau International Database. Although Türkiye still has a younger population than these countries, population projections show that it is an aging society; thus, we must plan our future before it turns into a crisis. Considering the old age dependency ratio, there was an increase from 7.1% in 1935 to 14.3% in 2021 (TurkStat Elderly Statistics, 2021). Before the COVID-19 pandemic, countries, such as Japan, characterized by an aging society for a long time and experiences of natural disasters, such as earthquakes (Junko, 2012), considered long-term challenges of an aging population in times of disasters. After the pandemic, identifying and applying effective strategies to support older adults in disasters has become a crucial topic for Türkiye along with the rest of the world.

Notes. This study used data from the SHARE Wave 8. COVID-19 Survey 1, DOI: 10.6103/SHARE.w8ca.800; see Börsch-Supan for methodological details. The SHARE data collection was funded by the European Commission, DG RTD through FP5 (QLK6-CT-2001-00360), FP6 (SHARE-I3: RII-CT-2006-062193, COMPARE: CIT5-CT-2005-028857, SHARELIFE: CIT4-CT-2006-028812), FP7 (SHARE-PREP: GA N°211909, SHARE-LEAP: GA N°227822, SHARE M4: GA N°261982, DASISH: GA N°283646), Horizon 2020 (SHARE-DEV3: GA N°676536, SHARE-COHESION: GA N°870628, SERISS: GA N°654221, SSHOC: GA N°823782, SHARE-COVID19: GA N°101015924), and DG Employment, Social Affairs & Inclusion through VS 2015/0195, VS 2016/0135, VS 2018/0285, VS 2019/0332, and VS 2020/0313. Additional funding from the German Ministry of Education and Research, the Max Planck Society for the Advancement of Science, the U.S. National Institute on Aging (U01_AG09740-13S2, P01_AG005842, P01_AG08291, P30_AG12815, R21_AG025169, Y1-AG-4553-01, IAG_BSR06-11, OGHA_04-064, HHSN271201300071C, RAG052527A), and various national funding sources is gratefully acknowledged (see www.share-project.org).