Trajetória de diminuição da solidão entre brasileiros com 50 anos ou mais durante a pandemia de COVID-19: ELSI-Brasil

Cad Saude Publica csp Cadernos de Saúde Pública Cad. Saúde Pública 0102-311X 1678-4464 Escola Nacional de Saúde Pública Sergio Arouca, Fundação Oswaldo Cruz Este estudo teve como objetivo estimar a prevalência de solidão entre idosos brasileiros nos primeiros sete meses da pandemia de COVID-19 e identificar os preditores das trajetórias de solidão, usando dados pré-pandemia oriundos de entrevistas presenciais de participantes do Estudo Longitudinal da Saúde dos Idosos Brasileiros (ELSI-Brasil) de 2019-2020, um estudo de representatividade nacional com residentes da comunidade com 50 anos ou mais. Os dados durante a pandemia foram coletados em três rodadas de entrevistas telefônicas com os participantes, realizadas de maio a outubro de 2020. A solidão foi medida por uma questão de item único, considerando os casos com pelo menos duas medidas repetidas. As variáveis explicativas incluíram depressão, morar sozinho, sair de casa na última semana e conexão virtual no último mês. A regressão logística de efeitos mistos foi utilizada para estimar as razões de chances com seus intervalos de 95% de confiança (IC95%) e investigar trajetórias de solidão e seus preditores. Foram incluídos 5.108 participantes. A prevalência global de solidão no período pré-pandemia foi de 33,1% (IC95%: 29,4-36,8), um valor superior ao período pandêmico (rodada 1: 23,6%, IC95%: 20,6-26,9; rodada 2: 20,5%, IC95%: 17,8-23,5; rodada 3: 20,6%, IC95%: 17,1-24,6). Uma interação significativa (p ≤ 0,05) foi encontrada apenas entre depressão e tempo; participantes com depressão apresentaram maior redução dos níveis de solidão. Embora os níveis de solidão no Brasil tenham diminuído durante a pandemia, esse padrão não se aplica a todos os idosos. Indivíduos com depressão tiveram uma redução mais significativa provavelmente por se sentirem mais próximos aos membros de suas redes sociais durante as recomendações de ficar em casa. Introduction The World Health Organization (WHO) recognizes loneliness as a major public health problem 1. Loneliness can be defined as the negative subjective feeling that results from a discrepancy between desired and actual social connections 2,3. Estimates have shown that over 20% of older adults in the United States 4, Eastern Europe 5, India 6, and six Latin American countries 7 reported feeling lonely. Due to the coronavirus disease 2019 (COVID-19), predictions that stay-at-home recommendations and social restrictions would negatively affect a broad set of mental health indicators were expected, including a rise in loneliness 8,9,10. Among community-dwelling older adults, longitudinal data from high-income countries, such as Austria 11, the Netherlands 12, Norway 8, and Switzerland 13 confirmed higher loneliness levels during the pandemic, in comparison with the pre-pandemic period. However, in the United Kingdom, similar proportions of respondents reported that they felt less lonely (13.6%) or lonelier (11.1%) during the pandemic 14. Data from the United States have shown mixed findings, with studies reporting either increased 14,15 or relatively stable loneliness levels during the pandemic 10,16. For older adults, predictors of increased loneliness trajectories after the pandemic onset include being a woman 14, older age, being single 12,17, living alone, poor self-rated health 14, having higher psychological distress 17, and less financial satisfaction 14. Although unexpected, having higher social support was also identified as a predictor of increased loneliness trajectories during the COVID-19 pandemic 17. As social support was measured by the number of people close to the participants previous to the onset of the pandemic, perhaps social distancing is comparatively more harmful for people with more social relationships 17. On the other hand, good physical functioning before the pandemic, high mastery 12, higher frequency of face-to-face contact outside the home 14, and satisfaction with the frequency of social interactions during the COVID-19 lockdown provided some protection against an increase in loneliness 13. Mixed results have been found regarding virtual contact (i.e., how often the individual interacts with children, other family members, and friends who live apart via telephone, email, and social media) 14,16. Albeit many people have resorted to virtual communication to maintain relationships with family members and friends who live apart, these controversial findings may suggest that interactions via telephone, video calls, text messaging, and social media may not be a qualitatively equivalent alternative to face-to-face contact. Firstly, digital media use depends on access to the internet and technological knowledge, which are often unevenly distributed across demographic and socioeconomic groups 14,18. Besides, the most common means of virtual contact among older adults - telephone calls and text messaging - are generally unable to simulate face-to-face contact due to their lack of visualization 14,19. Furthermore, digital media use can cause stress, and both digital stress and avoidance of use tend to be greater among older adults 14,20. Thus, virtual contact may have a limited role in maintaining older adults’ mental health, showing inconsistent benefits across groups. In low- and middle-income countries, little information is available on changes in loneliness levels in this age group since the onset of the pandemic. Although Chile has shown stable levels from the end of 2019 to September 2020 21, loneliness in Brazil has reduced from 32.8% (pre-pandemic) to 23.9% by June 2020 22. Predictors of loneliness trajectories remain not adequately explored. Brazil has been one of the most severely affected countries due to the COVID-19 pandemic 23. The community transmission throughout the country was declared in March 2020 by the Brazilian Ministry of Health, but the Federal Government did not issue a nationwide stay-at-home order, and the local governments imposed social restriction measures at different levels 24. By December 2021, the country had registered more than 22 million confirmed cases, and deaths had exceeded 614,000 25. A decline in 2020 life expectancy at birth of 1.3 years and 0.9 years in life expectancy at age 65 was estimated, setting Brazil back to 2014 and 2012 levels, respectively 23. Longitudinal changes in loneliness may differ across countries where different social restriction measures were introduced to face the COVID-19 pandemic. Also, little attention has been given to loneliness trajectories and their predictors, particularly in low- and middle-income countries with a high number of cumulative confirmed cases and deaths due to COVID-19. As loneliness damages both mental and physical health 1, and presents a similar mortality risk to well-established risk factors, such as smoking, physical inactivity, and obesity 26, its monitoring is crucial to the establishment of strategies and policies that could mitigate its harmful effects. Moreover, evidence shows that loneliness is associated with more visits to health care provider, increased hospitalization and readmission rates, and longer length of stay 27. Thus, this study aimed to estimate prevalence of loneliness among older Brazilian adults over the first seven months of the pandemic and to identify the predictors of loneliness trajectories. Method Data and sample The Brazilian Longitudinal Study of Aging (ELSI-Brazil) is a nationally representative study of persons aged 50 years and older, conducted in 70 municipalities in the five regions of Brazil 28. Pre-pandemic data for this analysis were obtained from the second wave of the ELSI-Brazil, conducted by face-to-face interview from August 2019 to March 2020, with 9,411 participants 29. This pre-pandemic survey was considered the baseline for this study. Data of the pandemic period were collected in 2020 from three rounds of telephone interviews, conducted in May-June (first round), July-August (second round), and September-October (third round). Detailed information on the ELSI-Brazil study and the telephone surveys design can be found elsewhere 28,30. Among baseline participants, 6,149 responded to the first round of the telephone survey; 6,752 responded to the second round; and 6,711 to the third. As loneliness is a subjective feeling, only the participants who answered the surveys without a proxy and those who had at least two repeated measures for loneliness (at baseline and at least one round) were included in this study. Thus, 5,108 participants (50 to 118 years), summing up to 20,432 interviews composed our sample. Figure 1 shows the study timeline and the national context regarding the number of cumulative confirmed cases in Brazil due to the COVID-19 pandemic at each time point. Figure 1 Study timeline and number of cases due to the COVID-19 pandemic in Brazil. Brazilian Longitudinal Study of Aging (ELSI-Brazil). Source: Mathieu et al. 25. Note: the figure reports the number of cumulative confirmed cases. The ELSI-Brazil (protocol 34649814.3.0000.5091) and the telephone surveys (protocol 34649814.3.0000.5091) were approved by the Research Ethics Committee of the Rene Rachou Institute, Oswaldo Cruz Foundation. Measures Loneliness was measured by the single-item question: “How often do you feel alone/lonely?”. For the telephone surveys, respondents were asked to rate the item referring to the last 30 days. Responses were scored on a three-point Likert scale: hardly ever/never, sometimes, or often. Following previous research, we created a dichotomous measure of loneliness (lonely versus not lonely) to aid in the results interpretation, coding persons who responded “sometimes” or “often” as lonely 31. Explanatory variables included the self-reported medical diagnosis of depression (yes or no), living alone (yes or no), leaving home in the last week (6 to 7 days, 3 to 5 days, 1 to 2 days, or did not leave home), and virtual connectedness in the last month (3 days or more per week, at least 1 day per week, less than 1 day per week, or no virtual connectedness). Leaving home was evaluated by the question: “In the last week, how many days did you leave home?”. Virtual connectedness was assessed by the frequency of contact with relatives and friends via telephone, Skype, WhatsApp, or social media in the past 30 days, considering those who did not live with the participant. Data on the medical diagnosis of depression and living arrangement were measured only at baseline whereas leaving home and virtual connectedness variables were assessed only during the telephone survey rounds. Covariates included age group (50-59, 60-69, 70-79, or ≥ 80), sex (female or male), schooling years (0-4, 5-8, 9-11, or ≥ 12), and place of residence (urban or rural). Statistical analyses Frequency distributions were performed to characterize the study population at baseline. Firstly, the prevalence of loneliness for the total population (overall prevalence) was estimated. Then, loneliness prevalence according to demographic, health, and social variables at each time point (baseline and in the three rounds performed during the COVID-19 pandemic) was calculated. Mixed-effects logistic regression was used to estimate odds ratios (OR) with their 95% confidence intervals (95%CI) and to investigate loneliness trajectory and its predictors (depression, living alone, leaving home, and virtual connectedness). Separate models for each explanatory variable were fitted, incorporating the time, then the time plus adjustment variables (age group, sex, schooling years, and place of residence), and, finally, an interaction term between time and explanatory variable (adjusted model + time interaction). Time was included in the models as categories (baseline, rounds 1, 2, and 3). The models included a random intercept and a random slope for time within-participant to account for the correlation between observations. Lastly, interaction terms were presented in plots. All analyses were performed using Stata 16.0 (https://www.stata.com). The svy command was used to account for the complex survey design and survey weights, which were used to compensate for unequal selection probabilities and differential nonresponse to telephone surveys. Significance level was set at 5%. Results At baseline, 51% of participants were female, 50% were aged 50-59 years, and 88.2% lived in urban areas. The most frequent education group reported ≥ 12 years of schooling (37.3%). Furthermore, 86.5% of participants reported no medical diagnosis of depression and 77.6% declared not living alone. Table 1 shows further details of the prevalence of loneliness according to demographic, health, and social variables at baseline and in the three rounds performed during the COVID-19 pandemic. Figure 2 shows the changes in loneliness prevalence over time. In the pre-pandemic period, the overall loneliness prevalence was 33.1% (95%CI: 29.4-36.8). It was higher than the prevalence observed in the three rounds carried out during the pandemic (round 1: 23.6%, 95%CI: 20.6-26.9; round 2: 20.5%, 95%CI: 17.8-23.5; round 3: 20.6%, 95%CI: 17.1-24.6). Table 1 Prevalence of loneliness according to demographic, health, and social variables at baseline and in three rounds performed during the COVID-19 pandemic. Brazilian Longitudinal Study of Aging (ELSI-Brazil). Characteristics Baseline (n = 5,108) Round 1 (n = 4,108) Round 2 (n = 4,471) Round 3 (n = 4,206) % 95%CI % 95%CI % 95%CI % 95%CI Age group (years) 50-59 34.3 29.1-39.7 23.7 19.4-28.6 18.8 15.1-23.3 19.7 15.1-25.1 60-69 29.6 25.0-34.7 20.1 16.6-24.1 20.1 17.1-23.4 21.1 16.0-27.3 70-79 31.4 26.3-37.1 26.0 21.7-30.6 22.4 17.7-27.8 19.6 14.7-25.8 ≥ 80 45.2 35.0-55.7 34.0 23.2-46.8 33.8 22.7-47.0 30.5 21.6-41.1 Sex Female 39.3 35.1-43.7 27.6 23.9-31.6 24.4 21.3-27.9 23.2 20.0-26.8 Male 26.5 22.6-30.8 19.4 15.9-23.5 16.3 12.6-20.7 17.8 13.0-24.0 Schooling (years) 0-4 35.6 31.6-39.8 25.1 21.8-28.8 21.8 19.1-24.8 24.5 21.8-27.6 5-8 35.7 30.9-40.9 26.4 20.6-33.2 24.8 20.6-29.6 21.2 17.0-26.1 9-11 30.2 25.9-34.8 21.5 17.9-25.5 16.6 13.5-20.3 17.5 13.8-21.9 ≥ 12 30.2 24.0-37.3 21.2 16.3-27.1 17.8 12.7-24.3 18.9 12.8-26.8 Place of residence Urban 33.3 29.6-37.1 23.5 20.3-27.0 20.9 17.9-24.2 20.8 17.0-25.2 Rural 31.5 24.9-38.8 24.4 19.2-30.5 17.1 13.2-21.9 19.2 12.4-28.5 Depression No 29.7 26.0-33.8 22.4 19.1-26.1 19.3 16.4-22.6 19.1 15.2-23.7 Yes 54.1 45.6-62.4 32.1 24.4-41.3 28.3 21.1-36.8 30.6 23.6-38.7 Living alone No 30.7 27.1-34.4 22.7 19.8-26.0 19.6 16.6-22.9 19.5 16.1-23.4 Yes 42.3 34.4-48.4 26.6 20.1-34.3 23.6 17.7-30.7 24.1 17.5-32.3 Leaving home in the last week (days) 6-7 - - 17.1 12.6-22.9 14.3 9.0-22.0 16.5 10.6-24.6 3-5 - - 24.0 17.8-31.4 20.0 15.1-25.9 21.3 14.9-29.5 1-2 - - 23.4 18.3-29.4 20.7 17.4-24.5 22.3 17.0-28.7 Did not leave home - - 28.2 23.8-33.0 25.8 21.0-31.2 22.5 17.9-27.8 Virtual connectedness in the last month 3 days or more per week - - 22.3 19.2-25.9 20.7 17.3-24.5 19.0 15.4-23.2 At least 1 day per week - - 23.4 17.9-29.9 18.0 14.1-22.8 24.3 16.0-35.1 Less than 1 day per week - - 35.4 23.7-49.2 25.1 16.5-36.1 22.6 14.6-33.3 No virtual connectedness - - 26.7 19.7-35.0 23.8 17.2-31.9 23.5 16.0-33.3 95%CI: 95% confidence interval. Figure 2 Prevalence of loneliness before (baseline) and during the COVID-19 pandemic (rounds 1, 2, and 3). Brazilian Longitudinal Study of Aging (ELSI-Brazil). Table 2 shows the results of the adjusted associations of each explanatory variable with loneliness. The fully adjusted models showed a positive association of loneliness with depression (OR = 2.49, 95%CI: 1.62-3.82), living alone (OR = 1.48, 95%CI: 1.04-2.10), and virtual connectedness in the last month - less than once per week (OR = 2.01, 95%CI: 1.19-3.39) and no virtual connectedness (OR = 1.54, 95%CI: 1.03-2.31). Table 2 Adjusted models of the association between time, depression, living alone, leaving home, virtual connectedness, and loneliness. Brazilian Longitudinal Study of Aging (ELSI-Brazil). Characteristics OR * 95%CI OR ** 95%CI Time Baseline 1.00 1.00 Round 1 0.50 0.37-0.67 0.50 0.37-0.67 Round 2 0.30 0.21-0.42 0.30 0.21-0.41 Round 3 0.21 0.12-0.38 0.21 0.12-0.36 Depression No 1.00 1.00 Yes 2.86 1.83-4.47 2.49 1.62-3.82 Living alone No 1.00 1.00 Yes 1.63 1.15-2.31 1.48 1.04-2.10 Leaving home in the last week (days) 6-7 1.00 1.00 3-5 1.59 0.99-2.56 1.48 0.95-2.30 1-2 1.53 0.81-2.89 1.30 0.72-2.32 Did not leave home 1.99 1.13-3.49 1.55 0.95-2.52 Virtual connectedness in the last month 3 days or more per week 1.00 1.00 At least 1 day per week 1.17 0.74-1.85 1.16 0.74-1.84 Less than 1 day per week 1.94 1.16-3.26 2.01 1.19-3.39 No virtual connectedness 1.53 1.05-2.23 1.54 1.03-2.31 95%CI: 95% confidence interval; OR: odds ratio. * All variables adjusted by time; ** Adjusted by time, age group, sex, schooling years, and place of residence. Finally, the adjusted associations of loneliness trajectory with each explanatory variable were separately tested. A significant interaction term was observed only between time and depression. Specifically, at baseline, participants with depression had a significantly higher predicted probability of loneliness compared to those without depression. However, no significant difference was observed in the predicted probability of loneliness between participants with and without depression in rounds 1, 2, and 3 (Figure 3). Figures 4, 5, and 6 show the non-significant interactions between loneliness trajectory and living alone, leaving home in the last week, and virtual connectedness in the last month, respectively. Figure 3 Predicted probability of loneliness before (baseline) and during the COVID-19 pandemic (rounds 1, 2, and 3), according to depression. Brazilian Longitudinal Study of Aging (ELSI-Brazil). Note: adjusted by time, age group, sex, schooling years, place of residence, and interaction time*depression. Figure 4 Predicted probability of loneliness before (baseline) and during the COVID-19 pandemic (rounds 1, 2, and 3), according to living alone. Brazilian Longitudinal Study of Aging (ELSI-Brazil). Note: adjusted by time, age group, sex, schooling years, place of residence, and interaction time*living alone. Figure 5 Predicted probability of loneliness during the COVID-19 pandemic (rounds 1, 2, and 3), according to the frequency with which participants left home in the last week. Brazilian Longitudinal Study of Aging (ELSI-Brazil). Note: adjusted by time, age group, sex, schooling years, place of residence, and interaction time*leaving home last week. Figure 6 Predicted probability of loneliness during the COVID-19 pandemic (rounds 1, 2, and 3), according to virtual connectedness in the last month. Brazilian Longitudinal Study of Aging (ELSI-Brazil). Note: adjusted by time, age group, sex, schooling years, place of residence, and interaction time*virtual connectedness. Discussion Our findings revealed a decrease in loneliness levels among older Brazilian adults, decreasing from 33.1% before pandemic to 20.5% during the first four months of the pandemic and then leveling off until October 2020. Depression was the only predictor of change in loneliness levels over the analyzed period. Participants with depression in the pre-pandemic period showed a greater reduction in loneliness levels during the pandemic compared to those without depression. Unlike previously published studies 10,11,12,13,14,15,16,17,21, to our knowledge only Brazil and Spain have observed decreased loneliness among older adults since the onset of the pandemic 32. However, Spain reported loneliness reduction as an immediate effect of the nationwide lockdown due to COVID-19, assessed 10 days after the government imposed strict stay-at-home orders. Furthermore, although middle-aged and older Spaniards have been included in the sample (mean age = 55.7), no information on the age range was available. Although unexpected, our findings show consistency over time, once loneliness prevalence was similar across the three-time point analysis carried out during the pandemic. A recent systematic review has recognized depression as a longitudinal predictor of loneliness, along with depressed mood, increase in depression, not being married/partnered, partner loss, limited social network, low level of social activity, and poor self-rated health 33. Although these predictors positively explain the levels of loneliness, they do not necessarily explain the loneliness changes over time, and predictors of loneliness trajectories remain rather unexplored. Our results indicated depression as the only predictor of loneliness trajectory. Although it also seems unexpected that older Brazilian adults with depression would report the greatest reduction in loneliness, some hypotheses can be raised. Perhaps older Brazilian adults with depression probably felt closer to their network members during the stay-at-home recommendations. Some family members could be busy with their work and school commitments before the pandemic, becoming more available and with a higher degree of freedom to connect with the older ones during the pandemic 9, despite living apart. Moreover, co-residing with extended family members during the pandemic may play an additional role, making older adults less susceptible to suffering from loneliness 22. Multigenerational households are a typical living arrangement in Brazil. This potential increase in social support may have reduced emotional loneliness. As loneliness is a multidimensional construct that represents a cluster of subjective and objective experiences of social integration and socioemotional states 32, some researchers consider its multiple facets and classify it into emotional, social, and existential loneliness 3,12,34. Whereas emotional loneliness involves the absence of an intimate attachment, social loneliness originates from the absence of being embedded in a broader community or social networks, such as coworkers and neighbors 3,12. On the other hand, existential loneliness is a sense of meaninglessness in life 12. Due to the single nature of our loneliness measure, we cannot distinguish which loneliness dimensions were affected. It is also possible that social integration and resources from before the pandemic have remained sustainable and provided extra protection against the occurrence or increase of loneliness during the pandemic 12 for many older Brazilian adults. Moreover, some details related to living arrangements, instrumental and emotional social support, and coping mechanisms might also partially justify our findings. Further investigations should explore the potential mediating and moderating effects of these variables in the relation between depression and loneliness. Virtual connectedness has been widely recommended as an alternative to face-to-face contact to offsetting loneliness during the COVID-19 pandemic. Nevertheless, our results showed no significant effect of virtual connectedness on the loneliness trajectory. Findings from other countries on the potential benefits of virtual contact in tackling loneliness among older adults during the pandemic are still controversial 14,16. These results suggest that virtual contact may not be a qualitatively equivalent alternative to face-to-face contact or that older adults who felt lonely were particularly more likely to initiate virtual contact 14. Additionally, data from two research reports have shown that, in 2020, only 50% of Brazilians aged 60 and older reported internet use 35 and 62% of individuals in this age group had never used social media, such as WhatsApp or Facebook 36. Therefore, digital exclusion in Brazil might also limit the potential benefits of virtual contact in decreasing loneliness levels. Although loneliness levels have decreased, this pattern is not present for all older adults. Many of those with relatively high loneliness levels before the pandemic could be among the loneliest respondents during the pandemic. For instance, a study carried out in the United Kingdom exploring the loneliness trajectories in 38,217 adults from March to May 2020 found relatively stable loneliness levels, but it has also shown that those with the highest initial status had no signs of improvement 37. As loneliness is a major public health problem, the health and care systems have a significant role in identifying and providing support for older adults at risk of or already experiencing loneliness 1. Primary prevention might design programs to improve the ability people have to connect and promote social prescribing to stimulate loneliness prevention 3. Social prescribing connects people to community resources (groups and services) that offer support for social, emotional, or practical needs, and may include activities such as arts, befriending, or sports, as well as debt, housing, or employment advice 38. Secondary prevention might involve screening, particularly for individuals at risk, such as those recently retired or widowed, and interventions to help prevent loneliness in this at-risk group. In turn, tertiary prevention might refer those who are acutely or chronically lonely to programs for social skills training or therapy, as indicated 3. Close monitoring by health care providers, mainly the primary care teams, knocking on doors in the community, keeping guided conversations, and motivational interviews may play a fundamental role for those who suffer from loneliness 1. Although many psychosocial interventions have been developed to tackle loneliness among older adults, adequate empirical evidence is still lacking. The current scenario is characterized by low-quality trials, small samples, a lack of theoretical frameworks, mixed measures of loneliness, and short follow-up periods to assess the longer-term impact 1,3. Key therapeutic elements of interventions must be identified, as well as their optimal intensity, frequency, and duration 39. Also, the current evidence for what works to reduce loneliness is primarily for individual-level interventions. Little is known about the effect of community-level strategies on loneliness levels, such as campaigns (e.g., to reduce stigma and raise awareness of the importance of strong social connections) or infrastructure improvement (e.g., transport and digital inclusion) 3,38. This study has several strengths. Most notably, its nationally representative sample of older adults enables the generalizability of our findings on the assessment of loneliness trajectories after the early stages of the COVID-19 pandemic. The availability of pre-pandemic data is also a further strength. To ensure the robustness of our results, alternative analyses based on latent class modeling (i.e., group-based trajectory) were conducted 40 (data not shown). However, the one-class trajectory obtained by the mixed-effects model provided the best fit, with the lowest values of both Akaike (AIC) and Bayesian information criteria (BIC). The main limitations to indicate are those inherent to telephone surveys. However, the response rates of the telephone surveys were considered satisfactory and the sociodemographic characteristics of the participants were quite similar to the Brazilian population within the same age group 30. Besides, considering the restrictions imposed by stay-at-home recommendations, a telephone survey was the most suitable interview mode. Secondly, pre-pandemic and pandemic questions used to measure loneliness were slightly different. At baseline, the measure has no reference to time, whereas in the telephone surveys, we considered the loneliness frequency in the last 30 days. However, both measures likely have similar capacities to capture the construct. A recent study assessing how large-scale aging studies assessed loneliness in 31 countries, including the ELSI-Brazil, has found high consistency levels across different measures 41. Thirdly, our sample excluded people living in long-term care facilities where higher loneliness levels and even loneliness increases could be expected. Nevertheless, older adults living in long-term care facilities represent less than 1% of the population aged 60 years and older in Brazil 42. Also, we could not study changes in living arrangements during the pandemic nor relevant variables that have a longitudinal impact on loneliness, such as coping behaviors, due to the short length of our telephone surveys. Finally, our findings contribute to fostering a better understanding of the loneliness trajectories for older adults living in low- and middle-income countries, which may facilitate individual and community strategies to target interventions. Further in-depth studies could provide a better understanding of this phenomenon. Moreover, monitoring of loneliness levels during the ongoing pandemic and in the post-COVID scenario is recommended and may be implemented in the next waves of ELSI-Brazil. Since no specific strategies such as programs or campaigns were developed at the national level for tackling loneliness among older Brazilian adults, our findings may corroborate the hypothesis that loneliness is a potentially modifiable outcome and can be improved without enlarging the individual’s social network. Acknowledgments ELSI-Brazil was supported by the Department of Science and Technology, Secretariat of Science, Technology, and Strategic Inputs, (DECIT/SCTIE/MS; grants: 404965/2012-1 and TED 28/2017); the Healthcare Coordination of Older Adults, Department of Strategic and Programmatic Actions, Secretariat of Health Care, Brazilian Ministry of Health (COSAPI/DAPES/SASMS; grants: 20836, 22566, 23700, 25560, 25552, and 27510); and the Brazilian National Research Council (CNPq; grant: 403473/2020-9). References 1 1. World Health Organization. Social isolation and loneliness among older people: advocacy brief. Geneva: World Health Organization; 2021. World Health Organization Social isolation and loneliness among older people: advocacy brief. Geneva World Health Organization 2021 2 2. Perlman D, Peplau L. Toward a social psychology of loneliness. In: Gilmour R, Duck S, editors. Personal relationships. London: Academic Press; 1981. p. 31-56. Perlman D Peplau L Toward a social psychology of loneliness. Gilmour R Duck S Personal relationships. London Academic Press 1981 31 56 3 3. Prohaska T, Burholt V, Burns A, Golden J, Hawkley L, Lawlor B, et al. Consensus statement: loneliness in older adults, the 21st century social determinant of health? BMJ Open 2020; 10:e034967. Prohaska T Burholt V Burns A Golden J Hawkley L Lawlor B Consensus statement: loneliness in older adults, the 21st century social determinant of health? BMJ Open 2020 10 e034967 4 4. Ong AD, Uchino BN, Wethington E. Loneliness and health in older adults: a mini-review and synthesis. Gerontology 2016; 62:443-9. Ong AD Uchino BN Wethington E Loneliness and health in older adults: a mini-review and synthesis. Gerontology 2016 62 443 449 5 5. Yang K, Victor C. Age and loneliness in 25 European nations. Ageing Soc 2011; 31:1368-88. Yang K Victor C Age and loneliness in 25 European nations. Ageing Soc 2011 31 1368 1388 6 6. Hossain M, Purohit N, Khan N, McKyer ELJ, Ma P, Bhattacharya S, et al. Prevalence and correlates of loneliness in India: a systematic review. Advance 2020; 17 apr. https://advance.sagepub.com/articles/preprint/Loneliness_in_India_A_systematic_review_of_empirical_studies/11533026. Hossain M Purohit N Khan N McKyer ELJ Ma P Bhattacharya S Prevalence and correlates of loneliness in India: a systematic review. Advance 17 04 2020 https://advance.sagepub.com/articles/preprint/Loneliness_in_India_A_systematic_review_of_empirical_studies/11533026 7 7. Gao Q, Prina AM, Prince M, Acosta D, Sosa AL, Guerra M, et al. Loneliness among older adults in Latin America, China, and India: prevalence, correlates and association with mortality. Int J Public Health 2020; 66:604449. Gao Q Prina AM Prince M Acosta D Sosa AL Guerra M Loneliness among older adults in Latin America, China, and India: prevalence, correlates and association with mortality. Int J Public Health 2020 66 604449 604449 8 8. Hansen T, Sevenius Nilsen T, Knapstad M, Skirbekk V, Skogen J, Vedaa O, et al. COVID-fatigued? A longitudinal study of Norwegian older adults’ psychosocial well-being before and during early and later stages of the COVID-19 pandemic. Eur J Ageing 2021; 19:463-73. Hansen T Sevenius Nilsen T Knapstad M Skirbekk V Skogen J Vedaa O COVID-fatigued? A longitudinal study of Norwegian older adults’ psychosocial well-being before and during early and later stages of the COVID-19 pandemic. Eur J Ageing 2021 19 463 473 9 9. Hwang TJ, Rabheru K, Peisah C, Reichman W, Ikeda M. Loneliness and social isolation during the COVID-19 pandemic. Int Psychoger 2020; 32:1217-20. Hwang TJ Rabheru K Peisah C Reichman W Ikeda M Loneliness and social isolation during the COVID-19 pandemic. Int Psychoger 2020 32 1217 1220 10 10. Luchetti M, Lee J, Aschwanden D, Sesker A, Strickhouser JE, Terracciano A, et al. The trajectory of loneliness in response to COVID-19. Am Psychol 2020; 75:897-908. Luchetti M Lee J Aschwanden D Sesker A Strickhouser JE Terracciano A The trajectory of loneliness in response to COVID-19. Am Psychol 2020 75 897 908 11 11. Stolz E, Mayerl H, Frieidl W. The impact of COVID-19 restriction measures on loneliness among older adults in Austria. Eur J Public Health 2021; 31:44-9. Stolz E Mayerl H Frieidl W The impact of COVID-19 restriction measures on loneliness among older adults in Austria. Eur J Public Health 2021 31 44 49 12 12. van Tilburg T. Emotional, social, and existential loneliness before and during the COVID-19 pandemic. J Gerontol B Psychol Sci Soc Sci 2022; 77:e179-84. van Tilburg T Emotional, social, and existential loneliness before and during the COVID-19 pandemic. J Gerontol B Psychol Sci Soc Sci 2022 77 e179 e184 13 13. Macdonald B, Hülür G. Well-being and loneliness in Swiss older adults during the COVID-19 pandemic: the role of social relationships. Gerontologist 2021; 61:240-50. Macdonald B Hülür G Well-being and loneliness in Swiss older adults during the COVID-19 pandemic: the role of social relationships. Gerontologist 2021 61 240 250 14 14. Hu Y, Qian Y. COVID-19, inter-household contact and mental well-being among older adults in the US and the UK. Front Sociol 2021; 6:714626. Hu Y Qian Y COVID-19, inter-household contact and mental well-being among older adults in the US and the UK. Front Sociol 2021 6 714626 714626 15 15. Krendl AC, Perry BL. The impact of sheltering in place during the COVID-19 pandemic on older adults’ social and mental well-being. J Gerontol B Psychol Sci Soc Sci 2021; 76:E53-8. Krendl AC Perry BL The impact of sheltering in place during the COVID-19 pandemic on older adults’ social and mental well-being. J Gerontol B Psychol Sci Soc Sci 2021 76 E53 E58 16 16. Peng S, Roth A. Social isolation and loneliness before and during the COVID-19 pandemic: a longitudinal study of US Adults over 50. J Gerontol B Psychol Sci Soc Sci 2022; 77:e185-90. Peng S Roth A Social isolation and loneliness before and during the COVID-19 pandemic: a longitudinal study of US Adults over 50. J Gerontol B Psychol Sci Soc Sci 2022 77 e185 e190 17 17. Hansen T, Nilsen TS, Yu B, Knapstad M, Skogen JC, Vedaa O, et al. Locked and lonely? A longitudinal assessment of loneliness before and during the COVID-19 pandemic in Norway. Scand J Public Health 2021; 49:766-73. Hansen T Nilsen TS Yu B Knapstad M Skogen JC Vedaa O Locked and lonely? A longitudinal assessment of loneliness before and during the COVID-19 pandemic in Norway. Scand J Public Health 2021 49 766 773 18 18. Seifert A, Cotten SR, Xie B. A double burden of exclusion? Digital and social exclusion of older adults in times of COVID-19. J Gerontol B Psychol Sci Soc Sci 2021; 76:e99-103. Seifert A Cotten SR Xie B A double burden of exclusion? Digital and social exclusion of older adults in times of COVID-19. J Gerontol B Psychol Sci Soc Sci 2021 76 e99 103 19 19. Holtzman S, DeClerck D, Turcotte K, Lisi D, Woodworth M. Emotional support during times of stress: can text messaging compete with in-person interactions? Comput Human Behav 2017; 71:130-9. Holtzman S DeClerck D Turcotte K Lisi D Woodworth M Emotional support during times of stress: can text messaging compete with in-person interactions? Comput Human Behav 2017 71 130 139 20 20. Reinecke L, Aufenanger S, Beutel ME, Dreier M, Quiring O, Stark B, et al. Digital stress over the life span: the effects of communication load and internet multitasking on perceived stress and psychological health impairments in a German probability sample. Media Psychol 2017; 20:90-115. Reinecke L Aufenanger S Beutel ME Dreier M Quiring O Stark B Digital stress over the life span: the effects of communication load and internet multitasking on perceived stress and psychological health impairments in a German probability sample. Media Psychol 2017 20 90 115 21 21. Herrera MS, Elgueta R, Fernández MB, Giacoman C, Leal D, Marshall P, et al. A longitudinal study monitoring the quality of life in a national cohort of older adults in Chile before and during the COVID-19 outbreak. BMC Geriatr 2021; 21:143. Herrera MS Elgueta R Fernández MB Giacoman C Leal D Marshall P A longitudinal study monitoring the quality of life in a national cohort of older adults in Chile before and during the COVID-19 outbreak. BMC Geriatr 2021 21 143 143 22 22. Torres JL, Braga LS, Moreira BS, Castro CMS, Vaz CT, Andrade ACS, et al. Loneliness and social disconnectedness in the time of pandemic period among Brazilians: evidence from the ELSI COVID-19 initiative. Aging Ment Health 2022; 26:898-904. Torres JL Braga LS Moreira BS Castro CMS Vaz CT Andrade ACS Loneliness and social disconnectedness in the time of pandemic period among Brazilians: evidence from the ELSI COVID-19 initiative. Aging Ment Health 2022 26 898 904 23 23. Castro M, Gurzenda S, Turra C, Kim S, Andrasfay T, Goldman N. Reduction in life expectancy in Brazil after COVID-19. Nat Med 2021; 27:1629-35. Castro M Gurzenda S Turra C Kim S Andrasfay T Goldman N Reduction in life expectancy in Brazil after COVID-19. Nat Med 2021 27 1629 1635 24 24. Aquino EML, Silveira IH, Pescarini JM, Aquino R, Souza-Filho JA, Rocha AS, et al. Social distancing measures to control the COVID-19 pandemic: potential impacts and challenges in Brazil. Ciênc Saúde Colet 2020; 25:2423-46. Aquino EML Silveira IH Pescarini JM Aquino R Souza-Filho JA Rocha AS Social distancing measures to control the COVID-19 pandemic: potential impacts and challenges in Brazil. Ciênc Saúde Colet 2020 25 2423 2446 25 25. Mathieu E, Ritchie H, Rodés-Guirao L, Appel C, Gavrilov D, Giattino C, et al. Coronavirus pandemic (COVID-19). https://ourworldindata.org/coronavirus (accessed on 14/Jan/2022). Mathieu E Ritchie H Rodés-Guirao L Appel C Gavrilov D Giattino C Coronavirus pandemic (COVID-19). https://ourworldindata.org/coronavirus 14/Jan/2022 26 26. Holt-Lunstad J, Smith TB, Baker M, Harris T, Stephenson D. Loneliness and social isolation as risk factors for mortality: a meta-analytic review. Perspect Psychol Sci 2015; 10:227-37. Holt-Lunstad J Smith TB Baker M Harris T Stephenson D Loneliness and social isolation as risk factors for mortality: a meta-analytic review. Perspect Psychol Sci 2015 10 227 237 27 27. National Academies of Sciencies, Engineering and Medicine. Social isolation and loneliness in older adults: opportunities for the health care system. Washington DC: The National Academies Press; 2020. National Academies of Sciencies, Engineering and Medicine Social isolation and loneliness in older adults: opportunities for the health care system. Washington DC The National Academies Press 2020 28 28. Lima-Costa MF, De Andrade FB, Souza Jr PRB, Neri AL, Duarte YAO, Castro-Costa E, et al. The Brazilian Longitudinal Study of Aging (ELSI-Brazil): objectives and design. Am J Epidemiol 2018; 187:1345-53. Lima-Costa MF De Andrade FB Souza PRB Jr Neri AL Duarte YAO Castro-Costa E The Brazilian Longitudinal Study of Aging (ELSI-Brazil): objectives and design. Am J Epidemiol 2018 187 1345 1353 29 29. Lima-Costa MF, Mambrini JVM, Andrade FB, Souza PRB, Vasconcellos MTL, Neri AL, et al. Cohort profile: the Brazilian Longitudinal Study of Aging (ELSI-Brazil). Int J Epidemiol 2022; [Online ahead of print]. Lima-Costa MF Mambrini JVM Andrade FB Souza PRB Vasconcellos MTL Neri AL Cohort profile: the Brazilian Longitudinal Study of Aging (ELSI-Brazil). Int J Epidemiol 2022 Online ahead of print 30 30. Lima-Costa MF, Macinko J, Andrade FB, Souza Júnior PRB, Vasconcellos MTL, Oliveira CM. ELSI-COVID-19 initiative: methodology of the telephone survey on coronavirus in the Brazilian Longitudinal Study of Aging. Cad Saúde Pública 2020; 36 Suppl 3:e00183120. Lima-Costa MF Macinko J Andrade FB Souza PRB Júnior Vasconcellos MTL Oliveira CM ELSI-COVID-19 initiative: methodology of the telephone survey on coronavirus in the Brazilian Longitudinal Study of Aging. Cad Saúde Pública 2020 36 Suppl 3 e00183120 31 31. Gerst-Emerson K, Jayawardhana J. Loneliness as a public health issue: the impact of loneliness on health care utilization among older adults. Am J Public Health 2015; 105:1013-9. Gerst-Emerson K Jayawardhana J Loneliness as a public health issue: the impact of loneliness on health care utilization among older adults. Am J Public Health 2015 105 1013 1019 32 32. Bartrés-Faz D, Macià D, Cattaneo G, Borràs R, Tarrero C, Solana J, et al. The paradoxical effect of COVID-19 outbreak on loneliness. BJPsych Open 2021; 7:e30. Bartrés-Faz D Macià D Cattaneo G Borràs R Tarrero C Solana J The paradoxical effect of COVID-19 outbreak on loneliness. BJPsych Open 2021 7 e30 33 33. Dahlberg L, McKee KJ, Frank A, Naseer M. A systematic review of longitudinal risk factors for loneliness in older adults. Aging Ment Health 2022; 26:225-49. Dahlberg L McKee KJ Frank A Naseer M A systematic review of longitudinal risk factors for loneliness in older adults. Aging Ment Health 2022 26 225 249 34 34. Russell D, Cutrona C, Rose J, Yurko K. Social and emotional loneliness: an examination of Weiss’s typology of loneliness. J Pers Soc Psychol 1984; 46:1313-21. Russell D Cutrona C Rose J Yurko K Social and emotional loneliness: an examination of Weiss’s typology of loneliness. J Pers Soc Psychol 1984 46 1313 1321 35 35. Brazilian Internet Steering Committee. Survey on the use of information and communication technologies in Brazilian households. ICT Households 2020 - COVID-19 edition: adapted methodology. São Paulo: Comitê Gestor da Internet no Brasil; 2021. Brazilian Internet Steering Committee Survey on the use of information and communication technologies in Brazilian households. ICT Households 2020 - COVID-19 edition: adapted methodology. São Paulo Comitê Gestor da Internet no Brasil 2021 36 36. SESC São Paulo; Fundação Perseu Abramo. Idosos no Brasil II: vivências, desafios e expectativas na 3ª idade. https://fpabramo.org.br/publicacoes/publicacao/idosos-no-brasil-vivencias-desafios-e-expectativas-na-terceira-idade/ (accessed on 21/Sep/2022). SESC São Paulo; Fundação Perseu Abramo Idosos no Brasil II: vivências, desafios e expectativas na 3ª idade. https://fpabramo.org.br/publicacoes/publicacao/idosos-no-brasil-vivencias-desafios-e-expectativas-na-terceira-idade/ 21/Sep/2022 37 37. Bu F, Steptoe A, Fancourt D. Loneliness during a strict lockdown: trajectories and predictors during the COVID-19 pandemic in 38,217 United Kingdom adults. Soc Sci Med 2020; 265:113521. Bu F Steptoe A Fancourt D Loneliness during a strict lockdown: trajectories and predictors during the COVID-19 pandemic in 38,217 United Kingdom adults. Soc Sci Med 2020 265 113521 113521 38 38. HM Government. A connected society: a strategy for tackling loneliness - laying the foundations for change. https://www.gov.uk/government/publications/a-connected-society-a-strategy-for-tackling-loneliness (accessed on 06/Dec/2021). HM Government A connected society: a strategy for tackling loneliness - laying the foundations for change. https://www.gov.uk/government/publications/a-connected-society-a-strategy-for-tackling-loneliness 06/Dec/2021 39 39. Fried L, Prohaska T, Burholt V, Burns A, Golden J, Hwakley L, et al. A unified approach to loneliness. Lancet 2020; 395:114. Fried L Prohaska T Burholt V Burns A Golden J Hwakley L A unified approach to loneliness. Lancet 2020 395 114 114 40 40. Nguena Nguefack HL, Pagé MG, Katz J, Choinière M, Vanasse A, Dorais M, et al. Trajectory modelling techniques useful to epidemiological research: a comparative narrative review of approaches. Clin Epidemiol 2020; 12:1205-22. Nguena Nguefack HL Pagé MG Katz J Choinière M Vanasse A Dorais M Trajectory modelling techniques useful to epidemiological research: a comparative narrative review of approaches. Clin Epidemiol 2020 12 1205 1222 41 41. Newmyer L, Verdery AM, Margolis R, Pessin L. Measuring older adult loneliness across countries. J Gerontol B Psychol Sci Soc Sci 2021; 76:1408-14. Newmyer L Verdery AM Margolis R Pessin L Measuring older adult loneliness across countries. J Gerontol B Psychol Sci Soc Sci 2021 76 1408 1414 42 42. Instituto de Pesquisa Econômica Aplicada. Infraestrutura social e urbana no Brasil: subsídios para uma agenda de pesquisa e formulação de políticas públicas (Comunicados do Ipea, nº 93). Condições de funcionamento e infraestrutura das instituições de longa permanência para idosos no Brasil. http://repositorio.ipea.gov.br/bitstream/11058/5208/1/Comunicados_n93_Condi%c3%a7%c3%b5es.pdf (accessed on 06/Dec/2021). Instituto de Pesquisa Econômica Aplicada Infraestrutura social e urbana no Brasil: subsídios para uma agenda de pesquisa e formulação de políticas públicas (Comunicados do Ipea, nº 93). Condições de funcionamento e infraestrutura das instituições de longa permanência para idosos no Brasil. http://repositorio.ipea.gov.br/bitstream/11058/5208/1/Comunicados_n93_Condi%c3%a7%c3%b5es.pdf 06/Dec/2021