We conducted a closed, web-based, and cross-sectional survey with members of a national retired educators’ organization (RTOERO) between May 6 and 19, 2020. More than three-quarters of the members (76.5%, 62,000/81,000) had registered an email address with the organization and were eligible to participate. Members were invited by email from RTOERO’s chief executive officer and sent reminders at 7 and 10 days. Study materials were provided in English and French. The questionnaire was co-designed and pretested with RTOERO leadership and included 32 questions that examined the impact of COVID-19 on daily life, loneliness, and the use of digital technologies for social connectivity [10].

Social media and video communication use was measured by asking “Do you use any social networking websites (eg, Facebook) or apps (eg, Zoom or FaceTime) to communicate with friends and family?” (yes/no). Respondents were also asked about their internet connection, smartphone access, loneliness, and sociodemographic characteristics including age, gender, ethnicity, language, living arrangement, self-perceived health status and location of residence (rural/urban). The questionnaire was pretested in English with 18 RTOERO board members and staff and French with 1 staff member for usability, technical functionality, clarity, flow, sensitive questions, and timing. The pretest results were not included in the final analysis. In the pretest, it took respondents on average 13 minutes to complete the survey (median: 14 minutes).

Surveys were only analyzed if the respondent clicked “Submit” and responded to more than 1 question. Nonusers of social media were compared to users through Pearson chi-square tests in univariable analyses on sociodemographic factors (age, gender, rurality, health status, and ethnicity), access factors (internet connection and smartphone access), and relational factors (living arrangement, loneliness, communication frequency, and social support). Corresponding adjusted associations were made using multivariable logistic regression including all covariates in the model. Survey questions on gender and ethnicity included “Prefer to self-identify” or “Prefer not to say” response options, which were collapsed into an “Other” category and retained in the regression analysis. Otherwise, respondents with missing (ie, don’t know or blank) covariate values were excluded from the model.

We conducted an inductive thematic analysis of free-text responses to 4 survey questions that included an open-ended response option; these questions asked respondents to describe feelings of loneliness, strategies they use to avoid feeling lonely, how RTOERO could support members, and if they had any other comments or suggestions. Responses were reviewed for descriptions of experiences with virtual connection and a set of 14 initial codes were generated by the analyst (SDN) and discussed with the study team. Themes were identified by examining patterns across the codes and were presented, along with illustrative quotations, to both the study team and RTOERO members for input and reflection.

Participation was voluntary, and informed consent was obtained electronically. The study was approved by the Research Ethics Board at Women’s College Hospital (#2020-0051-E) and reporting followed the Checklist for Reporting Results of Internet E-Surveys [11].

In a survey of 4879 older retirees of the education sector, we found that 1 in 7 were not using social networking websites or apps to communicate with friends and family early in the COVID-19 pandemic. We identified several characteristics associated with an increased likelihood of nonuse including male gender, advanced age, living alone, poorer health, poor internet connection, a lack of smartphone access, and lower social support. Our findings are consistent with studies of technology use in older adults, which similarly show that advanced age, male gender, being unmarried, and poorer self-reported health are important predictors of nonuse and telemedicine unreadiness [3,12].

We found that sentiments toward technology use during the early weeks of the pandemic were mostly positive, with respondents describing how technology facilitated social connections, allowed them to maintain routines, or helped them adjust to a new way of life. Qualitative studies on the experiences of older adults during the pandemic’s first wave report similar findings, with technology being described as having a facilitative role in allowing older adults to engage in the things that matter most to them despite ongoing restrictions [13-15].

Loneliness and social isolation are top concerns of older adults during the COVID-19 pandemic [10,16], so developing effective strategies to maintain social connection while adhering to public health recommendations is essential. For many older adults, digital communication has been a source of joy and comfort during the often uncertain times of the pandemic [16]. Efforts are needed to enable access and build digital literacy to support those who wish to use these technologies but are currently unable to. This includes rigorous research on what strategies are effective and for whom and addressing structural barriers to technology use, including universal internet access [4]. Such improvements could have far-reaching impacts beyond enhanced social connection. Now more than ever, digital technologies are being used to deliver lifesaving health care and health promotion services, including access to COVID-19 testing and vaccines and other essential services including education, groceries, banking, shopping, and government resources and support. Improving the adoption and use of these technologies in older adults and removing structural barriers to access will ensure the benefits of technology are equitably realized. This is especially important as telehealth expands and is incorporated into routine clinical practice [12].

Despite its benefits, technology-based solutions are not a panacea for social connection. In our study and others [17], older adults have acknowledged that even with virtual connection, gaps in social connection and support remain. A recent comparative study using nationally representative longitudinal survey data in the United States and United Kingdom found that virtual contact during the pandemic did not compensate for the lack of in-person contact in terms of supporting older adults’ mental well-being in either country [18]. In fact, the authors found that in the United States, older adults with more frequent virtual contact were more likely to feel lonely during the pandemic and have become lonelier in both countries [18]. Although we did not find an association between loneliness and social media use/nonuse, future research is needed to better understand temporal associations and the potential unintended consequences of virtual connection, particularly for subgroups like older women, who have been reported to be more likely to rely on virtual-only contact during the pandemic [18].

These findings underscore that the value of in-person and offline connections cannot be forgotten. Weekly telephone call interventions with volunteers have shown benefit to both older adults living in long-term care [19] and those in the community [20] during the COVID-19 pandemic. Supporting older adults to remain socially connected through more traditional means is an important and more accessible complement to technological solutions; both are needed to ensure solutions to the issue of loneliness do not further exclude older adults who are not on the internt [4,21,22]. Outdoor environments and spaces are a safe option for connecting in-person at a distance, and the use of these spaces have been shown to be an important predictor of social connectedness pre–COVID-19 [23]. Improving equitable access to safe, accessible open spaces within neighborhoods and enhancing existing ones to better facilitate social interaction would again pay dividends now and in the future.

Our study has several limitations. Our study is based on a convenience sample of community-dwelling, retired educators who had internet access. Due to this sample, the prevalence of social media use was likely overestimated in our study. Estimates of social media use by older adults during the pandemic for social connection are wide-ranging (eg, 37% to 80% based on surveys using similar methodologies conducted during the first wave in the United States and Canada) [24,25]. The rates of use are expected to be lower in the general population of older adults, who may face access barriers due to lower levels of education, lower income, health challenges, and ethnicity or race. In a racially diverse study of US older adults, 76% reported minimal video-based socializing; access to and discomfort with technology were described as a key barrier to coping, maintaining social connection, and accessing health care during the pandemic [26]. There are other factors that are associated with nonuse of digital technologies that we did not evaluate, including cognitive impairment, frailty, vision or hearing impairments, and non-English/French speaking ability [27,28]. The unique needs and barriers of these population subgroups warrant thorough investigation and consideration when designing interventions, programs, and services.

Among a sample of retired educators with internet access, we found that male gender, advanced age, living alone, poorer health, and lower social support increased the odds of nonuse of social media to communicate with friends and family during the first wave of the COVID-19 pandemic. The reliability of internet connection and smartphone access also predicted nonuse. Strategies to improve the uptake of digital technologies for social connection must address structural barriers to accessing technology and build digital literacy among older adults. Complementary approaches, such as telephone outreach, are also needed to improve social connection without further excluding older adults who remain offline.