The novel coronavirus (COVID-19) pandemic poses global threats to the wellbeing of families and individuals. Increased family-related mental burdens and personal unhappiness have been reported amidst social disruptions and financial insecurity (1–3). Health, social, and economic challenges and adverse consequences from the pandemic are increasing. Posited by Prime and Wade's framework, individuals and families would be differentially influenced by the pandemic and well-functioning families, consisting of effective communication, organization, and belief systems, would be more resilient to these risks (4).

Family communication is a core process of the concepts of family wellbeing and family system and is foundational in maintaining family relationships and fostering individual wellbeing, as stressed in many Western theories and in Chinese culture (4–7). The pandemic triggers various stressors and intensifies the needs of feeling safe, hopeful, and socially connected (8). It has dramatically changed the habitual way many families interact and communicate, especially among those who were separated across different households amidst lockdowns and other social distancing restrictions (9, 10). With such challenges disrupting usual face-to-face communication, communication via digital tools such as instant messaging (IM) has increased substantially (11). Further understanding of how and what families are communicating amidst the pandemic is needed in preparation of the “new normal” and future pandemics.

Previous studies mostly focused on IM communication in social networks in romantic relationships, friendship, and working relationships, but rarely within the whole family (12). IM users did not have emotionally closer feelings with network members when offline (13), and longer duration of IM interactions did not predict better subjective happiness (14). We searched PubMed and Web of Science using a combination of keywords including “COVID-19”, “coronavirus”, “family”, “communication”, “face-to-face”, “instant messaging” and “wellbeing” up to 21 August 2021 to identify IM use in closed communication circles such as families. Only four of our previous papers separately reported the use of family IM chat groups associated with higher family functioning and wellbeing (15) and with personal happiness amidst the pandemic (16), sharing of family life information associated with higher family wellbeing (17), and socioeconomic disparities in using different digital communication methods for family communication (18). We found no reports on the contents of family communication or their associations with family and personal wellbeing amidst the pandemic.

Throughout the life course, family communication is crucial for a balanced family system and involves the exchange and sharing of information, knowledge, values, and beliefs (6, 19). The content of family communication and type of information being shared can shape behaviors, emotions, and sense of self (20, 21), and may thus influence family and individual wellbeing. In addition to providing support against social isolation and loneliness during the pandemic, family communication can also be an important source of health information (22, 23). Hong Kong has had no lockdowns during the pandemic, but almost 100% voluntary mask-wearing in the public started within a few weeks of the first COVID-19 outbreak. The government implemented social distancing regulations since 29 March 2020 (24), which deterred family gatherings and face-to-face interaction with family members. With such disruptions to in-person communication, communication via IM platforms such as WhatsApp and WeChat became popular alternatives, allowing for convenient and instant exchange of texts, voice messages, images including pictures and photos, videos, and audio clips. The pandemic has led to more older adults making use of the internet and smartphones to stay connected to family members (25, 26). In Hong Kong, both face-to-face communication and IM were previously found to be common methods of family communication (18).

We hypothesized that, (i) the use of face-to-face and IM family communication and their corresponding contents amidst the COVID-19 pandemic are associated with family and personal wellbeing, (ii) these associations are mediated by family communication quality, and (iii) communication methods moderate the associations between communication contents and outcomes. This study aimed to examine the use and contents of face-to-face and IM family communication amidst the COVID-19 pandemic, their associations with family wellbeing and personal happiness, the mediating effects of communication quality, and whether the moderating effects of communication methods exist.

Of the 4,891 respondents included in this study, after weighting, 52.9% of them were female, with the mean age of 43.5 years (37.7% aged 45–64 years and 21.3% aged ≥65 years). Details of their sociodemographic characteristics have been previously reported (3, 29). Table 1 shows that after weighting, over half of respondents (53.4%) communicated with family members using both methods (face-to-face and IM messages), followed by IM only (26.7%), face-to-face only (12.7%), and no family communication (7.1%). Compared with those using both methods, more respondents having no family communication or having face-to-face communication only were male (52.5 and 65.1%, respectively, vs. 46.2%), at younger ages, and had lower socioeconomic status (low: 61.6 and 52.8%, respectively, vs. 48.7%) and fewer cohabitants (all P ≤ 0.03); while more of those using IM only were female (61.2 vs. 53.8%), at older ages, and had lower socioeconomic status (low: 56.6 vs. 48.7%) and fewer cohabitants (all P ≤ 0.001). The unweighted characteristics are shown in Supplementary Table 1.

Table 2 shows that after weighting, in face-to-face communication, self/family-related things in daily life (79.5%) were the most frequent contents, followed by information of COVID-19 (78.2%), self/family-related happy/funny things (47.1%), others (e.g., daily life information, news, and current affairs, etc.,) (46.0%), showing care (42.8%), other health information (30.8%), self/family-related unhappy things (29.3%), self/family-unrelated happy/funny things (26.9%), encouragements (14.9%), appreciations (12.1%), and good wishes (11.9%). In IM messages, information of COVID-19 (80.6%) was the most common contents, followed by self/family-related things in daily life (59.6%), showing care (41.2%), and others (40.2%), whereas encouragements (16.9%), good wishes (16.3%), self/family-related unhappy things (16.0%), and appreciations (8.7%) were the least 4 common ones. When contents were grouped into 4 kinds, fewer respondents reported neutral (88.1 vs. 98.4%), positive (43.9 vs. 60.8%), and negative contents (16.0 vs. 29.3%) in IM communication than in face-to-face communication (all P <0.001). The unweighted percentages of contents are shown in Supplementary Table 2.

Compared with face-to-face communication only, IM only included less self/family-related things in daily life (51.7 vs. 73.3%) and self/family-related unhappy things (10.9 vs. 19.4%), but more information of COVID-19 (81.7 vs. 74.4%), showing care (47.0 vs. 36.5%), other health information (38.0 vs. 29.3%), encouragements (18.3 vs. 7.2%), and good wishes (20.1 vs. 5.9%) (all P < 0.001). For the 4 kinds of contents, it included less neutral (83.1 vs. 97.3%) and negative contents (10.9 vs. 19.4%) but more supportive contents (53.7 vs. 37.5%) (all P < 0.001). Using both methods included higher percentages of almost all contents than using one method only (all P ≤ 0.02) except good wishes compared with IM only (18.9 vs. 20.1%, P = 0.36). In general, using both methods contained more neutral, positive, and negative contents than using one method only (P < 0.001), except supportive contents (both 54.8% vs. IM only 53.7%) (P = 0.54).

Table 3 shows that after excluding those with no communication, when communication methods and contents were included in the same models, compared with using face-to-face communication only, using IM only and using both methods were associated with higher levels of family communication quality, family wellbeing, and personal happiness (Model III, adjusted βs: 0.30–0.48, all P < 0.001). Only positive, supportive and negative contents were associated with higher levels of family communication quality (adjusted βs: 0.27–0.70, all P < 0.001), and only positive and supportive contents were associated with higher levels of family wellbeing and personal happiness (adjusted βs: 0.45–0.74, all P < 0.001).

Family communication quality partially mediated the associations of communication methods and positive contents with family wellbeing (proportion mediated: 50.1–76.1%) and personal happiness (proportion mediated: 35.2–51.4%), and the associations of supportive contents with personal happiness (73.9%), and almost fully mediated the association of supportive contents with family wellbeing (93.5%, P for direct effect = 0.35) (Table 4).

Communication methods moderated the associations of positive contents with family wellbeing (P for interaction=0.006) (Figure 1). Positive contents had stronger associations with better family wellbeing in using both methods (estimated score changes: 0.71, 95% CI: 0.57–0.85) and face-to-face only (0.78, 95% CI: 0.53–1.03) than in IM only (0.37, 95% CI: 0.18–0.56) (P = 0.004 and 0.009, respectively).

This is the first study to report family communication methods and contents were independently associated with family communication quality, family wellbeing, and personal happiness, showing that better family wellbeing and personal happiness were associated with using IM only and both methods than face-to-face communication only, and were associated with having positive and supportive contents in family communication. About half to almost all these associations were mediated by communication quality. Communication methods moderated the association between positive contents and family wellbeing, showing stronger associations of family wellbeing with both methods and face-to-face only than IM only. These results are consistent with our three hypotheses.

Our results highlight the importance of using IM messages to communicate with family members amidst the COVID-19 pandemic, which showed better outcomes than face-to-face communication only. Due to physical isolation and social distancing, traditional face-to-face family communication and usual family gatherings have been disrupted, while digital communication via IM tools is increasingly used and often the only option in maintaining familial interactions in separated households (43). Those who could only use IM communication may value the interaction more and thus feel connected and supported. More men and younger people used face-to-face communication only probably because some might choose not to or seldom use IM messages for family communication even under social distancing restrictions. Consistent with previous studies, women and older people cared more about family affairs and participated in more IM family communication (17, 44). Their behaviors may be encouraged by perceived usefulness and enjoyment, attachment motivation, and relationship commitment, which were predictors of the intention of continuously using IM to sustain interpersonal relationships (45).

We found that when using both IM and face-to-face communication within the family context, almost all contents increased to a greater or lesser degree (15.9–87.2%) than using one method only (face-to-face only: 5.9–74.4%, IM only: 9.4–81.7%), except good wishes which were similar to IM only. Using both methods means more communication. IM messages may act as a supplement of face-to-face communication as an additional way of conversing even when face-to-face is possible, as IM is more convenient. For example, more COVID-19 and health-related information can be promptly shared in their original formats via IM messages, such as long texts, web links, photos, and short videos. Also, as Chinese people tend to indirectly express their encouragements, appreciations, and good wishes to family members (37–39), IM has become a brand-new platform to deliver and convey supportive contents beyond merely text, with emojis and many readily available e-messages for more vivid and intimate expressions (46). Face-to-face communication contains more self/family-related things in daily life and unhappy things and can provide better communication satisfaction through verbal, facial, and body language but is restricted by physical location (47).

All kinds of contents were associated with better family communication quality when sociodemographic characteristics were controlled (Model II), but only positive and supportive contents remained associated with family wellbeing and personal happiness when communication methods were also controlled (Model III). In a well-functioning family, members should be willing to share all kinds of contents, including negative ones. An effective coping mechanism in combating stress of negative life events is to seek comfort and help within social support networks such as family (48), through the sharing of affection and empathy, and giving encouragement, advice, and practical help such as health information (49). COVID-19 and health-related information have been widely spread amidst the pandemic, and the related sharing and forwarding behaviors could mean both showing care to family members and a cause of the ongoing infodemic, pandemic fear, and mental health burdens (50, 51). Previous studies have linked positive and supportive contents with confidence and competence among family members, while negative contents such as criticism were associated with lower self-esteem and defiance (20, 52). While open and direct expression of affection, both verbal and non-verbal, are encouraged in Western families (53), such as saying “I love you”, this is not common in Chinese households, where strong emotions are typically held back, stemming from a historical emphasis on the regulation of social behaviors and expression of emotions (54, 55). Such differences can also be observed in the discussion of funny and humorous topics within families (56). Considering the relatively low weighted percentages of positive and supportive contents (positive: 42.1–62.2%; supportive: 37.5–54.8% vs. neutral: 83.1–99.3%), increasing these contents through IM first may lead to increased use in face-to-face communication, which may promote family wellbeing and personal happiness. Intervention studies on IM use to deliver such contents to promote family and individual wellbeing are warranted.

The moderating effect of communication methods showed that positive contents in face-to-face communication only and both methods were more strongly associated with better family wellbeing than in IM messages, suggesting that sharing self/family-related and unrelated happy/funny things by IM only may be less effective for maintaining and nourishing family relationships. In face-to-face communication, non-verbal language, such as laughter and smiles, can give real-time positive feedbacks and immediately create a happy and enjoyable atmosphere (57). According to the attachment theory, pleasant and frequent interactions with others contribute to individual mental and emotional wellbeing (45), which may evoke better family wellbeing in family communication.

The mediation effects of family communication quality can provide new evidence to Prime and Wade's framework (4). Amidst the COVID-19 pandemic, family communication provides clear information, emotional sharing, collaborative problem-solving, and dyadic and family coping to connect family members and share beliefs. Quality communication, such as using both face-to-face and IM communication and including positive or supportive contents, can thus provide security and hope for vulnerable members during periods of stress (4), shown as higher perceived family wellbeing and personal happiness in the present study.

In Hong Kong, the most westernized city in China, the high penetration rate of smartphones (91.5% in 2019) and the Internet (87.0% in 2019) means most people can conveniently use social media and IM messages (58, 59). With 93.6% of the population being Chinese, Hong Kong people highly value family relationships, which are influenced by collectivism and Confucius ideals in traditional Chinese culture (31). However, we found that 7.1% of people had no family communication and they reported the lowest family communication quality, family wellbeing, and personal happiness. They tended to be in low socioeconomic status and could be vulnerable and more adversely impacted by the pandemic than others. This is an example of digital inequality, shown as the inequality in terms of access, usage, skills, and self-perceptions to digital engagement in individual and macro-level domains (60). Urgent attention and assistance should be given to these vulnerable people from policymakers and social welfare organizations.

Our study had some limitations. First, recall bias and social desirability bias could not be avoided in self-administered questionnaires. However, the use of communication methods and contents in family communication when the pandemic was severe was asked during the easing period of the pandemic, and recall errors would be little within such recent time periods. We used an online survey via emails without interviewers, which could help reduce social desirability bias (61). Second, although we tried to provide a clearer temporal sequence by asking the perceived outcomes during the easing period and the communication methods and contents during an earlier period when the pandemic was severe, due to the cross-sectional observational study design, we could not rule out reverse causality. Future prospective studies are needed to verify the associations and mediation effects we observed. Also, systematic bias due to residual confounding might exist. For example, people having face-to-face communication only could be lack of health literacy to share digital information with family, while sharing family life information through ICT tools were found to be associated with family wellbeing (17). Also, those having IM only might live separately with their family so face-to-face communication was unavailable. Such separation from family, especially amidst the pandemic, could lead to low family wellbeing or personal happiness. Third, as the COVID-19 pandemic changes rapidly and unpredictably, we tried to collect the largest possible sample within a short period, non-response bias could be present as younger and better-educated respondents were included. Generalization could be limited. Finally, details of the contents were not asked, and more in-depth information should be collected in future studies.

We have first shown that, amidst the COVID-19 pandemic, better family wellbeing and personal happiness were associated with family communication using IM only and both methods than face-to-face only, and with positive and supportive contents. These associations were partially or almost fully mediated by communication quality. Family IM communication and positive and supportive contents may promote family wellbeing and personal happiness. People with no family communication may need urgent attention and assistance. Prospective studies are needed to verify the associations and mediations.

The dataset presented in this article is not readily available because the sharing of data to third parties was not mentioned in subjects' consent. Requests to access the dataset can be directed to the corresponding author.

The studies involving human participants were reviewed and approved by Institutional Review Board of the University of Hong Kong/Hospital Authority Hong Kong West Cluster (Reference Number: UW 20-238). Informed consent was obtained from all participants included in this study.

WG and SS: formal analysis and writing—original draft. BW: data curation, project administration, and writing—review and editing. SW: methodology and writing—review and editing. AL: conceptualization and writing—review and editing. SH: conceptualization, methodology, and writing—review and editing. MW and TL: supervision, conceptualization, and writing—review and editing. All authors participated in the critical review of this study and provided final approval for publication submission.

This study was funded by the Hong Kong Jockey Club Charities Trust.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.