Closeness of Social Media Contacts and Loneliness Among US Adults: A Nationally Representative Study

Abstract

Objectives:

Loneliness is strongly linked to morbidity and mortality. While social media use has been associated with increased loneliness among adolescents, little research has examined this association among adults. It is also not clear whether it is better to communicate only with close personal friends or with people one has never met in person. We examined whether real-life closeness of social media contacts was associated with loneliness among a nationally representative sample of US adults.

Methods:

Participants included 1512 US adults aged 30 to 70 years who were surveyed in July and August 2023. We assessed loneliness using the National Institutes of Health Patient-Reported Outcomes Measurement Information System’s 4-item scale. Participants self-reported the proportion of their social media contacts whom they had “never met in person” (NMP) and whom they considered “close personal friends” (CPFs). We used logistic regression to examine associations between NMPs/CPFs and loneliness, controlling for 7 sociodemographic covariates and including survey weights to make results nationally representative.

Results:

In fully adjusted weighted multivariable models, participants in the highest quartile for NMPs had more than double the odds of loneliness (adjusted odds ratio = 2.33; 95% CI, 1.52-3.58) than those in the lowest quartile, and the association between NMPs and loneliness was linear (P < .001). However, we found no significant association between CPFs and loneliness (P = .93).

Conclusion:

While having more social media interactions with relative strangers was linked to increased loneliness, having more social media contacts who were CPFs was not linked to reduced loneliness. Future research should examine temporal associations and reasons for these findings. It may be useful for interventions to focus on reducing interactions with NMPs.

Keywords

social media loneliness closeness in-person adult mental health PROMIS Patient-Reported Outcomes Measurement Information System

Loneliness is an emerging global epidemic associated with substantial morbidity, mortality, and health care costs.¹ About half of people in the United States are lonely,^1,2 and people who are lonely experience negative health consequences, including poor mental health,³ cardiovascular disease and stroke,⁴ substance use,⁵ and increased mortality from all causes.^1,6 The 2023 Surgeon General’s report Our Epidemic of Loneliness and Isolation equated morbidity and mortality from loneliness to smoking 15 cigarettes per day.²

While social media use has been identified as potentially related to loneliness,⁷ findings have been mixed.^8
-10 Some studies have suggested that social media use might be associated with positive outcomes such as increased social capital, which may help alleviate loneliness.^11,12 However, other studies suggest associations between increased social media use and increased loneliness.^10,13,14

Systematic reviews emphasize the need for more research among adults because most studies have focused on adolescents and young adults.^13,14 This gap in the literature is important because adults represent 75% of the US population, they are heavily exposed to social media,¹⁵ and many clinical sequelae associated with loneliness are particularly severe during adulthood.¹

Another gap in the literature is that most prior work has focused on social media use as a monolithic exposure rather than acknowledging the myriad contextual factors related to its use. Some research has examined the varying content encountered while using social media, such as whether experiences are positive or negative.^13,16,17 However, an important understudied contextual factor involves with whom one communicates on social media.

Conceptually, one might expect that interacting primarily with strangers might be different from interacting primarily with close real-life contacts.¹⁸ Yet, research is mixed on this issue. For example, some studies highlight that increased measures of overall social media communication, such as the number of social media contacts, are associated with positive mental health outcomes regardless of closeness of contacts, because all contacts can still be good sources of information and/or support.^19
-21 Other research cautions that overall social media interaction may be associated with increased loneliness, regardless of the closeness of contacts, potentially because of magnification of social comparison and/or displacement of beneficial in-person activities.^10,22,23 Other research suggests a nuanced association, such as benefits when contacts are close but drawbacks when contacts are not known in person.^24
-26 Better understanding varying risks related to the makeup of one’s contacts might help inform recommendations aimed at preventing loneliness and its sequelae.

Our objective was to assess associations between closeness of social media contacts and loneliness in a nationally representative sample of US adults. We hypothesized that (1) interaction with an increased proportion of strangers would be associated with increased loneliness and (2) interaction with an increased proportion of close friends with whom users have existing offline relationships would be associated with decreases in loneliness.

Methods

Participants and Procedures

To obtain the sample, Qualtrics Sampling Services (QSS)²⁷ recruited a national sample of US adults in July and August 2023. The only selection criteria were that participants were required to read English, complete an online questionnaire, and be aged 30 to 70 years, inclusive. After obtaining informed consent for all participants, QSS administered this sample an online cross-sectional questionnaire focusing on social media behaviors and mental health outcomes. QSS compensated participants for their time with credits that could be redeemed for cash awards. The study was approved by the Oregon State University Institutional Review Board (protocol HE-2023-374).

QSS first conducted a pilot survey of 50 participants. Because no unforeseen challenges arose, these 50 participants were included in the final sample. Of the 2461 adults initially invited to participate, 1512 (61.4%) were included in the final analyses. This figure reflects that 418 did not complete the survey and 531 were not included because of poor data quality. QSS assesses data quality according to an established QSS algorithm,²⁸ through which data are removed if clear inconsistencies indicate poor effort—for example, a participant claiming to have never smoked cigarettes in 1 item but then claiming to smoke 1 pack per day in a subsequent item. QSS also removed data for straightlining, which occurs when a participant responds quickly with the same response for multiple items in a row when uniform responses were not expected. While excluded respondents were more likely than included respondents to be younger, male, and non-White (P < .001 for all), we applied weighting adjustments to address this nonresponse bias.

Measures

Primary dependent variable: loneliness

We assessed loneliness by using the Patient-Reported Outcome Measurement Information System (PROMIS) short-form version 4a of the Perceived Social Isolation Scale.²⁹ Participants were asked how strongly they agreed, during the past 7 days, with the following 4 statements: “I felt left out,” “I felt that people barely know me,” “I felt isolated from others,” and “I felt that people were around me but not with me.” Participants responded to each item on a 5-point Likert scale, where 0 = never and 4 = always. The 4-item scale had strong internal consistency reliability (α = .92), which is consistent in other studies. Summing scores for the 4 items, each of which was operationalized as 0 to 4, yielded a total score of 0 through 16. For primary analyses, we dichotomized the item to define loneliness using a cutoff score of 10 based on PROMIS guidance.²⁹ However, we also conducted sensitivity analyses varying this cutoff to 9 and 11 to ensure robustness of results.

Independent variables: never met in person and close personal friends

Participants used 2 separate sliders to estimate the percentage of their social media contacts whom they had never met in person (NMP) and those that they deemed close personal friends (CPFs). Each slider indicated a percentage from 0% to 100% in gradations of 10. Responses were not required to sum to 100 because participants often had contacts whom they had met in person but did not consider close. In pretesting, asking participants to subjectively determine whom they considered close had greater face validity than asking other items to assess closeness, such as indicating the frequency of interaction with specific people. We found that subjective determination of closeness was superior to other methods for 2 reasons. First, items such as those asking specific frequencies of interaction with specific individuals were associated with confusion and respondent burden. Second, such items examining frequency of interaction did not accurately evaluate closeness, because people might interact frequently with people not considered close, and we aimed to capture participants’ subjective experiences. We selected gradations of 10 in a 0 to 100 scale for 2 reasons. First, this system has been used successfully in prior work.^16,24 Second, in pretesting, participants found finer gradations to involve too much of a burden but larger blocks to be insufficient. For the purposes of analysis and to improve ease of interpreting the results, we combined responses into quartiles. However, we also conducted analyses with independent variables as continuous to assess the robustness of results.

Sociodemographic factors

We assessed key sociodemographic factors potentially related to loneliness: age (30-39, 40-49, 50-59, 60-70 y), gender identity (man, woman), sexuality (heterosexual, nonheterosexual), race and ethnicity (non-Hispanic White, non-Hispanic Black, other [Hispanic, Asian American, Pacific Islander, American Indian, Alaska Native, multiracial]), education level (no degree, associate’s or bachelor’s degree, or master’s or doctoral degree), employment status (employed, partially employed [part-time or sporadic work], unemployed), and marital status (single, married, other [divorced, separated, or widowed]). In primary multivariable analyses, we included all covariates. In sensitivity analyses, we included a more parsimonious set of sociodemographic covariates that had bivariate associations with loneliness at P < .15.³⁰

Social media use

The primary independent variables of the current study related to the makeup of social media contacts in terms of the perceived proportion of NMPs and CPFs. The independent variables were not related to the volume and/or frequency of social media use. However, it is still relevant to the context of the overall study that we comprehensively assessed social media use in terms of time and frequency. To do so, we first provided a list of the 10 most commonly used platforms at the time of the survey (Facebook, X, Reddit, YouTube, LinkedIn, Instagram, TikTok, Snapchat, Pinterest, and WhatsApp); at that time, these 10 platforms constituted >95% of all social media use.³¹ We clarified in all questions that participants were to report social media use only for personal reasons and not related to work.³² Participants estimated social media use in terms of time and frequency.

Analysis

We first described participant characteristics using percentages and frequencies. We then assessed bivariate associations between sociodemographic factors and the presence or absence of loneliness using the Pearson χ² test.

We then conducted 3 separate logistic regression models, all of which focused on loneliness as the dependent variable. Because the first model included only NMP contact percentage as the independent variable, it served to assess hypothesis 1. Because the second model included only CPF contact percentage as the independent variable, it assessed hypothesis 2. We applied survey weights for both models to address nonresponse bias and make the findings representative of the general US population in terms of age, gender identity, sexuality, race and ethnicity, and education.

We conducted multiple sensitivity analyses to examine the robustness of results: (1) various cutoffs for loneliness according to the PROMIS scale, (2) independent variables as continuous instead of quartiles, (3) including both independent variables in the same model, (4) using only covariates with bivariable associations with anxiety at P < .15, and (5) not including survey weights.

We performed all statistical analyses using Stata version 18 (StataCorp). For all analyses, we defined significance using a 2-tailed α of .05.

Results

One-quarter (25%) of participants were categorized as lonely when we used a cutoff of 10 on the PROMIS scale (Table 1). The distribution of NMP was slightly skewed left (median, 30%; IQR, 10%-60%) while that of CPF was slightly skewed right (median, 60%; IQR, 20%-80%). We combined these variables into quartiles due to their skewness. Combining NMP into quartiles yielded the following 4 categories: 0% (quartile 1), 10% to 20% (quartile 2), 30% to 60% (quartile 3), and 70% to 100% (quartile 4). Combining CPF into quartiles yielded the following 4 categories: 0% to 20% (quartile 1), 30% to 50% (quartile 2), 60% to 80% (quartile 3), and 90% to 100% (quartile 4). Mean (SD) values for NMP and CPF were 36% (32%) and 55% (32%), respectively. Loneliness in bivariate analyses was associated with NMP contacts, being nonheterosexual, younger age, not having a college degree, being unemployed, and being unmarried.

Table 1.

Closeness of social media contacts and sociodemographic characteristics of adults aged 30 to 70 years with and without loneliness, United States, July–August 2023^a

	Loneliness, %^b
Variable	Total	No	Yes	P value^c
Actual sample size, no. (%)	1512 (100)	1120 (74)	392 (26)
Weighted percentage		74	25
Contacts: never met in person, %^d				<.001
0	17	19	11
10-20	27	28	26
30-60	30	30	30
70-100	26	24	33
Contacts: close personal friend, %^d				.32
0-20	25	24	26
30-50	21	21	21
60-80	33	32	35
90-100	22	23	18
Gender identity				.33
Man	50	51	48
Woman	50	49	52
Sexuality				.002
Heterosexual	92	93	87
Nonheterosexual	8	7	13
Age, y				<.001
30-39	30	30	33
40-49	25	23	30
50-59	21	20	25
60-70	24	27	13
Race and ethnicity				.24
Non-Hispanic White	75	75	73
Non-Hispanic Black	14	13	17
Other^e	11	12	10
Education level				.03
No degree	49	47	56
Associate’s or bachelor’s degree	33	35	29
Master’s or doctoral degree	17	18	15
Employment status				.01
Employed	65	66	62
Partially employed^f	28	29	27
Unemployed	7	5	10
Marital status				<.001
Single	25	22	35
Married	55	58	44
Other^g	20	20	21

Data source: online questionnaire conducted by Qualtrics Sampling Services in July–August 2023.

Data are presented as percentage unless noted otherwise. Loneliness was defined per the Patient-Reported Outcomes Measurement Information System–Perceived Social Isolation Scale.²⁹ Participants were asked how strongly they agreed, during the past 7 days, with each of the following 4 statements: “I felt left out,” “I felt that people barely know me,” “I felt isolated from others,” and “I felt that people were around me but not with me.”

Pearson χ² tests were used, and significance was defined as P < .05.

For “never met in person,” participants estimated the percentage of their social media contacts whom they had never met in person. For close personal friends, participants estimated the percentage of their social media contacts whom they considered close personal friends. Each slider indicated a percentage from 0% to 100% in gradations of 10.

Hispanic, Asian American, Pacific Islander, American Indian, Alaska Native, and multiracial.

Part-time or sporadic work.

Divorced, separated, or widowed.

In model 1, we found a significant association between NMP contacts and loneliness (Table 2). For example, when compared with those in the lowest quartile of NMP contacts, those in the highest quartile were 2.33 times (95% CI, 1.52-3.58) more likely to be lonely. In addition, loneliness was associated with being nonheterosexual, younger age, being partially employed or unemployed, and being unmarried. In model 2, we found no association between the percentage of CPF contacts and loneliness, and associations with demographic factors were similar to those from model 1. We identified a significant linear association between NMP contacts and loneliness (P < .001) but not between CPF contacts and loneliness (P = .93). Therefore, our first hypothesis was supported but our second hypothesis was not.

Table 2.

Multivariable associations between closeness of social media contacts and loneliness among a nationally representative sample of adults aged 30 to 70 years, United States, July–August 2023^a

	Adjusted odds ratio (95% CI)^b
Variable	Model 1	Model 2
Contact: never met in person, %^c
None	1 [Reference]^d
10-20	1.57^e (1.03-2.39)
30-60	1.62^e (1.06-2.48)
70-100	2.33^f (1.52-3.58)
Contact: close personal friend, %^c
0-20		1 [Reference]^g
30-50		0.86 (0.59-1.27)
60-80		1.02 (0.72-1.46)
90-100		0.71 (0.47-1.06)
Gender identity^h
Man	1 [Reference]	1 [Reference]
Woman	1.22 (0.92-1.61)	1.17 (0.89-1.54)
Sexuality
Heterosexual	1 [Reference]	1 [Reference]
Nonheterosexual	1.62^e (1.05-2.52)	1.66^e (1.07-2.58)
Age, y
30-39	1 [Reference]	1 [Reference]
40-49	1.08 (0.76-1.54)	1.04 (0.73-1.49)
50-59	0.97 (0.67-1.40)	0.91 (0.62-1.31)
60-70	0.34^f (0.21-0.56)	0.30^f (0.18-0.50)
Race
Non-Hispanic White	1 [Reference]	1 [Reference]
Non-Hispanic Black	0.99 (0.67-1.46)	1.02 (0.70-1.50)
Otherⁱ	0.74 (0.48-1.13)	0.75 (0.49-1.16)
Education level
No degree	1 [Reference]	1 [Reference]
Associate’s or bachelor’s degree	0.79 (0.59-1.05)	0.79 (0.59-1.06)
Master’s or doctoral degree	0.88 (0.57-1.36)	0.94 (0.61-1.45)
Employment status
Employed	1 [Reference]	1 [Reference]
Partially employed^j	1.50^e (1.07-2.10)	1.45^e (1.03-2.03)
Unemployed	1.84^e (1.14-2.97)	1.71^e (1.06-2.74)
Marital status
Single	1 [Reference]	1 [Reference]
Married	0.58^f (0.43-0.80)	0.57^f (0.42-0.78)
Other^k	0.87 (0.59-1.29)	0.86 (0.58-1.26)

Data source: online questionnaire conducted by Qualtrics Sampling Services in July–August 2023.

Odds ratio, adjusted for all variables in the table, for loneliness as defined by the Patient-Reported Outcomes Measurement Information System–Perceived Social Isolation Scale.²⁹ Data were also adjusted by survey weights to make findings nationally representative.

Each slider indicated a percentage from 0% to 100% in gradations of 10.

P < .001 for linearity between social media contacts never met in person and loneliness.

Significant at P < .05.

Significant at P < .001.

P = .93 for linearity between social media contacts with close friends and loneliness.

The 9 people who indicated a gender identity other than man or woman were not included in multivariable analyses to avoid instability of models resulting from small cell sizes.

Hispanic, Asian American, Pacific Islander, American Indian, Alaska Native, and multiracial.

Part-time or sporadic work.

Divorced, separated, or widowed.

Because all sensitivity analyses yielded similar results in significance and magnitude, we included only results from primary analyses here.

Discussion

This nationally representative study found that among US adults aged 30 to 70 years, the percentage of social media contacts whom one has NMP is strongly and linearly associated with loneliness as assessed by a well-accepted and validated measure, supporting our first hypothesis. However, contrary to our initial second hypothesis, we found no significant association between the percentage of social media contacts who are also CPFs and loneliness.

Our finding of a linear association between NMPs and loneliness aligns with several studies conducted among adolescents and young adults.^23,33 This finding is consistent with population-based studies focusing on depression rather than loneliness.^24,34 One reason that interacting with strangers on social media may be associated with loneliness relates to the various ways in which social media interactions can facilitate misinterpretation. It is known that social media interactions can result in idealization of others’ friendships with each other, exacerbating social comparison.^22,35 This idealization may be of stronger magnitude when NMPs are involved because there is no personal experience to counter that idealization. Interactions with strangers also may be superficial and short-lived, further increasing one’s sense of isolation. Furthermore, these interactions may be particularly problematic among populations that are already lonely, facilitating a downward spiral.^10,34,36

However, other research has demonstrated that some interactions with NMPs may lead to more responsive and supportive interactions. For example, more frequent use of a social media platform connecting patients with cancer who had minor children was associated with less loneliness.²¹ In addition, some studies have found associations between overall numbers of friends and reduced loneliness, regardless of whether those friends were NMPs. For example, a study in Scotland demonstrated an inverse association between the number of Facebook friends and loneliness among 332 university students aged 18 to 29 years.¹⁹ Similar results were identified among a population of 536 Facebook users recruited online.²⁰ While these associations occurred among a relatively narrow population and involved the total number of contacts as the independent variable, they still suggest that more contacts, regardless of NMP status, can be valuable in some instances.

However, the results of the current study suggest that in broad population-based data, those benefits seem to be outweighed by more common social media experiences that are associated with a perception of loneliness. Therefore, it will be valuable for future research to determine optimal ways of leveraging social media for interactions targeted toward forming connections while avoiding more general use that can facilitate misperceptions and a subjective feeling of loneliness. It will also be useful for future work to examine other relevant nuances related to interactions with NMPs. For example, factors such as the platform used, the reason for the connection, and the expectations of emotional support may all be relevant. Qualitative studies may be particularly useful for examining these nuanced factors.

We had hypothesized that having a higher proportion of CPFs as social media contacts would be associated with lower odds of loneliness. However, our findings did not support this hypothesis. It seems intuitive that interacting with CPFs on social media could build on those real-life relationships, alleviating perceptions of loneliness.³⁷ In addition, having CPFs as social media contacts could reduce the likelihood of misunderstandings or gaffes that might lead to a feeling of being marginalized. For these reasons, it is not surprising that some research has found improved mental health among populations who have a larger proportion of social media contacts who are close offline friends.^24,25

One possible explanation of our finding in the context of this earlier work is that our sample involved adults aged 30 to 70 years, while most research to date has focused on adolescents and young adults. Nearly all participants in our sample could be described as “digital immigrants”—individuals who grew up without social media but now interact with it on a continuous basis—while most adolescents and young adults are “digital natives,” defined as individuals who grew up in the digital age. Thus, it may be that people who learned to socialize offline are not as adept at leveraging newer technology to enhance offline experiences. Future research should investigate this possibility. If it is true, educational interventions have the potential to improve the ability of adults to use social media more effectively.

It should be noted, however, that not all prior research demonstrates that online interactions with CPFs are associated with fewer mental health problems, even among adolescents and young adults.²⁶ For example, some research conducted during the COVID-19 pandemic suggested that only in-person communication with close ties significantly protected against loneliness.³⁸ Similarly, a large population-based study in the United States demonstrated that interaction with what people perceived as close, emotionally supportive people paradoxically was associated with poorer emotional health outcomes.³⁴ Other studies have had mixed findings.^9,26,39

Varied results regarding associations between mental health outcomes and the composition of social media contacts may be due to nuanced context-related factors, such as people’s reasons for using social media, the social media platforms used, the in-person network size, and the personality of the user.^39
-41 The type of interactions that users have on social media can also modify the perceived closeness of the relationships between themselves and their social media contacts.²⁶ Thus, it would be useful for future work to examine various nuances such as these.

Limitations

This study had several limitations. First, the study’s cross-sectional design precluded interpretation of directionality and/or causality. While interactions with strangers may lead to loneliness, lonely people may also seek out strangers as social media contacts. Longitudinal assessments would help examine directionality. Second, it was difficult to measure a complex construct such as “closeness of social media ties.” Further evaluation of reliability and validity of these measures could be investigated by using more rigorous designs. Third, because our sample included only adults aged 30 to 70 years, inferences could not be made outside this demographic. Fourth, while we did apply survey weights accounting for sex, race and ethnicity, sexual orientation, and educational attainment to make results as nationally representative as possible, it should be noted that large online panel data such as those from Qualtrics can be limited by data quality in ways that cannot be fully corrected by using survey weights. Fifth, although we included a comprehensive set of sociodemographic control variables, including marital status, we did not examine related assessments such as companionship.

Conclusion

This nationally representative study of US adults demonstrates that having more strangers as social media contacts is associated with increased loneliness; however, having more close contacts is not associated with decreased loneliness. These results suggest that people experiencing loneliness may wish to critically examine their interactions with strangers on social media and to prioritize in-person connections over social media ones, even when those social media connections are considered close.

Footnotes

ORCID iDs

Brian A. Primack, MD, PhD

Hyosin Kim, PhD

Geethika Koneru, MD, MPH

Memuna Aslam, MS

Jessica R. Gorman, PhD, MPH

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The study was funded by the National Institutes of Health (award R01-CA225773). The funders had no role in the study design, data collection, analysis, preparation of the article, or approval of the article.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

Office of the Surgeon General. Our Epidemic of Loneliness and Isolation: The US Surgeon General’s Advisory on the Healing Effects of Social Connection and Community. US Department of Health and Human Services; 2023.

The Cigna Group. Loneliness in America. 2025. Accessed March 26, 2026. https://newsroom.thecignagroup.com/loneliness-in-america

Mann

Wang

Pearce

, et al. Loneliness and the onset of new mental health problems in the general population. Soc Psychiatry Psychiatr Epidemiol. 2022;57(11):2161-2178. doi:10.1007/s00127-022-02261-7

Valtorta

Kanaan

Gilbody

Ronzi

Hanratty

Loneliness and social isolation as risk factors for coronary heart disease and stroke: systematic review and meta-analysis of longitudinal observational studies. Heart. 2016;102(13):1009-1016. doi:10.1136/heartjnl-2015-308790

Hosseinbor

Yassini Ardekani

Bakhshani

Emotional and social loneliness in individuals with and without substance dependence disorder. Int J High Risk Behav Addict. 2014;3(3):e22688. doi:10.5812/ijhrba.22688

Steptoe

Shankar

Demakakos

Wardle

Social isolation, loneliness, and all-cause mortality in older men and women. Proc Natl Acad Sci U S A. 2013;110(15):5797-5801. doi:10.1073/pnas.1219686110

Chen

Gao

Effects of social media self-efficacy on informational use, loneliness, and self-esteem of older adults. Int J Human–Computer Interact. 2023;39(5):1121-1133. doi:10.1080/10447318.2022.2062855

Cauberghe

Van Wesenbeeck

De Jans

Hudders

Ponnet

How adolescents use social media to cope with feelings of loneliness and anxiety during COVID-19 lockdown. Cyberpsychol Behav Soc Netw. 2021;24(4):250-257. doi:10.1089/cyber.2020.0478

Mao

Huang

JJ.

The double-edged sword effects of active social media use on loneliness: the roles of interpersonal satisfaction and fear of missing out. Front Psychol. 2023;14:1108467. doi:10.3389/fpsyg.2023.1108467

10.

Primack

Shensa

Sidani

, et al. Social media use and perceived social isolation among young adults in the US. Am J Prev Med. 2017;53(1):1-8. doi:10.1016/j.amepre.2017.01.010

11.

Dong

Lee

Wang

Shannon

DM.

Impact on social capital and learning engagement due to social media usage among the international students in the US. Educ Inf Technol (Dordr). Published online December 14, 2022. doi:10.1007/s10639-022-11520-8

12.

Panzarasa

Griffiths

Sastry

De Simoni

Social medical capital: how patients and caregivers can benefit from online social interactions. J Med Internet Res. 2020;22(7):e16337. doi:10.2196/16337

13.

Marciano

Lin

Sato

Saboor

Viswanath

Does social media use make us happy? A meta-analysis on social media and positive well-being outcomes. SSM–Mental Health. 2024;6:100331. doi:10.1016/j.ssmmh.2024.100331

14.

Ivie

Pettitt

Moses

Allen

NB.

A meta-analysis of the association between adolescent social media use and depressive symptoms. J Affect Disord. 2020;275:165-174. doi:10.1016/j.jad.2020.06.014

15.

Howarth

Worldwide daily social media usage (new 2025 data). 2025. Accessed March 26, 2026. https://explodingtopics.com/blog/social-media-usage

16.

Primack

Karim

Shensa

Bowman

Knight

Sidani

JE.

Positive and negative experiences on social media and perceived social isolation. Am J Health Promot. 2019;33(6):859-868. doi:10.1177/0890117118824196

17.

Rzewnicki

Shensa

Levenson

Primack

Sidani

JE.

Associations between positive and negative social media experiences and sleep disturbance among young adults. 2020;6(5):671-675. doi:10.1016/j.sleh.2020.02.013

18.

Berkman

Glass

Brissette

Seeman

TE.

From social integration to health: Durkheim in the new millennium. Soc Sci Med. 2000;51(6):843-857. doi:10.1016/s0277-9536(00)00065-4

19.

Phu

Gow

AJ.

Facebook use and its association with subjective happiness and loneliness. Comput Hum Behav. 2019;92:151-159. doi:10.1016/j.chb.2018.11.020

20.

Jin

How lonely people use and perceive Facebook. Comput Hum Behav. 2013;29(6):2463-2470. doi:10.1016/j.chb.2013.05.034

21.

Kosugi

Nishiguchi

Miura

, et al. Association between loneliness and the frequency of using online peer support groups among cancer patients with minor children: a cross-sectional web-based study. J Pain Symptom Manage. 2021;61(5):955-962. doi:10.1016/j.jpainsymman.2020.09.035

22.

Warrender

Milne

How use of social media and social comparison affect mental health. Nursing Times Online. 2020;116(3):56-59.

23.

Costa

Patrão

Machado

Problematic internet use and feelings of loneliness. Int J Psychiatry Clin Pract. 2019;23(2):160-162. doi:10.1080/13651501.2018.1539180

24.

Shensa

Sidani

Escobar-Viera

, et al. Real-life closeness of social media contacts and depressive symptoms among university students. J Am Coll Health. 2018;66(8):747-753. doi:10.1080/07448481.2018.1440575

25.

Riley

Stuart

Barber

BL.

Contemporary friendships and social vulnerability among youth: understanding the role of online and offline contexts of interaction in friendship quality. J Soc Pers Relat. 2021;38(12):3451-3471. doi:10.1177/02654075211029384

26.

Taylor

Zhao

Bazarova

NN.

Social media and close relationships: a puzzle of connection and disconnection. Curr Opin Psychol. 2022;45:101292. doi:10.1016/j.copsyc.2021.12.004

27.

Qualtrics Survey Platform. Qualtrics. Accessed March 26, 2026. https://www.qualtrics.com

28.

Qualtrics. Online research panels and samples for surveys. 2025. Accessed March 26, 2026. https://www.qualtrics.com/research-services/online-sample

29.

Patient-Reported Outcomes Measurement Information System. Social isolation scoring manual. 2021. Accessed March 26, 2026. https://www.healthmeasures.net/images/PROMIS/manuals/Scoring_Manuals_/PROMIS_Social_Isolation_Scoring_Manual.pdf

30.

Hosmer

Jr Lemeshow

Sturdivant

RX.

Applied Logistic Regression. 3rd ed. Wiley; 2013.

31.

Gelles-Watnick

Americans’ use of mobile technology and home broadband. Pew Research Center. January 31, 2024. Accessed March 26, 2026. https://www.pewresearch.org/internet/2024/01/31/americans-use-of-mobile-technology-and-home-broadband

32.

Pew Research Center. Social media fact sheet. November 13, 2024. Updated November 20, 2025. Accessed March 26, 2026. https://www.pewresearch.org/internet/fact-sheet/social-media

33.

Yavich

Davidovitch

“Together but separate”—what is the correlation, if any, between social loneliness, virtual friendship, and use of social networks. J Educ Train Stud. 2019;7(10):1. doi:10.11114/jets.v7i10.4395

34.

Shensa

Sidani

Escobar-Viera

Switzer

Primack

Choukas-Bradley

Emotional support from social media and face-to-face relationships: associations with depression risk among young adults. J Affect Disord. 2010;260:38-44. doi:10.1016/j.jad.2019.08.092

35.

Chao

Lei

Jiang

Yang

TikTok use and psychosocial factors among adolescents: comparisons of non-users, moderate users, and addictive users. Psychiatry Res. 2023;325:115247. doi:10.1016/j.psychres.2023.115247

36.

Primack

Shensa

Sidani

Escobar-Viera

Fine

MJ.

Temporal associations between social media use and depression. Am J Prev Med. 2021;60(2):179-188. doi:10.1016/j.amepre.2020.09.014

37.

Meng

Chung

Cox

Linking network structure to support messages: effects of brokerage and closure on received social support. J Commun. 2016;66(6):982-1006. doi:10.1111/jcom.12268

38.

Kovacs

Caplan

Grob

King

Social networks and loneliness during the COVID-19 pandemic. Socius: Sociological Research for a Dynamic World. 2021;7:2378023120985254. doi:10.1177/2378023120985254

39.

Shensa

Sidani

Lin

Bowman

Primack

BA.

Social media use and perceived emotional support among US young adults. J Community Health. 2016;41(3):541-549. doi:10.1007/s10900-015-0128-8

40.

Matthews

Arseneault

Bryan

, et al. Social media use, online experiences, and loneliness among young adults: a cohort study. Ann N Y Acad Sci. 2025;1548(1):194-205. doi:10.1111/nyas.15370

41.

Smith

Leonis

Anandavalli

Belonging and loneliness in cyberspace: impacts of social media on adolescents’ well-being. Aust J Psychol. 2021;73(1):12-23. doi:10.1080/00049530.2021.1898914