As a social species, humans tend to have a strong need for social connection. The negative emotions experienced when individuals’ needs for social connection are not met are collectively referred to as loneliness (Abdellaoui et al., 2019). Investigation of the correlates and effects of loneliness is important because loneliness increases the chance of psychological distress, physical health problems, and even mortality (Lim et al., 2020). Previous researchers have found that loneliness may be influenced by personality, which means not everyone is equally susceptible to loneliness given similar levels of objective social isolation (Buecker et al., 2020; Schermer & Martin, 2019). Accordingly, we have expanded on previous examinations of loneliness and personality by exploring not only the relationship between interpersonal sensitivity and loneliness, but also the largely unknown influence of trait interpersonal sensitivity on loneliness.

Interpersonal sensitivity (IS), as defined by Boyce and Parker (1989), is a personality trait characterized by excessive awareness of others’ feelings and behavior. Individuals with IS are continually preoccupied with interpersonal relationships, vigilantly regarding others’ moods and behavior (Bell & Freeman, 2014; Boyce & Parker, 1989). This stable personality trait leads to continuous attention to negative social evaluation (Boyce & Parker, 1989). Other researchers, however, regard IS as a psychological symptom reflecting the temporary symptomatic psychological patterns in psychiatric patients or in society (Wedgeworth et al., 2017). To distinguish between them, we have used the terms state IS to refer to temporal and fluctuating IS levels, and trait IS to refer to general or chronic IS levels, the latter of which we used as the focal variable in this study.

Although Jiang et al. (2019) found a positive correlation between state IS and loneliness in a sample of gay Chinese men, the relationship for trait IS still unclear. However, as a key component in trait IS, rejection sensitivity is positively correlated with loneliness (Watson & Nesdale, 2012). This is because individuals with rejection sensitivity react intensely to rejection, which undermines their ability to develop and benefit from positive relationships; further, poor-quality relationships and loneliness are closely related (Zhang et al., 2015). Trait IS (vs. rejection sensitivity) has a broader definition, showing a strong reaction not only to rejection but also in general relationships. Thus, as individuals with trait IS are more likely to feel discomfort during interpersonal interactions, this may lead to loneliness. Those with trait IS are also susceptible to symptoms such as depression and anxiety (Harb et al., 2002; Otani et al., 2018). Otani et al.’s (2018) finding that trait IS is closely related to core negative self-beliefs suggests that this personality trait can be regarded as a cognitive vulnerability to depression. Social anxiety characteristics also correspond to many aspects of the definition of trait IS: interpersonal vigilance, fear of rejection, misinterpretation of others’ behavior, inferiority thoughts, nonassertive behavior, and avoidance of interpersonal situations (Harb et al., 2002). Further, depression and anxiety interrelate with loneliness (Domènech-Abella et al., 2019). Thus, it seems reasonable that trait IS can be an important personality predictor of loneliness.

Individuals with trait IS, especially the attribute of high sensitivity to external evaluation, incorporate more negative external evaluation into their own concepts (Boyce & Parker, 1989). That is, trait IS is closely related to core negative self-beliefs. Previous researchers have found that self-esteem relates to confidence, adjustment, and self-belief (Zhao et al., 2017). Therefore, self-blame caused by trait IS leads to a lack of personal value and ability, further damaging self-esteem. It is proposed in mental dissonance theory that higher self-esteem leads to less loneliness (van Baarsen, 2002). Researchers have found that high susceptibility to loneliness in early adulthood is attributable to low self-esteem (Hall & Pennington, 2013; Zhao et al., 2017). These findings show that self-esteem is an important influencing factor in loneliness, because individuals with low self-esteem blame themselves for social failure or for having low social contact (Zhao et al., 2017), thereby reinforcing their loneliness.

Because they are overly sensitive to criticism and rejection, individuals with high trait IS continually feel uncomfortable in social interactions (Masillo et al., 2012) and exhibit interpersonal avoidance behavior (Hiçdurmaz & Öz, 2016), which impedes their perceived social support. For instance, Siegel and colleagues (2007) found that trait IS was significantly associated with lower perceived partner support. This result occurred because the feeling of embarrassment and vulnerability caused by trait IS can lead to avoidance and poor communication. Perceived social support is not the same as actual support from a social network. It is a cognitive component that is a perception of a close relationship (Wohn et al., 2016), whereas loneliness is an affective component involving a subjective feeling of distress when social connections are perceived to be lacking. Perceived social support has been found to negatively predict loneliness (Cao & Liang, 2020; Segrin & Passalacqua, 2010); therefore, we proposed that perceived social support would mediate the relationship between trait IS and loneliness.

In addition, the relationship between self-esteem and perceived social support appears to be bidirectional, such that the level of self-esteem predicts perceived social support, and vice versa (Lee et al., 2014). Individuals with low (vs. high) self-esteem are more likely to perceive others’ behavior more negatively and to show excessive social avoidance (Routledge et al., 2010), which reduces perceived social support (Świtaj et al., 2015; Wohn et al., 2016). In contrast, perceived social support fosters feelings of belonging, security, and a sense of control over the environment, which may enhance self-esteem (Lee et al., 2014). Therefore, we believe it is necessary to re-examine the self-esteem–perceived social support relationship, which we proposed would be bidirectional and predictive.

Method

Participants and Procedure

We collected data from 761 undergraduates (403 women, 358 men; M_age = 19.36 years, SD = 1.20) at two universities in Shanghai and Nanjing, China. Of the original distributed 804 questionnaires, 43 were excluded owing to respondents failing to complete all the items. The study was approved by the Ethics Committee of East China Normal University. Participants all gave informed consent before taking part in the study. In the first data collection phase, 397 participants were recruited through an online survey invitation to complete the Interpersonal Sensitivity Measure. Participants were compensated with a random amount of money (USD 0.50–1.00). In the second data collection phase, another 364 participants were recruited to complete a series of offline questionnaires within 20 minutes, including demographics and the measures of IS, self-esteem, perceived social support, and loneliness. On completion of the questionnaires, participants were compensated with a small gift of a pen worth about USD 1.00.

Measures

Interpersonal Sensitivity
We adopted the 36-item Interpersonal Sensitivity Measure (Boyce & Parker, 1989) to measure participants’ trait IS. Participants assess the items on a 4-point Likert scale ranging from 1 (very inconsistent) to 4 (very consistent). Higher scores indicate greater trait IS. As the factor structure and psychometric properties of the scale among Chinese people is unknown, and too many items in the original scale may cause varying fatigue effects, we deleted 17 items when simplifying the scale, and measured its reliability and construct validity (see Appendix Table A1). Cronbach’s alpha for the simplified scale was .82, and for the subscales it ranged between .72 and .85 in this study.

Loneliness
We adopted a Chinese version (Cao & Liang, 2020) of the UCLA Loneliness Scale to assess participants’ loneliness with 20 items (11 positive-scored items and nine reverse-scored items). Participants assess the items on a 4-point Likert scale ranging from 1 (never) to 4 (always). Sample items are “I feel isolated from others” and “I lack companionship.” Higher scores indicate greater loneliness. Cronbach’s alpha was .74 in this study.

Self-Esteem
The 10-item Rosenberg Self-Esteem Scale (Rosenberg, 1965) was adopted to assess participants’ global self-esteem. Participants assessed the items on a 4-point Likert scale ranging from 1 (very inconsistent) to 4 (very consistent). A sample item is “I am able to do things as well as most other people.” The eighth question (“I hope I can win more respect for myself”) was deleted because it is unsuitable for the measurement of Chinese self-esteem (Lumei, 2006). Cronbach’s alpha was .78 in this study.

Perceived Social Support
We adopted the 12-item Chinese version (Cao & Liang, 2020) of Zimet et al.’s (1988) Multidimensional Scale of Perceived Social Support to assess perceived social support. Three aspects of social support were assessed: family (FA), friends (FR), and significant others (SO). Sample items are “I can talk about my problems with my family” (FA), “My friends really try to help me” (FO), and “There is a special person who is around when I am in need” (SO). Participants evaluate the items on a 5-point Likert scale ranging from 1 (completely disagree) to 5 (completely agree). Higher scores indicate greater perceived social support. Cronbach’s alpha for the scale was .90, and for the subscales for FA, FR, and SO, were .79, .82, and .89, respectively, in this study.

Results

Simplification of the Interpersonal Sensitivity Measure

We used SPSS to perform an item analysis and an exploratory factor analysis (EFA), and Amos to perform a confirmatory factor analysis (CFA). First, we performed the project analysis on the first-phase data (n = 397), and deleted Item 20 (the correlation coefficient with the total scale score was nonsignificant), then Items 21 and 25 (composite reliability values > 3) sequentially. The scale’s internal consistency after we had deleted each item was not substantially improved (see Appendix Table A1).

We then performed a factor analysis of the remaining items using half of the randomly selected data (n = 198). The results demonstrate that the data were suitable for EFA (Kaiser–Meyer–Olkin = .82, Bartlett’s test of sphericity, χ² = 2269.48, p < .001). We employed principal component analysis to extract the common factors of the questionnaire items, and orthogonally rotated the factors, resulting in a rotated factor matrix. Contemporarily, we deleted, in sequence, Items 15, 19, and 34 because of low factor loadings (< .40), and Items 2, 10, 12, 13, and 29 because of high cross loadings (absolute value of the load difference between the two factors of the item < .20). We also deleted factors containing too few items (each factor contains at least three items) and deleted Items 6, 7, 16, 28, 31, and 32 in sequence. We eventually extracted five common factors with characteristic roots greater than 1 (19 items), and the total percentage of explained variance was 56.45% (see Appendix Table A2).

Finally, we used the remaining data (n = 199) to perform a CFA of the remaining items. To test the model fit, we adopted the conventional cutoff criterion, which is .90 or higher for goodness-of-fit index (GFI) and comparative fit index (CFI) and .08 or lower for root mean square error of approximation (RMSEA) and standardized root mean square residual (SRMR; Kline, 2011). The results show that the five-factor model provided a suitable fit to the data, χ² = 288.31, χ²/df = 2.07, RMSEA = .05, GFI = .93, CFI = .90, SRMR = .03. Each path coefficient reached significance. We referenced the factors in the original scale when naming each factor: fragile inner self (six items), separation anxiety (three items), interpersonal awareness (four items), need for approval (three items), and timidity (three items).

Descriptive Statistics

Means, standard deviations, and correlations for the study variables are presented in Table 1.

Structural Model

Amos was used to conduct structural equation modeling to test our hypothesized model, with maximum likelihood estimation. We also tested the mediation effects with the bootstrapping method, using 1,000 bootstrapped resamples to calculate 95% confidence intervals (CIs). We used a parceling approach when analyzing the hypothesized model to reduce the chance of residuals being correlated (Little et al., 2002). The model detected if the effect of IS on loneliness was mediated by self-esteem and perceived social support (see Figure 1). The hypothesized model had a good fit to the data, χ²/df = 1.43, CFI = .97, GFI = .97, Tucker–Lewis Index (TLI) = .96, incremental fit index (IFI) = .97, RMSEA = .05. However, there were three nonsignificant pathways, from IS to perceived social support, β = .72, p = .16, from trait IS to loneliness, β = .20, p = .10, and from perceived social support to self-esteem, β = - .12, p = .35. After we had removed these pathways, the indices of the revised model had a good fit to the data, χ²/df = 1.44, CFI = .97, GFI = .97, TLI = .96, IFI = .97, RMSEA = .05. Results show that the indirect effect of trait IS on loneliness through self-esteem was significant, β = .18, 95% CI [0.09, 0.28]. The indirect effect of trait IS on loneliness, through both self-esteem and perceived social support, was also significant, β = .08, 95% CI [0.04, 0.14].

Discussion

We examined the underlying mechanisms of self-esteem and perceived social support in the relationship between trait IS and individuals’ feelings of loneliness. Our results show that trait IS had a significant positive predictive effect on loneliness, the indirect effect of self-esteem was statistically significant, and the serial indirect effect of self-esteem with perceived social support was statistically significant.

Our finding that individuals with (vs. without) trait IS were more likely to feel lonely corroborates previous results that state IS (Jiang et al., 2019) and rejection sensitivity (Watson & Nesdale, 2012) lead to loneliness. However, it remains undefined how trait IS significantly increases individuals’ loneliness feelings. Our examination of the trait IS–loneliness relationship adds knowledge to the literature. We also explored the mediating roles of self-esteem and perceived social support to confirm that they interpret the strength of the trait IS–loneliness relationship. The findings show the indirect effect of self-esteem on this relationship, highlighting that trait IS causes low self-esteem, thereby triggering loneliness because self-blame caused by trait IS leads to a lack of personal value and ability, further lowering self-esteem (Bell & Freeman, 2014; Boyce & Parker, 1989).

Notably, we did not strengthen Siegel et al.’s (2007) finding that trait IS leads to less perceived social support. In model testing, trait IS did not directly predict perceived social support, but indirectly predicted perceived social support through self-esteem. Individuals with positive self-evaluation can adopt active and adaptive coping strategies (Świtaj et al., 2015). However, lower self-esteem reduces the use of beneficial coping strategies for seeking social support, and reduced perceived social support leads to intensification of feelings of loneliness. Further, Wohn et al. (20126) showed that the correlation between individuals’ perceived social support and social support was moderate: Those with low (vs. high) self-esteem were more likely to have less stable relationships, and were less likely to subjectively see the provider as a supportive individual (see also Świtaj et al., 2015). Overall, we believe that individuals with trait IS incorporate negative external assessments in their minds, which results in lower self-esteem, and that individuals with low self-esteem perceive less support from important others because of their low social relationship evaluation, which further increases their sense of loneliness.

Implications

Our results contribute to the trait IS–loneliness relationship research in three ways. First, we generated a short version of the Interpersonal Sensitivity Measure, which is more suitable for use by Chinese people. We deleted some items, tested the factor structure of the revised scale, and provided additional evidence for the measure’s validity and reliability. The structure of the retained factors is similar to that of the original measure (Boyce & Parker, 1989). Second, as few researchers have explored the potential influence of trait IS on loneliness, we have attempted to fill this gap. Finally, we have gained new insights into the underlying mechanism of trait IS’s influence on loneliness, by showing that trait IS is negatively correlated with self-esteem, which predicts perceived social support, thereby contributing to feelings of loneliness.

Limitations and Directions for Future Research

There are some limitations in this study. First, it is unclear if alternative models would fit the data better than the proposed model. Although our findings support the idea that self-esteem is an antecedent of perceived social support, which is consistent with previous studies (Tan et al., 2016), future researchers could explore the impact of perceived social support on self-esteem (Cao & Liang, 2017). Second, our conclusions cannot reflect the dynamic causality of the model owing to the use of a cross-sectional research design. Future researchers could use experimental or longitudinal methods to verify causality. Third, we focused only on self-esteem and perceived social support. Future researchers could use other variables (e.g., social exclusion, interpersonal satisfaction, field dependence/independence) to explain the trait IS–loneliness relationship.

Conclusion

Our serial mediation model offers new insights for a pathway from self-esteem to perceived social support in the trait IS–loneliness relationship. The model provides a comprehensive viewpoint on why individuals with high trait IS tend to experience a high level of loneliness, and provides valuable guidance on the implementation of psychological interventions aimed at reducing the loneliness of these individuals. It is critical for individuals to be taught how to form high self-esteem objectively and, thereby, improve their perceived social support.

Appendix

Table A1. Item Analysis of Interpersonal Sensitivity Measure

Table A2. Factor Loadings and Additional Estimates for the 20-Item Chinese Interpersonal Sensitivity Measure–Revised

Abdellaoui, A., Chen, H.-Y., Willemsen, G., Ehli, E. A., Davies, G. E., Verweij, K. J. H., … Cacioppo, J. T. (2019). Associations between loneliness and personality are mostly driven by a genetic association with neuroticism. Journal of Personality, 87(2), 386–397.
https://doi.org/10.1111/jopy.12397

Bell, V., & Freeman, D. (2014). A pilot trial of cognitive behavioural therapy for interpersonal sensitivity in individuals with persecutory delusions. Journal of Behavior Therapy and Experimental Psychiatry, 45(4), 441–446.
https://doi.org/10.1016/j.jbtep.2014.06.001

Boyce, P., & Parker, G. (1989). Development of a scale to measure interpersonal sensitivity. Australian and New Zealand Journal of Psychiatry, 23(3), 341–351.
https://doi.org/10.1177/000486748902300320

Buecker, S., Maes, M., Denissen, J. J. A., & Luhmann, M. (2020). Loneliness and the Big Five Personality Traits: A meta-analysis. European Journal of Personality, 34(1), 8–28.
https://doi.org/10.1002/per.2229

Cao, Q., & Liang, Y. (2020). Perceived social support and life satisfaction in drug addicts: Self-esteem and loneliness as mediators. Journal of Health Psychology, 25(7), 976–985.
https://doi.org/10.1177/1359105317740620

Domènech-Abella, J., Mundó, J., Haro, J. M., & Rubio-Valera, M. (2019). Anxiety, depression, loneliness and social network in the elderly: Longitudinal associations from The Irish Longitudinal Study on Ageing (TILDA). Journal of Affective Disorders, 246, 82–88.
https://doi.org/10.1016/j.jad.2018.12.043

Hall, J. A., & Pennington, N. (2013). Self-monitoring, honesty, and cue use on Facebook: The relationship with user extraversion and conscientiousness. Computers in Human Behavior, 29(4), 1556–1564.
https://doi.org/10.1016/j.chb.2013.01.001

Harb, G. C., Heimberg, R. G., Fresco, D. M., Schneier, F. R., & Liebowitz, M. R. (2002). The psychometric properties of the Interpersonal Sensitivity Measure in social anxiety disorder. Behaviour Research and Therapy, 40(8) 961–979.
https://doi.org/10.1016/S0005-7967(01)00125-5

Hiçdurmaz, D., & Öz, F. (2016). Interpersonal sensitivity, coping ways and automatic thoughts of nursing students before and after a cognitive-behavioral group counseling program. Nurse Education Today, 36, 152–158.
https://doi.org/10.1016/j.nedt.2015.09.006

Jiang, D., Hou, Y., Chen, X., Wang, R., Fu, C., Li, B., … Liu, X. (2019). Interpersonal sensitivity and loneliness among Chinese gay men: A cross-sectional survey. International Journal of Environmental Research and Public Health, 16(11), Article 2039.
https://doi.org/10.3390/ijerph16112039

Kline, R. B. (2011). Principles and practice of structural equation modeling (3rd ed.). Guilford Press.

Lee, C., Dickson, D. A., Conley, C. S., & Holmbeck, G. N. (2014). A closer look at self-esteem, perceived social support, and coping strategy: A prospective study of depressive symptomatology across the transition to college. Journal of Social and Clinical Psychology, 33(6), 560–585.
https://doi.org/10.1521/jscp.2014.33.6.560

Lim, M. H., Holt-Lunstad, J., & Badcock, J. C. (2020). Loneliness: Contemporary insights into causes, correlates, and consequences. Social Psychiatry and Psychiatric Epidemiology, 55(7), 789–791.
https://doi.org/10.1007/s00127-020-01891-z

Little, T. D., Cunningham, W. A., Shahar, G., & Widaman, K. F. (2002). To parcel or not to parcel: Exploring the question, weighing the merits. Structural Equation Modeling: A Multidisciplinary Journal, 9(2), 151–173.
https://doi.org/10.1207/S15328007SEM0902_1

Lumei, T. (2006). Shortcoming and merits of the Chinese version of the Rosenberg (1965) Self-Esteem Scale [In Chinese]. Psychological Exploration, 26(2), 88–92.

Masillo, A., Day, F., Laing, J., Howes, O., Fusar-Poli, P., Byrne, M., … Valmaggia, L. R. (2012). Interpersonal sensitivity in the at-risk mental state for psychosis. Psychological Medicine, 42(9), 1835–1845.
https://doi.org/10.1017/S0033291711002996

Otani, K., Suzuki, A., Matsumoto, Y., & Shirata, T. (2018). Close relation of interpersonal sensitivity with negative core beliefs about the self, the central construct of cognitive vulnerability to depression. Psychiatry Research, 263, 162–165.
https://doi.org/10.1016/j.psychres.2018.03.015

Rosenberg, M. (1965). Society and the adolescent self-image. Princeton University Press.

Routledge, C., Ostafin, B., Juhl, J., Sedikides, C., Cathey, C., & Liao, J. (2010). Adjusting to death: The effects of mortality salience and self-esteem on psychological well-being, growth motivation, and maladaptive behavior. Journal of Personality and Social Psychology, 99(6), 897–916.
https://doi.org/10.1037/a0021431

Schermer, J. A., & Martin, N. G. (2019). A behavior genetic analysis of personality and loneliness. Journal of Research in Personality, 78, 133–137.
https://doi.org/10.1016/j.jrp.2018.11.011

Segrin, C., & Passalacqua, S. A. (2010). Functions of loneliness, social support, health behaviors, and stress in association with poor health. Health Communication, 25(4), 312–322.
https://doi.org/10.1080/10410231003773334

Siegel, S. D., Molton, I., Penedo, F. J., Llabre, M. M., Kinsinger, D. P., Traeger, L., … Antoni, M. H. (2007). Interpersonal sensitivity, partner support, patient–physician communication, and sexual functioning in men recovering from prostate carcinoma. Journal of Personality Assessment, 89(3), 303–309.
https://doi.org/10.1080/00223890701629847

Świtaj, P., Grygiel, P., Anczewska, M., & Wciórka, J. (2015). Experiences of discrimination and the feelings of loneliness in people with psychotic disorders: The mediating effects of self-esteem and support seeking. Comprehensive Psychiatry, 59, 73–79.
https://doi.org/10.1016/j.comppsych.2015.02.016

Tan, C.-S., Krishnan, S., & Lee, Q.-W. (2016). The role of self-esteem and social support in the relationship between extraversion and happiness: A serial mediation model. Current Psychology, 36(3), 556–564.
https://doi.org/10.1007/s12144-016-9444-0

van Baarsen, B. (2002). Theories on coping with loss: The impact of social support and self-esteem on adjustment to emotional and social loneliness following a partner’s death in later life. The Journals of Gerontology: Series B: Psychological Sciences and Social Sciences, 57(1), S33–S42.
https://doi.org/10.1093/geronb/57.1.S33

Watson, J., & Nesdale, D. (2012). Rejection sensitivity, social withdrawal, and loneliness in young adults. Journal of Applied Social Psychology, 42(8), 1984–2005.
https://doi.org/10.1111/j.1559-1816.2012.00927.x

Wedgeworth, M., LaRocca, M. A., Chaplin, W. F., & Scogin, F. (2017). The role of interpersonal sensitivity, social support, and quality of life in rural older adults. Geriatric Nursing, 38(1), 22–26.
https://doi.org/10.1016/j.gerinurse.2016.07.001

Wohn, D. Y., Carr, C. T., & Hayes, R. A. (2016). How affective is a “like”?: The effect of paralinguistic digital affordances on perceived social support. Cyberpsychology, Behavior and Social Networking, 19(9), 562–566.
https://doi.org/10.1089/cyber.2016.0162

Zhang, B., Gao, Q., Fokkema, M., Alterman, V., & Liu, Q. (2015). Adolescent interpersonal relationships, social support and loneliness in high schools: Mediation effect and gender differences. Social Science Research, 53, 104–117.
https://doi.org/10.1016/j.ssresearch.2015.05.003

Zhao, L., Zhang, X., & Ran, G. (2017). Positive coping style as a mediator between older adults’ self-esteem and loneliness. Social Behavior and Personality: An international journal, 45(10), 1619–1628.
https://doi.org/10.2224/sbp.6486

Zimet, G. D., Dahlem, N. W., Zimet, S. G., & Farley, G. K. (1988). The Multidimensional Scale of Perceived Social Support. Journal of Personality Assessment, 52(1), 30–41.
https://doi.org/10.1207/s15327752jpa5201_2

This work was supported by the National Social Science Foundation of China (15ZDB121).

Yongfang Liu, Department of Psychology and Cognitive Science, East China Normal University, 3663 North Zhongshan Road, Shanghai 200062, People’s Republic of China. Email: [email protected]