Abstract
Background
Thus far, no studies have focused on hugging and social disconnectedness outcomes comprehensively (loneliness, social isolation, social withdrawal) in the post-pandemic era. Our aim was to examine whether hugging is associated with loneliness, social isolation, and social withdrawal and whether these associations vary by gender.
Methods
We used cross-sectional data collected via an online survey conducted in January 2025 (N = 3,270 adults aged 18 to 74 years across Germany). To ensure representativeness, a quota sampling method was employed based on gender, age, and federal state. Participants reported on their loneliness, social isolation, and social withdrawal and how many individuals they hugged each day. Linear regression models with robust standard errors were used.
Results
After adjusting for multiple covariates, hugging on average one (or more) individuals per day (compared to lack of hugging) was associated with fewer feelings of loneliness, social isolation, and social withdrawal in the total sample. Among men, hugging was associated with reduced feelings of loneliness and social withdrawal but not social isolation; hugging was related to all three social disconnectedness metrics among women and the associations were stronger than for men, as indicated by corresponding interaction terms (hugging x gender).
Conclusion
Our findings suggested an association between hugging frequency and measures of social disconnectedness, particularly among women. Efforts to increase the frequency of hugs, ideally to at least 2–3 individuals on average per day, may be beneficial for social outcomes, pending longitudinal and experimental evidence.
Keywords: embracing, hikikomori, hug, hugging, loneliness, social exclusion, social isolation, social withdrawal
1. Introduction
Loneliness is a negative emotion that refers to the perceived difference between actual and desired social relationships (1), perceived social isolation refers to a feeling of not belonging to the society (2), while and social withdrawal refers to retreating from society (e.g., in terms of one’s relationships and social interactions) (3). These aspects of social disconnection are prevalent across the lifespan, and not limited to old age, and contribute to morbidity and mortality (4, 5) and other unfavorable outcomes (6).
There are various factors are associated with these components of social disconnectedness. Spousal loss, for example, is associated with worse social disconnectedness outcomes (7, 8). Physical and emotional closeness may be important for loneliness, social isolation and social withdrawal (9). Hugs (act of embracing another person, often as a sign of comfort, affection or greeting) can be one such an expression of physical and emotional closeness (10). When individuals are sad, for example, after the end of a relationship, a hug may make them feel less lonely. A daily hug may also intensify relationships as a sign of affection. Similarly, hugs as a sign of greeting may symbolize closeness and a certain depth and quality of friendship.
One previous study that used data from a representative sample of American adults aged 18 to 94 years found that hugging or kissing a romantic partner was significantly associated with less loneliness, whereas hugging or kissing a family member was not significantly associated with loneliness (11). A more recent qualitative study also showed that the lack of hugs during the pandemic was burdensome for low-income adolescents in São Paulo, Brazil (12). However, no studies have focused specifically on hugging and social disconnectedness outcomes more comprehensively (loneliness, social isolation, social withdrawal) in the post-pandemic era. Therefore, we aimed to examine the association between hugging and loneliness, social isolation, and social withdrawal and whether these associations vary by gender. Since the meaning and importance of hugs may vary depending on gender (with potentially stronger associations among women) (13, 14), the role of gender in the association between hugging and social disconnectedness outcomes will be explored. Such knowledge is of great importance to better characterize individuals at risk of high levels of loneliness, social isolation and social withdrawal. This in turn is important as such social disconnectedness outcomes can eventually lead to chronic conditions and premature death (4, 5).
2. Methods
2.1. Sample
Data were drawn from an online survey that was conducted in January 2025, involving 3,270 participants aged 18 to 74 years who were residing in Germany. The only inclusion criteria were being within the specified age range and living in Germany. The survey was administered by Bilendi (a reputable ISO-certified market research company). To ensure that the sample accurately reflected the German adult population (in terms of gender, age, and federal state) a quota sampling strategy was used.
2.2. Dependent variables
Loneliness was assessed using the 6-item De Jong Gierveld (DJG) Loneliness Scale (15). Items were averaged. The average score could range from 0 to 6; higher scores indicated higher loneliness levels (Cronbach’s alpha = 0.77; McDonald’s omega = 0.75).
Social isolation was measured using the 4-item Bude and Lantermann tool (2). Responses were averaged to create a final score that ranged from 1 to 4, higher scores indicated higher levels of social isolation (Cronbach’s alpha and McDonald’s omega both = 0.91).
Social withdrawal was measured with the validated German version (16) of the 25-item Hikikomori Questionnaire (HQ-25) (17). Items were summed, and scores ranged from 0 to 100, higher scores indicated more social withdrawal (Cronbach’s alpha and McDonald’s omega both = 0.92).
2.3. Main independent variable: frequency of hugs
Participants reported the number of individuals they hugged on average on a day. Following former research (18), individuals could indicate a frequency ranging from 0 to 99 individuals. Based on this and similar to previous research (18), four categories were developed: The first category included individuals who answered 0 (i.e., on average, they did not hug anyone per day). The second category included individuals who hugged one individual per day on average. The third category included people who hugged two or three individuals per day on average. The fourth category included individuals who hugged four or more individuals per day on average.
2.4. Covariates
Covariates for regression analysis were selected based on prior research (19, 20) and theoretical grounds. For example, lifestyle factors may affect the likelihood of hugs (e.g., before or after joint sporting activities). Physical activity can also have a direct effect on outcomes such as loneliness (21). Moreover, health-related factors such as chronic illnesses may affect the ability to maintain social contacts (e.g., receiving hugs) and the quality of relationships (which can affect social outcomes such as loneliness) (22).
Sociodemographic covariates were age, gender (men, women, diverse), marital status (single, divorced, widowed, cohabiting—either married/in a partnership, living separately, but married/in a partnership), size of the city (less than 2,000 inhabitants, 2,000 to less than 5,000 inhabitants, 5,000 to less than 20,000 inhabitants, 20,000 to less than 50,000 inhabitants, 50,000 to less than 100,000 inhabitants, 100,000 to less than 500,000 inhabitants, 500,000 inhabitants and more), education (according to the ISCED-97 framework (23): low, medium, or high), employment status (full-time employed, retired, or other), and religious affiliation (no religious affiliation, Christianity, Islam, other). Lifestyle factors were frequency of physical activity (from no activity to more than 4 h weekly; five categories in total), alcohol intake (daily, several times a week, once a week, 1–3 times monthly, less often, or never), and smoking history (never smoked, former smoker, occasional smoker, or daily smoker).
Health-related factors included self-rated health (single item, from 1 = very poor to 5 = very good), sum of chronic conditions (based on 15 chronic conditions such as cancer or diabetes), and probable depression [Patient Health Questionnaire-9 (PHQ-9) ≥ 10 (24)].
2.5. Statistical analysis
Descriptive statistics were computed to evaluate the characteristics of the sample (also stratified by the frequency of hugs). Unadjusted and adjusted linear regression analyses were calculated with robust standard errors. In the adjusted models, initial covariates were sociodemographic factors. Lifestyle-related factors were then incorporated as additional covariates, followed by health-related covariates. Moreover, moderation by gender was tested (by including interaction terms for frequency of hugging x gender).
Effect sizes were also calculated (Cohen’s d for unadjusted comparisons and partial Eta2 for regression analysis). Cohen’s d = 0.20 reflects a small effect, d = 0.50 reflects a medium effect and d = 0.80 reflects a large effect size (25). Partial eta-squared values can be classified as follows (25): 0.01 as “small,” 0.06 as “medium, and 0.14 as “large”, respectively. McDonald’s omega was calculated using the “omegacoef” command in Stata to estimate internal consistency for the scales (26).
The threshold for statistical significance was set at p < 0.05. All analyses were conducted using Stata Now 18.5 MP-Parallel Edition (StataCorp, College Station, Texas).
3. Results
3.1. Sample characteristics
Characteristics of the sample are in Table 1 (n = 3,270 individuals). Overall, 50.4% of the respondents were women and the average age was 47.0 years (SD: 15.3 years, range 18–74 years).
Table 1.
Sample characteristics stratified by frequency of hugs (n = 3,270).
| Variables | Average number of individuals hugged per day | Total | p-value | |||
|---|---|---|---|---|---|---|
| 0 | 1 | 2 or 3 | 4 or more | |||
| In total | 814 (24.9) | 1,251 (38.3) | 998 (30.5) | 207 (6.3) | 3,270 (100.0) | |
| Gender: N (%) | 0.24 | |||||
| Men | 434 (53.3) | 605 (48.4) | 477 (47.8) | 98 (47.3) | 1,614 (49.4) | |
| Women | 377 (46.3) | 643 (51.4) | 518 (51.9) | 109 (52.7) | 1,647 (50.4) | |
| Diverse | 3 (0.4) | 3 (0.2) | 3 (0.3) | 0 (0.0) | 9 (0.3) | |
| Age: mean (SD) | 49.6 (15.2) | 47.6 (15.9) | 44.6 (14.3) | 43.7 (13.7) | 47.0 (15.3) | <0.001 |
| Marital situation: N (%) | <0.001 | |||||
| Single | 402 (49.4) | 308 (24.6) | 160 (16.0) | 38 (18.4) | 908 (27.8) | |
| Divorced | 133 (16.3) | 77 (6.2) | 52 (5.2) | 15 (7.2) | 277 (8.5) | |
| Widowed | 56 (6.9) | 27 (2.2) | 17 (1.7) | 3 (1.4) | 103 (3.1) | |
| Married, cohabiting with spouse | 192 (23.6) | 790 (63.1) | 737 (73.8) | 148 (71.5) | 1867 (57.1) | |
| Married, not cohabiting with spouse | 31 (3.8) | 49 (3.9) | 32 (3.2) | 3 (1.4) | 115 (3.5) | |
| Educational level: N (%) | <0.001 | |||||
| Low | 114 (14.0) | 94 (7.5) | 107 (10.7) | 22 (10.6) | 337 (10.3) | |
| Medium | 412 (50.6) | 627 (50.1) | 422 (42.3) | 91 (44.0) | 1,552 (47.5) | |
| High | 288 (35.4) | 530 (42.4) | 469 (47.0) | 94 (45.4) | 1,381 (42.2) | |
| Labor force participation: N (%) | <0.001 | |||||
| Full-time employed | 360 (44.2) | 611 (48.8) | 540 (54.1) | 118 (57.0) | 1,629 (49.8) | |
| Retired | 209 (25.7) | 292 (23.3) | 129 (12.9) | 19 (9.2) | 649 (19.8) | |
| Other | 245 (30.1) | 348 (27.8) | 329 (33.0) | 70 (33.8) | 992 (30.3) | |
| City size: N (%) | 0.11 | |||||
| Less than 2,000 inhabitants | 50 (6.1) | 86 (6.9) | 61 (6.1) | 13 (6.3) | 210 (6.4) | |
| 2,000 to less than 5,000 inhabitants | 66 (8.1) | 85 (6.8) | 69 (6.9) | 13 (6.3) | 233 (7.1) | |
| 5,000 to less than 20,000 inhabitants | 197 (24.2) | 270 (21.6) | 210 (21.0) | 47 (22.7) | 724 (22.1) | |
| 20,000 to less than 50,000 inhabitants | 132 (16.2) | 217 (17.3) | 184 (18.4) | 17 (8.2) | 550 (16.8) | |
| 50,000 to less than 100,000 inhabitants | 89 (10.9) | 114 (9.1) | 101 (10.1) | 23 (11.1) | 327 (10.0) | |
| 100,000 to less than 500,000 inhabitants | 124 (15.2) | 204 (16.3) | 177 (17.7) | 40 (19.3) | 545 (16.7) | |
| 500,000 inhabitants and more | 156 (19.2) | 275 (22.0) | 196 (19.6) | 54 (26.1) | 681 (20.8) | |
| Religious affiliation: N (%) | <0.001 | |||||
| No religious affiliation | 392 (48.2) | 628 (50.2) | 365 (36.6) | 74 (35.7) | 1,459 (44.6) | |
| Christianity | 388 (47.7) | 578 (46.2) | 582 (58.3) | 111 (53.6) | 1,659 (50.7) | |
| Islam | 21 (2.6) | 24 (1.9) | 31 (3.1) | 14 (6.8) | 90 (2.8) | |
| Other | 13 (1.6) | 21 (1.7) | 20 (2.0) | 8 (3.9) | 62 (1.9) | |
| Smoking status: N (%) | <0.001 | |||||
| Yes, daily | 221 (27.1) | 246 (19.7) | 219 (21.9) | 51 (24.6) | 737 (22.5) | |
| Yes, sometimes | 49 (6.0) | 117 (9.4) | 140 (14.0) | 41 (19.8) | 347 (10.6) | |
| No, not anymore | 203 (24.9) | 378 (30.2) | 255 (25.6) | 55 (26.6) | 891 (27.2) | |
| Never smoker | 341 (41.9) | 510 (40.8) | 384 (38.5) | 60 (29.0) | 1,295 (39.6) | |
| Alcohol consumption: N (%) | <0.001 | |||||
| Daily | 54 (6.6) | 57 (4.6) | 58 (5.8) | 16 (7.7) | 185 (5.7) | |
| Several times a week | 116 (14.3) | 205 (16.4) | 193 (19.3) | 46 (22.2) | 560 (17.1) | |
| Once a week | 110 (13.5) | 244 (19.5) | 191 (19.1) | 33 (15.9) | 578 (17.7) | |
| 1–3 times a month | 121 (14.9) | 230 (18.4) | 181 (18.1) | 26 (12.6) | 558 (17.1) | |
| Less often | 212 (26.0) | 315 (25.2) | 211 (21.1) | 44 (21.3) | 782 (23.9) | |
| Never | 201 (24.7) | 200 (16.0) | 164 (16.4) | 42 (20.3) | 607 (18.6) | |
| Sports activity: N (%) | <0.001 | |||||
| No sports activity | 322 (39.6) | 312 (24.9) | 181 (18.1) | 28 (13.5) | 843 (25.8) | |
| Less than 1 h a week | 124 (15.2) | 259 (20.7) | 187 (18.7) | 35 (16.9) | 605 (18.5) | |
| Regularly, 1–2 h a week | 160 (19.7) | 309 (24.7) | 287 (28.8) | 57 (27.5) | 813 (24.9) | |
| Regularly, 2–4 h a week | 95 (11.7) | 210 (16.8) | 204 (20.4) | 41 (19.8) | 550 (16.8) | |
| Regularly, more than 4 h a week | 113 (13.9) | 161 (12.9) | 139 (13.9) | 46 (22.2) | 459 (14.0) | |
| Probable depression (PHQ-9): N (%) | <0.01 | |||||
| No | 541 (66.5) | 915 (73.1) | 739 (74.0) | 148 (71.5) | 2,343 (71.7) | |
| Yes | 273 (33.5) | 336 (26.9) | 259 (26.0) | 59 (28.5) | 927 (28.3) | 0.00 |
| Count of chronic conditions: mean (SD) | 2.0 (2.0) | 1.7 (1.7) | 1.5 (1.7) | 1.6 (1.7) | 1.7 (1.8) | <0.001 |
| Self-rated health: mean (SD) | 3.3 (0.9) | 3.5 (0.8) | 3.7 (0.8) | 3.8 (0.8) | 3.6 (0.8) | <0.001 |
| Loneliness (DJG-6): mean (SD) | 4.1 (1.9) | 3.2 (2.0) | 2.9 (2.0) | 2.8 (2.0) | 3.3 (2.0) | <0.001 |
| Social isolation (Bude and Lantermann tool): mean (SD) | 2.1 (0.9) | 1.9 (0.8) | 1.8 (0.8) | 1.9 (0.8) | 2.0 (0.8) | <0.001 |
| Social withdrawal (HQ-25-G): mean (SD) | 46.9 (17.0) | 37.4 (16.0) | 32.6 (16.3) | 33.3 (17.5) | 38.0 (17.3) | <0.001 |
p-values are based on one-way ANOVAs or chi-square tests, as appropriate.
In total, 24.9% of respondents reported that they did not hug anyone on average daily, 38.3% of respondents reported that they hugged 1 individual per day, 30.5% of respondents reported that they hugged 2 to 3 individuals on average per day, and 6.3% of respondents reported that they hugged 4 or more individuals on average per day.
Depending on frequency of hugs, the mean loneliness score varied from 2.8 (SD: 2.0; among individuals hugging 4 or more individuals on average each day) to 4.1 (SD: 1.9; among individuals hugging nobody on average each day). This difference corresponded to an effect size Cohen’s d = 0.7. Mean social isolation varied from 1.8 (SD: 0.8; among individuals hugging two or three individuals on average daily) to 2.1 (SD: 0.9; among individuals hugging nobody on average daily). This difference corresponded to an effect size Cohen’s d = 0.4. Mean social withdrawal varied from 32.6 (SD: 16.3; among individuals hugging two or three individuals on average daily) to 46.9 (SD: 17.0; among individuals hugging nobody on average daily). This difference corresponded to an effect size Cohen’s d = 0.9. Additional details are in Table 1.
3.2. Regression analysis
The results of the unadjusted and adjusted linear regressions (examining the association between the frequency of hugs and loneliness, social isolation and social withdrawal) are in Table 2. Gender-stratified regressions are in Table 3 (men) and Table 4 (women). The fully-adjusted models indicated that hugging on average one or more individuals per day (compared to the lack of hugging) was associated with less loneliness, social isolation, and social withdrawal in the total sample (see Table 2). The largest difference in terms of loneliness was between individuals who hug on average nobody daily, and people who hug at least four people daily on average (β = −1.03, p < 0.001). The largest differences for social isolation and social withdrawal were between individuals who hug two to three individuals on average every day and those who do not hug anyone (β = −0.20, p < 0.001 and β = −10.54, p < 0.001, respectively).
Table 2.
Association of hugging frequency with loneliness, social isolation, and social withdrawal among the total sample based on linear regressions.
| Independent variables | Loneliness | Perceived social isolation | Social withdrawal | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Model 1 | Model 2 | Model 3 | Model 4 | Model 1 | Model 2 | Model 3 | Model 4 | Model 1 | Model 2 | Model 3 | Model 4 | |
| Hugging on average one individual per day (compared to the lack of hugging) | −0.86*** | −0.71*** | −0.72*** | −0.59*** | −0.17*** | −0.16*** | −0.17*** | −0.10** | −9.44*** | −8.18*** | −7.54*** | −6.45*** |
| (−1.03–−0.69) | (−0.89–−0.53) | (−0.90–−0.54) | (−0.76–−0.42) | (−0.25–−0.10) | (−0.24–−0.08) | (−0.25–−0.09) | (−0.17–−0.03) | (−10.91–−7.98) | (−9.71–−6.65) | (−9.05–−6.03) | (−7.87–−5.03) | |
| Hugging on average two or three individuals per day | −1.16*** | −1.05*** | −1.08*** | −0.85*** | −0.28*** | −0.30*** | −0.31*** | −0.20*** | −14.30*** | −13.11*** | −12.40*** | −10.54*** |
| (−1.34–−0.99) | (−1.24–−0.86) | (−1.27–−0.88) | (−1.04–−0.66) | (−0.36–−0.20) | (−0.38–−0.22) | (−0.40–−0.23) | (−0.28–−0.13) | (−15.84–−12.75) | (−14.76–−11.46) | (−14.04–−10.76) | (−12.11–−8.98) | |
| Hugging on average four or more individuals per day | −1.26*** | −1.20*** | −1.26*** | −1.03*** | −0.19** | −0.22*** | −0.24*** | −0.14* | −13.61*** | −12.72*** | −12.03*** | −10.23*** |
| (−1.56–−0.96) | (−1.51–−0.90) | (−1.56–−0.96) | (−1.32–−0.74) | (−0.32–−0.06) | (−0.34–−0.10) | (−0.36–−0.12) | (−0.25–−0.02) | (−16.26–−10.96) | (−15.33–−10.10) | (−14.59–−9.47) | (−12.70–−7.75) | |
| Sociodemographic covariates | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | |||
| Lifestyle-related covariates | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ||||||
| Health-related covariates | ✓ | ✓ | ✓ | |||||||||
| R 2 | 3,270 | 3,270 | 3,270 | 3,270 | 3,270 | 3,270 | 3,270 | 3,270 | 3,270 | 3,270 | 3,270 | 3,270 |
| Observations | 0.05 | 0.10 | 0.12 | 0.23 | 0.02 | 0.12 | 0.13 | 0.30 | 0.10 | 0.17 | 0.20 | 0.31 |
*** p < 0.001, ** p < 0.01, * p < 0.05, + p < 0.10, unstandardized beta-coefficients are shown (95% CI are displayed in parentheses); sociodemographic factors encompass gender, age, size of the city, education, marital situation, labor force participation, and religious affiliation; lifestyle-related factors encompass smoking status, sports activities and alcohol consumption; health-related factors encompass probable depression, the number of chronic illnesses, and self-rated health. Model 1 is without covariates, model 2 includes sociodemographic covariates, model 3 includes sociodemographic and lifestyle-related covariates, and model 4 includes sociodemographic, lifestyle-related, and health-related covariates.
Table 3.
Association of hugging frequency with loneliness, social isolation, and social withdrawal among men based on linear regressions.
| Independent variables | Loneliness | Perceived social isolation | Social withdrawal | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Model 1 | Model 2 | Model 3 | Model 4 | Model 1 | Model 2 | Model 3 | Model 4 | Model 1 | Model 2 | Model 3 | Model 4 | |
| Hugging on average one individual per day (compared to the lack of hugging) | −0.68*** | −0.43*** | −0.45*** | −0.36** | −0.10+ | −0.03 | −0.04 | 0.01 | −7.68*** | −5.79*** | −5.18*** | −4.43*** |
| (−0.91–−0.45) | (−0.67–−0.18) | (−0.70–−0.20) | (−0.60–−0.13) | (−0.20–0.01) | (−0.14–0.08) | (−0.15–0.07) | (−0.09–0.10) | (−9.72–−5.64) | (−7.96–−3.62) | (−7.34–−3.02) | (−6.41–−2.44) | |
| Hugging on average two or three individuals per day | −0.96*** | −0.73*** | −0.74*** | −0.55*** | −0.16** | −0.13* | −0.15* | −0.05 | −12.54*** | −10.71*** | −10.02*** | −8.42*** |
| (−1.21–−0.71) | (−1.00–−0.46) | (−1.02–−0.47) | (−0.82–−0.29) | (−0.27–−0.05) | (−0.24–−0.01) | (−0.26–−0.03) | (−0.15–−0.06) | (−14.73–−10.34) | (−13.09–−8.33) | (−12.38–−7.65) | (−10.66–−6.19) | |
| Hugging on average four or more individuals per day | −0.82*** | −0.70** | −0.76*** | −0.64** | 0.03 | 0.03 | −0.00 | 0.05 | −10.24*** | −8.99*** | −8.71*** | −7.93*** |
| (−1.26–−0.38) | (−1.16–−0.25) | (−1.21–−0.31) | (−1.05–−0.22) | (−0.14–0.21) | (−0.15–−0.20) | (−0.18–−0.18) | (−0.12–−0.22) | (−14.16–−6.32) | (−12.91–−5.07) | (−12.52–−4.91) | (−11.47–−4.39) | |
| Sociodemographic covariates | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | |||
| Lifestyle-related covariates | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ||||||
| Health-related covariates | ✓ | ✓ | ✓ | |||||||||
| R 2 | 1,614 | 1,614 | 1,614 | 1,614 | 1,614 | 1,614 | 1,614 | 1,614 | 1,614 | 1,614 | 1,614 | 1,614 |
| Observations | 0.04 | 0.09 | 0.10 | 0.21 | 0.01 | 0.11 | 0.13 | 0.33 | 0.08 | 0.15 | 0.19 | 0.32 |
*** p < 0.001, ** p < 0.01, * p < 0.05, + p < 0.10, unstandardized beta-coefficients are shown (95% CI are displayed in parentheses); sociodemographic factors encompass age, size of the city, education, marital situation, labor force participation, and religious affiliation; lifestyle-related factors encompass smoking status, sports activities and alcohol consumption; health-related factors encompass probable depression, the number of chronic illnesses, and self-rated health. Model 1 is without covariates, model 2 includes sociodemographic covariates, model 3 includes sociodemographic and lifestyle-related covariates, and model 4 includes sociodemographic, lifestyle-related, and health-related covariates.
Table 4.
Association of hugging frequency with loneliness, social isolation and social withdrawal among women based on linear regressions.
| Independent variables | Loneliness | Perceived social isolation | Social withdrawal | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Model 1 | Model 2 | Model 3 | Model 4 | Model 1 | Model 2 | Model 3 | Model 4 | Model 1 | Model 2 | Model 3 | Model 4 | |
| Hugging on average one individual per day (compared to the lack of hugging) | −1.04*** | −0.99*** | −0.97*** | −0.81*** | −0.26*** | −0.30*** | −0.30*** | −0.22*** | −11.23*** | −10.82*** | −10.05*** | −8.78*** |
| (−1.29–-0.79) | (−1.25–-0.72) | (−1.24–-0.71) | (−1.06–-0.55) | (−0.37–-0.16) | (−0.41–-0.19) | (−0.41–-0.19) | (−0.32–-0.12) | (−13.35–-9.11) | (−13.01–-8.62) | (−12.21–-7.88) | (−10.84–-6.73) | |
| Hugging on average two or three individuals per day | −1.38*** | −1.38*** | −1.40*** | −1.13*** | −0.41*** | −0.48*** | −0.48*** | −0.35*** | −16.08*** | −15.82*** | −14.98*** | −13.00*** |
| (−1.63–-1.12) | (−1.65–-1.10) | (−1.67–-1.12) | (−1.40–-0.86) | (−0.52–-0.30) | (−0.59–-0.36) | (−0.59–-0.36) | (−0.46–-0.25) | (−18.28–-13.88) | (−18.14–-13.50) | (−17.29–-12.68) | (−15.21–-10.78) | |
| Hugging on average four or more individuals per day | −1.68*** | −1.70*** | −1.73*** | −1.44*** | −0.41*** | −0.47*** | −0.48*** | −0.34*** | −16.77*** | −16.75*** | −15.59*** | −13.30*** |
| (−2.09–-1.27) | (−2.11–-1.29) | (−2.13–-1.33) | (−1.84–-1.04) | (−0.59–-0.23) | (−0.64–-0.30) | (−0.65–-0.31) | (−0.50–-0.18) | (−20.32–-13.21) | (−20.21–-13.28) | (−19.04–-12.14) | (−16.76–-9.83) | |
| Sociodemographic covariates | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | |||
| Lifestyle-related covariates | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ||||||
| Health-related covariates | ✓ | ✓ | ✓ | |||||||||
| R 2 | 1,647 | 1,647 | 1,647 | 1,647 | 1,647 | 1,647 | 1,647 | 1,647 | 1,647 | 1,647 | 1,647 | 1,647 |
| Observations | 0.07 | 0.12 | 0.15 | 0.25 | 0.03 | 0.14 | 0.16 | 0.31 | 0.13 | 0.20 | 0.23 | 0.31 |
*** p < 0.001, ** p < 0.01, * p < 0.05, + p < 0.10, unstandardized beta-coefficients are shown (95% CI are displayed in parentheses); Model 1 sociodemographic factors encompass age, size of the city, education, marital situation, labor force participation, and religious affiliation; lifestyle-related factors encompass smoking status, sports activities and alcohol consumption; health-related factors encompass probable depression, the number of chronic illnesses, and self-rated health. Model 1 is without covariates, model 2 includes sociodemographic covariates, model 3 includes sociodemographic and lifestyle-related covariates, and model 4 includes sociodemographic, lifestyle-related, and health-related covariates.
Among men, hugging on average one or more individuals per day (compared to the lack of hugging) was associated with less loneliness and social withdrawal (see Table 3). Significant differences between those who hug two to three individuals on average daily and those who do not hug anyone were reduced to non-significance adjusting for health-related covariates among men.
Among women, hugging on average one or more individuals per day (compared to the lack of hugging) was associated with less loneliness, social isolation and social withdrawal (see Table 4). It is worth noting that all associations between the frequency of hugging and the three outcomes were significantly more pronounced among women, as indicated by significant interaction terms for frequency of hugging x gender. Please see Table 5 for further details.
Table 5.
Association of hugging frequency with loneliness, social isolation and social withdrawal among the total sample based on linear regressions (with interaction term: hugging x gender).
| Independent variables | Loneliness | Social isolation | Social withdrawal |
|---|---|---|---|
| Hugging on average one individual per day (compared to the lack of hugging) | −0.42*** | −0.02 | −4.69*** |
| (−0.64–−0.19) | (−0.11–0.08) | (−6.57–−2.81) | |
| Hugging on average two or three individuals per day | −0.63*** | −0.08 | −8.75*** |
| (−0.88–−0.39) | (−0.18–0.02) | (−10.84–−6.67) | |
| Hugging on average four or more individuals per day | −0.70*** | 0.02 | −7.86*** |
| (−1.10–−0.29) | (−0.14–0.18) | (−11.29–−4.42) | |
| Gender: women (compared to men) | 0.16 | 0.14** | −0.02 |
| (−0.09–0.41) | (0.04–0.25) | (−2.16–2.12) | |
| Interaction term (hugging x gender): Hugging on average one individual per day x women | −0.37* | −0.19** | −3.77** |
| (−0.69–−0.05) | (−0.31–−0.06) | (−6.40–−1.13) | |
| Hugging on average two or three individuals per day x women | −0.46** | −0.25*** | −3.75** |
| (−0.79–−0.12) | (−0.38–−0.11) | (−6.54–−0.97) | |
| Hugging on average four or more individuals per day x women | −0.69* | −0.34** | −4.84* |
| (−1.24–−0.15) | (−0.55–−0.12) | (−9.52–−0.17) | |
| Sociodemographic covariates | ✓ | ✓ | ✓ |
| Lifestyle-related covariates | ✓ | ✓ | ✓ |
| Health-related covariates | ✓ | ✓ | ✓ |
| R 2 | 3,261 | 3,261 | 3,261 |
| Observations | 0.23 | 0.31 | 0.31 |
*** p < 0.001, ** p < 0.01, * p < 0.05, + p < 0.10, unstandardized beta-coefficients are shown (95% CI are displayed in parentheses); sociodemographic factors encompass gender, age, size of the city, education, marital situation, labor force participation, and religious affiliation; lifestyle-related factors encompass smoking status, sports activities and alcohol consumption; health-related factors encompass probable depression, the number of chronic illnesses, and self-rated health. The model includes sociodemographic, lifestyle-related, and health-related covariates.
Among the total sample, the effect sizes (in terms of partial eta-squared) for the frequency of hugging were 0.03 for loneliness, 0.01 for social isolation, and 0.06 for social withdrawal, corresponding to small to medium effect sizes. Among men, the effect sizes for frequency of hugging were 0.01 for loneliness, 0.001 for social isolation, and 0.04 for social withdrawal, reflecting small effect sizes. Among women, the effect sizes for frequency of hugging were 0.05 for loneliness, 0.03 for social isolation, and 0.08 for social withdrawal, corresponding to small to medium effect sizes.
4. Discussion
Our objective was to investigate whether hugging was associated with loneliness, social isolation, and social withdrawal and whether these associations varied by gender. Our main findings were that hugging on average one or more individuals per day (compared to no hugging) was associated with feeling less loneliness, social isolation, and social withdrawal among the total sample. These associations were more pronounced among women. The associations were weaker and still apparent for loneliness and social withdrawal among men.
As there has been little research to date on the association of hugging with loneliness, social isolation and social withdrawal it is challenging to contextualize our findings in terms of previous research. As outlined in the introduction, one former study (11) indicated that hugging or kissing a romantic partner was associated with lower loneliness among American adults in the early stages of the Covid-19 pandemic. Our current study builds on this previous Covid-19 specific American study (11) and shows that the frequency of hugging is associated with loneliness and also social isolation and social withdrawal in Germany in the post-pandemic era.
The frequency of daily hugging may be particularly driven by hugs as a sign of greeting or affection rather than comfort. In this respect, the physical closeness of a hug may make individuals aware of their favorable and high-quality current relationships (e.g., to friends and acquaintances), ultimately contributing to lower levels of loneliness, social isolation and social withdrawal. Even if sad, being hugged might make individuals feel connected and included in society (12). Interestingly, being hugged on average two or three times daily seems to be sufficient (in terms of the outcomes used in this study) (27). This frequency of hugging may reflect a mix of adequate (in terms of number of social contacts) yet high-quality social relationships (11). Future research is needed to investigate this in more detail.
In accordance with our initial assumptions, the associations of interest were significantly more pronounced among women than men. This difference may be explained by the fact that hugs might be more relevant for women as an expression of physical closeness in their friendships (or other relationships) (14). Hugs may be more important to express their emotions and receive support (14). Emotional closeness (also expressed by hugs) may be particularly important for women to feel connected in their social relationships. Cultural norms and social expectations may also be of relevance. More precisely, we assume that hugging as a sign of physical closeness may be more common among women friends in Germany than among men friends. If women then hug less, for example, they could quickly feel lonely, isolated and withdrawn because their expectations are not met. Biologically, physical touches such as hugs may promote the release of oxytocin (a hormone related to stress reduction, trust, and bonding) (28, 29). For reasons of biological responses (30) and social customs, women may particularly benefit from such effects.
When interpreting our present findings, it is important to consider some key strengths and limitations of this work. The data were from a large sample that reflects the German adult population – in terms of gender, age groups, and federal states. Psychometrically sound tools were used to quantify the outcomes. However, solely subjective tools were used. Future research could focus on, for example, objective factors to quantify social withdrawal. While the tool used to quantify hugging has a high face validity, more sophisticated tools are recommended. For example, future research could distinguish between different types of hugs, such as hugs between relatives, friends or partners. It would also be interesting to examine the duration of hugs in more detail. Furthermore, the context and intimacy of hugs should be explored in future research. The observational study design has shortcomings regarding the directionality and causality of the relationship (e.g., individuals who withdraw from society may be also less likely to hug others). It is possible that people who are feeling lonely and isolated are too distressed to socialize, or have fewer people to hug, or they may even prefer greater social distance. Even if social disconnection leads to fewer hugs, it is still possible that an intervention based on hugs could reduce loneliness and social disconnection. Therefore, experimental longitudinal studies are recommended to get a better understanding of the directionality and evaluate causality. Additionally, a potential online response bias cannot be completely excluded. Nonetheless, we believe that responding to an online survey may lead to more honest answers (e.g., in terms of the actual number of hugs per day) compared to being interviewed in person (31). Another limitation is that it was not possible to adjust for social network size. Such a factor could be included in future research.
In conclusion, our findings suggested an association between hugging frequency and measures of social disconnectedness – among women in particular. Efforts to increase the frequency of hugs, ideally to at least two to three individuals on average daily may be beneficial for such social disconnectedness outcomes, pending longitudinal and experimental evidence. Since the frequency and social meaning of hugging may vary between countries (e.g., Western and Asian cultures), future cross-country comparisons may be of interest. Some individuals experienced post-pandemic stress disorder due to the COVID-19 pandemic, with resulting behavioral changes. Such a topic could also be further explored in future studies in this context by explicitly focusing on this subgroup.
Funding Statement
The author(s) declared that financial support was not received for this work and/or its publication.
Footnotes
Edited by: Nayara Mota, Catholic University of Salvador, Brazil
Reviewed by: Kiffer Card, Simon Fraser University, Canada
Roger Manktelow, Ulster University, United Kingdom
Data availability statement
The datasets presented in this article are not readily available because the data are not publicly available due to ethical restrictions. Requests to access the datasets should be directed to the corresponding author and the Department for Health Economics and Health Services Research, University Medical Center Hamburg-Eppendorf: igv@uke.de.
Ethics statement
The studies involving humans were approved by Psychological Ethics Committee of the University Medical Center Hamburg-Eppendorf (LPEK-0849). The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.
Author contributions
AH: Conceptualization, Data curation, Formal analysis, Methodology, Project administration, Visualization, Writing – original draft, Writing – review & editing. ASu: Conceptualization, Visualization, Writing – review & editing, Writing – original draft. ML: Conceptualization, Visualization, Writing – review & editing, Writing – original draft. YS: Conceptualization, Visualization, Writing – review & editing, Writing – original draft. KP: Conceptualization, Visualization, Writing – review & editing, Writing – original draft. SP: Conceptualization, Visualization, Writing – review & editing, Writing – original draft. DY: Conceptualization, Visualization, Writing – review & editing, Writing – original draft. RG: Conceptualization, Visualization, Writing – review & editing, Writing – original draft. ASt: Conceptualization, Visualization, Writing – original draft, Writing – review & editing. AT: Conceptualization, Visualization, Writing – original draft, Writing – review & editing. H-HK: Conceptualization, Resources, Supervision, Visualization, Writing – original draft, Writing – review & editing.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that Generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
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.
References
- 1.Perlman D, Peplau LA. Toward a Social Psychology of Loneliness. London: Academic Press; (1981). [Google Scholar]
- 2.Bude H, Lantermann E-D. Soziale Exklusion Und Exklusionsempfinden. Kölner Zeitschrift für Soziologie und Sozialpsychologie. (2006) 58:233–52. doi: 10.1007/s11575-006-0054-1 [DOI] [Google Scholar]
- 3.Rubin KH, Coplan RJ, Bowker JC. Social withdrawal in childhood. Annu Rev Psychol. (2009) 60:141. doi: 10.1146/annurev.psych.60.110707.163642, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Deason KG, Luchetti M, Karakose S, Stephan Y, O'Súilleabháin PS, Hajek A, et al. Neuroticism, loneliness, all-cause and cause-specific mortality: a 17-year study of nearly 500,000 individuals. J Affect Disord. (2025) 368:274–81. doi: 10.1016/j.jad.2024.09.077, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Hajek A, Kretzler B, König H-H. Multimorbidity, loneliness, and social isolation. A systematic review. Int J Environ Res Public Health. (2020) 17:8688. doi: 10.3390/ijerph17228688, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Ejiri M, Kawai H, Ishii K, Oka K, Obuchi S. Predictors of older adults’ objectively measured social isolation: a systematic review of observational studies. Arch Gerontol Geriatr. (2021) 94:104357. doi: 10.1016/j.archger.2021.104357, [DOI] [PubMed] [Google Scholar]
- 7.Hajek A, Gyasi RM, König H-H. Factors associated with loneliness among individuals aged 80 years and over: findings derived from the nationally representative" old age in Germany (D80+)" study. Arch Gerontol Geriatr. (2024) 123:105443. doi: 10.1016/j.archger.2024.105443, [DOI] [PubMed] [Google Scholar]
- 8.Hajek A, Sutin AR, Posi G, Stephan Y, Peltzer K, Terracciano A, et al. Chronic loneliness and chronic social isolation among older adults. A systematic review, Meta-analysis and Meta-regression. Aging Ment Health. (2025) 29:185–200. doi: 10.1080/13607863.2024.2385448, [DOI] [PubMed] [Google Scholar]
- 9.Mullins LC, Mushel M. The existence and emotional closeness of relationships with children, friends, and spouses: the effect on loneliness among older persons. Res Aging. (1992) 14:448–70. doi: 10.1177/0164027592144002 [DOI] [Google Scholar]
- 10.Trinh E. How hugging mom teaches me the meaning of love and perhaps beyond. J Faith Educ Community. (2018) 2:2. [Google Scholar]
- 11.Rosenberg M, Luetke M, Hensel D, Kianersi S, Fu T-c, Herbenick D. Depression and loneliness during April 2020 COVID-19 restrictions in the United States, and their associations with frequency of social and sexual connections. Soc Psychiatry Psychiatr Epidemiol. (2021) 56:1221–32. doi: 10.1007/s00127-020-02002-8, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Cabral CS, Guimarães JS, Teixeira A, Generoso NK, França I, Borges ALV. “We want to hug a friend”: the Covid-19 pandemic among low-income adolescents. Rev Saude Publica. (2023) 57:5s. doi: 10.11606/s1518-8787.2023057004778 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Murphy ML, Janicki-Deverts D, Cohen S. Receiving a hug is associated with the attenuation of negative mood that occurs on days with interpersonal conflict. PLoS One. (2018) 13:e0203522. doi: 10.1371/journal.pone.0203522, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Berretz G, Cebula C, Wortelmann BM, Papadopoulou P, Wolf OT, Ocklenburg S, et al. Romantic partner embraces reduce cortisol release after acute stress induction in women but not in men. PLoS One. (2022) 17:e0266887. doi: 10.1371/journal.pone.0266887, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Gierveld JDJ, Tilburg TV. A 6-item scale for overall, emotional, and social loneliness: confirmatory tests on survey data. Res Aging. (2006) 28:582–98. doi: 10.1177/0164027506289723 [DOI] [Google Scholar]
- 16.Hajek A, Teo AR, Zwar L, König HH. Validation of the German version of the 25-item hikikomori questionnaire 25. Int J Methods Psychiatr Res. (2024) 33:e2027. doi: 10.1002/mpr.2027, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Teo AR, Chen JI, Kubo H, Katsuki R, Sato-Kasai M, Shimokawa N, et al. Development and validation of the 25-item hikikomori questionnaire (Hq-25). Psychiatry Clin Neurosci. (2018) 72:780–8. doi: 10.1111/pcn.12691, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Hajek A, König H-H. Interest in anime and manga: relationship with (mental) health, social disconnectedness, social joy and subjective well-being. J Public Health. (2024):1–11. doi: 10.1007/s10389-024-02341-9 [DOI] [Google Scholar]
- 19.Hajek A, König H-H. Prevalence and correlates of loneliness, perceived and objective social isolation during the COVID-19 pandemic. Evidence from a representative survey in Germany. Soc Psychiatry Psychiatr Epidemiol. (2022) 57:1969–78. doi: 10.1007/s00127-022-02295-x, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Pengpid S, Peltzer K. Loneliness and associated factors among middle-aged and older adults: cross-sectional and longitudinal survey results from the Haalsi cohort in South Africa. Aging Ment Health. (2024) 28:1179–87. doi: 10.1080/13607863.2024.2345777, [DOI] [PubMed] [Google Scholar]
- 21.Ahn J, Falk EB, Kang Y. Relationships between physical activity and loneliness: a systematic review of intervention studies. Curr Res Behav Sci. (2024) 6:100141. doi: 10.1016/j.crbeha.2023.100141 [DOI] [Google Scholar]
- 22.Kristensen K, König H-H, Hajek A. The Association of Multimorbidity, loneliness, social exclusion and network size: findings from the population-based German ageing survey. BMC Public Health. (2019) 19:1383. doi: 10.1186/s12889-019-7741-x, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.UNESCO . International Standard Classification of Education. ISCED 1997. Re-edition ed. Paris: UNESCO; (2006). [Google Scholar]
- 24.Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med. (2001) 16:606–13. doi: 10.1046/j.1525-1497.2001.016009606.x, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Cohen J. Statistical Power Analysis for the Behavioral Sciences. New York: Lawrence Erlbaum Associates; (1988). [Google Scholar]
- 26.Shaw BP. Meeting assumptions in the estimation of reliability. Stata J: Promoting Commun Stat Stata. (2021) 21:1021–7. doi: 10.1177/1536867X211063407 [DOI] [Google Scholar]
- 27.Thompson A, Smith MA, McNeill A, Pollet TV. Friendships, loneliness and psychological wellbeing in older adults: a limit to the benefit of the number of friends. Ageing Soc. (2024) 44:1090–115. doi: 10.1017/S0144686X22000666 [DOI] [Google Scholar]
- 28.Gordon I, Zagoory-Sharon O, Leckman JF, Feldman R. Oxytocin, cortisol, and triadic family interactions. Physiol Behav. (2010) 101:679–84. doi: 10.1016/j.physbeh.2010.08.008, [DOI] [PubMed] [Google Scholar]
- 29.Field T. Touch for socioemotional and physical well-being: a review. Dev Rev. (2010) 30:367–83. doi: 10.1016/j.dr.2011.01.001 [DOI] [Google Scholar]
- 30.Taylor SE, Klein LC, Lewis BP, Gruenewald TL, Gurung RA, Updegraff JA. Biobehavioral responses to stress in females: tend-and-befriend, not fight-or-flight. Psychol Rev. (2000) 107:411–29. doi: 10.1037/0033-295X.107.3.411, [DOI] [PubMed] [Google Scholar]
- 31.Taddicken M. Methodeneffekte Von Web-Befragungen: Soziale Erwünschtheit Vs. Soziale Entkontextualisierung. Springer: Wiesbaden; (2009). [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets presented in this article are not readily available because the data are not publicly available due to ethical restrictions. Requests to access the datasets should be directed to the corresponding author and the Department for Health Economics and Health Services Research, University Medical Center Hamburg-Eppendorf: igv@uke.de.