Association of hugs with depression, anxiety, and suicidal ideation: findings based on data from the general adult population in Germany
- Open Access
- 25.10.2025
- Original Article
Abstract
Introduction
Depression, anxiety, and suicidal ideation pose a tremendous challenge for societies and economies (GBD 2019 Mental Disorders Collaborators 2022). All of them occur frequently (Hajek and König 2021; Hajek et al. 2022a; Kino et al. 2022) and are associated with unfavorable factors such as unsteady employment biographies, social withdrawal, loneliness, physical illnesses, and suicide risk (Campbell et al. 2022; Celano and Huffman 2011; Hajek et al. 2025, 2024; Hakulinen et al. 2019; Khan et al. 2002). Moreover, they place a high economic burden on the healthcare system (Bock et al. 2017; Buczak-Stec et al. 2022; Hohls et al. 2018). They can also be a great encumbrance for friends and families (Radfar et al. 2014; Skundberg-Kletthagen et al. 2014). Therefore, it is important to identify the factors associated with them.
Thus far, there have been numerous studies examining the factors associated with depression, anxiety, and suicidal ideation (Gastaldon et al. 2022; Pelton et al. 2021; Yuan et al. 2022). For instance, socioeconomic factors such as financial difficulties, lifestyle-related factors such as physical activity, and health-related factors such as self-rated health have all been linked to the aforementioned mental health outcomes (Hajek et al. 2022b; Östberg and Nordin 2022; Stickley et al. 2023; Wanjau et al. 2023).
Anzeige
However, as yet, far less is known about the association of hugging with depression, anxiety, and suicidal ideation. A hug occurs when one person physically embraces another individual, often as a sign of greeting, affection or comfort (Forsell and Åström 2012). When individuals are sad, a hug may reduce negative emotions (Gordon et al. 2010). Regular hugs may also strengthen bonds and can express affection (Forsell and Åström 2012). As greetings, they may signify closeness and friendship (Forsell and Åström 2012). A recent qualitative study also emphasized the need to be hugged of patients with substance use disorder after hospital discharge (Incze et al. 2025). A quantitative study (Rosenberg et al. 2021) showed that kissing or hugging either a family member or a partner almost every day (compared to not at all) were both significantly associated with a lower prevalence of probable depression in adjusted regression analyses. That study used data from a generalizable sample of adults (spanning an age range from 18 to 94 years) in the United States during the beginning of the COVID-19 pandemic (Rosenberg et al. 2021). However, quantitative studies are completely lacking on the association of hugging with a broader range of mental health outcomes (in terms of depression, anxiety and suicidal ideation) in the post-pandemic period (i.e., in a time period when there are no longer any contact restrictions). Hence, the aim of this study was to investigate the association of the frequency of hugs with probable depression, probable anxiety, and suicidal ideation in the German adult population. Such knowledge is important for better characterizing individuals at risk of depression, anxiety, and suicidal ideation. This is also relevant because those mental health outcomes may contribute to chronic illnesses and mortality (Pan et al. 2011; Walker et al. 2015).
Methods
Sample
Data were collected through an online survey (January 2025). Overall, n = 3270 individuals aged 18 to 74 years and residing in Germany took part in the survey. The only inclusion criteria were that participants had to be within the specified age range and living in Germany. The survey was administered by Bilendi, which is a reputable, International Organization for Standardization (ISO)-certified market research firm. Quota sampling was employed to ensure that the sample accurately represented the German adult population in terms of sex, age, and federal state.
All participants provided informed consent before taking part in the study. Ethical approval was granted by the Psychological Ethics Committee of the University Medical Center Hamburg-Eppendorf (number: LPEK-0849). All methods were performed in accordance with the relevant guidelines and regulations.
Dependent variables
The well-known Patient Health Questionnaire-9 (PHQ-9) (Kroenke et al. 2001) was used to quantify probable depression. A final score was generated by summing all nine items. This score varies from 0 to 27, whereby higher values denote more depressive symptoms. The established cutoff score of ≥ 10 was used to indicate probable depression (Kroenke et al. 2001). In our study, Cronbach’s alpha was 0.90 (McDonald’s omega = 0.90).
Anzeige
The widely used Generalized Anxiety Disorder-7 (GAD-7) scale (Spitzer et al. 2006) was employed to measure probable anxiety. Summing up all seven items, the final score thus ranges from 0 to 21, whereby higher values denote more anxiety symptoms. As suggested by the scale’s developers, a cutoff score of ≥ 10 was used to indicate probable anxiety (Spitzer et al. 2006). In this study, Cronbach’s alpha and McDonald’s omega were both 0.92.
The last item (item 9) of the aforementioned Patient Health Questionnaire (PHQ-9) (Kroenke et al. 2001) was used to quantify the frequency of suicidal ideation in the previous 2 weeks. The four response options for this item were as follows: “not at all,” “several days,” “more than half the days,” or “nearly every day.” As is common (Huarcaya-Victoria et al. 2025; Tassone et al. 2024), responses were dichotomized (0 = “not at all” or 1 = at least “several days”). This single item has been used extensively to quantify suicidal ideation in previous research (for an overview see Kim et al. 2021). It is also worth noting that this form of assessment predicts higher suicide risk across several groups (Kim et al. 2021; Louzon et al. 2016; Rossom et al. 2017; Simon et al. 2013).
Key independent variable: frequency of hugs
The average daily number of individuals hugged served as a measure of hugging frequency. Aligning with previous research (Hajek and König 2024), participants could indicate a range from 0 to 99 individuals. Four categories were established (see Hajek and König 2024): the first category comprised individuals who reported zero hugs (indicating they did not engage in hugging on average each day), while the second category included those who hugged on average one individual per day. The third category encompassed individuals who hugged on average two or three individuals daily, and the fourth category consisted of individuals who hugged on average four or more individuals per day.
Covariates
The selection of covariates was made based on previous research and grounded in theoretical considerations (Gastaldon et al. 2022; Hajek et al. 2022b; Yuan et al. 2022). Sociodemographic covariates included age, sex (male, female, diverse), federal state (separating the 16 federal states in Germany), marital status (single, divorced, widowed, or cohabiting—either married or in a partnership, living separately but married, or partnered), level of education (ISCED-97 framework: low, medium or high; UNESCO 2006), employment status (full-time, retired, other), and religious affiliation (no religion, Christianity, Islam, or other faiths). Lifestyle factors considered were the frequency of physical activity (no physical activity, less than 1 h per week, 1–2 h per week, 2–4 h per week, more than 4 h per week), alcohol intake (daily, several times a week, once a week, 1–3 times monthly, less often, or never), smoking history (never smoked, former smoker, occasional smoker, or daily smoker), and health-conscious diet (extent to which attention is paid to having a health-conscious diet, with four categories: not at all, a little, strongly, very strongly). Health-related covariates included self-rated health quantified by a single item (from 1 = very poor to 5 = very good), and a count of chronic conditions (based on 15 conditions such as stroke, diabetes, or heart disease).
Statistical analysis
First, the sample characteristics were calculated among the total sample and also stratified by the three mental health outcomes. The p-values for the group differences are based on independent t-tests or chi-square tests, as appropriate. Following this, both unadjusted and adjusted logistic regression analyses were undertaken. In the adjusted models, the sociodemographic variables were entered as the initial covariates. Thereafter, the lifestyle-related variables were added as additional covariates, followed by the inclusion of the health-related covariates.
In a robustness check for probable depression, the ninth item of the PHQ-9 was removed to avoid overlapping mental health outcomes, creating the PHQ-8 scale (ranging from 0 to 24, with higher scores reflecting more depressive symptoms). For the PHQ-8, a cut-point of ≥ 10 is also recommended given that suicidal thoughts are fairly uncommon in previous general adult population samples, and the effect of this item’s removal on scoring is often minor (Kroenke et al. 2009). We thus used this cut-point in our analysis. In a second robustness check, the number of children living in one’s own household (ranging from 0 to 9) and the living situation (living alone in a private household, living with a partner in a private household, living with relatives in a private household, living with friends/acquaintances in a private household, living in assisted living, living in a nursing home) were added as covariates to the main model (to take account of hugs within the household/family in particular).
To assess internal consistency, McDonald’s omega was calculated (using the “omegacoef” command in Stata 18). The significance threshold was set at p < 0.05. All analyses were carried out using Stata Now 19.5, MP-Parallel Edition (StataCorp, College Station, TX, USA).
Results
Sample characteristics
The sample characteristics are presented in Table 1 (with n = 3270 individuals). The mean age of the sample was 47 years (SD = 15.3 years, varying from 18 to 74 years), and slightly more than 50% were female.
Table 1
Characteristics of the sample (total sample and also stratified by the outcomes, n = 3270)
Variables | Total | Absence of probable depression | Presence of probable depression | p-value | Absence of probable anxiety | Presence of probable anxiety | p-value | Absence of suicidal ideation | Presence of suicidal ideation | |
|---|---|---|---|---|---|---|---|---|---|---|
n (%) | 3270 (100.0) | 2343 (71.7) | 927 (28.3) | 2553 (78.1) | 717 (21.9) | 2458 (75.2) | 812 (24.8) | |||
Sex: N (%) | 0.01 | 0.07 | < 0.01 | |||||||
Men | 1614 (49.4) | 1195 (51.0) | 419 (45.2) | 1286 (50.4) | 328 (45.7) | 1174 (47.8) | 440 (54.2) | |||
Women | 1647 (50.4) | 1142 (48.7) | 505 (54.5) | 1261 (49.4) | 386 (53.8) | 1278 (52.0) | 369 (45.4) | |||
Diverse | 9 (0.3) | 6 (0.3) | 3 (0.3) | 6 (0.2) | 3 (0.4) | 6 (0.2) | 3 (0.4) | |||
Age: Mean (SD) | 47.0 (15.3) | 49.6 (14.8) | 40.3 (14.4) | < 0.001 | 49.0 (14.9) | 39.7 (14.2) | < 0.001 | 49.1 (14.9) | 40.4 (14.6) | < 0.001 |
Marital situation: N (%) | < 0.001 | < 0.001 | < 0.001 | |||||||
Single | 908 (27.8) | 579 (24.7) | 329 (35.5) | 656 (25.7) | 252 (35.1) | 623 (25.3) | 285 (35.1) | |||
Divorced | 277 (8.5) | 208 (8.9) | 69 (7.4) | 227 (8.9) | 50 (7.0) | 213 (8.7) | 64 (7.9) | |||
Widowed | 103 (3.1) | 76 (3.2) | 27 (2.9) | 89 (3.5) | 14 (2.0) | 84 (3.4) | 19 (2.3) | |||
Married, cohabiting with spouse | 1867 (57.1) | 1406 (60.0) | 461 (49.7) | 1500 (58.8) | 367 (51.2) | 1451 (59.0) | 416 (51.2) | |||
Married, not cohabiting with spouse | 115 (3.5) | 74 (3.2) | 41 (4.4) | 81 (3.2) | 34 (4.7) | 87 (3.5) | 28 (3.4) | |||
Educational level: N (%) | < 0.001 | < 0.001 | < 0.001 | |||||||
Low | 337 (10.3) | 205 (8.7) | 132 (14.2) | 238 (9.3) | 99 (13.8) | 206 (8.4) | 131 (16.1) | |||
Medium | 1552 (47.5) | 1115 (47.6) | 437 (47.1) | 1205 (47.2) | 347 (48.4) | 1187 (48.3) | 365 (45.0) | |||
High | 1381 (42.2) | 1023 (43.7) | 358 (38.6) | 1110 (43.5) | 271 (37.8) | 1065 (43.3) | 316 (38.9) | |||
Employment status: N (%) | < 0.001 | < 0.001 | < 0.001 | |||||||
Full-time employed | 1629 (49.8) | 1184 (50.5) | 445 (48.0) | 1269 (49.7) | 360 (50.2) | 1213 (49.3) | 416 (51.2) | |||
Retired | 649 (19.8) | 522 (22.3) | 127 (13.7) | 566 (22.2) | 83 (11.6) | 543 (22.1) | 106 (13.1) | |||
Other | 992 (30.3) | 637 (27.2) | 355 (38.3) | 718 (28.1) | 274 (38.2) | 702 (28.6) | 290 (35.7) | |||
Religious affiliation: N (%) | < 0.001 | < 0.001 | < 0.001 | |||||||
No religious affiliation | 1459 (44.6) | 1093 (46.6) | 366 (39.5) | 1188 (46.5) | 271 (37.8) | 1148 (46.7) | 311 (38.3) | |||
Christianity | 1659 (50.7) | 1177 (50.2) | 482 (52.0) | 1276 (50.0) | 383 (53.4) | 1223 (49.8) | 436 (53.7) | |||
Islam | 90 (2.8) | 41 (1.7) | 49 (5.3) | 51 (2.0) | 39 (5.4) | 51 (2.1) | 39 (4.8) | |||
Other | 62 (1.9) | 32 (1.4) | 30 (3.2) | 38 (1.5) | 24 (3.3) | 36 (1.5) | 26 (3.2) | |||
Smoking status: N (%) | < 0.001 | < 0.001 | < 0.001 | |||||||
Yes, daily | 737 (22.5) | 517 (22.1) | 220 (23.7) | 556 (21.8) | 181 (25.2) | 536 (21.8) | 201 (24.8) | |||
Yes, sometimes | 347 (10.6) | 187 (8.0) | 160 (17.3) | 230 (9.0) | 117 (16.3) | 189 (7.7) | 158 (19.5) | |||
No, not anymore | 891 (27.2) | 677 (28.9) | 214 (23.1) | 734 (28.8) | 157 (21.9) | 700 (28.5) | 191 (23.5) | |||
Never smoker | 1295 (39.6) | 962 (41.1) | 333 (35.9) | 1033 (40.5) | 262 (36.5) | 1033 (42.0) | 262 (32.3) | |||
Alcohol consumption: N (%) | 0.68 | 0.04 | < 0.001 | |||||||
Daily | 185 (5.7) | 133 (5.7) | 52 (5.6) | 139 (5.4) | 46 (6.4) | 127 (5.2) | 58 (7.1) | |||
Several times a week | 560 (17.1) | 394 (16.8) | 166 (17.9) | 423 (16.6) | 137 (19.1) | 383 (15.6) | 177 (21.8) | |||
Once a week | 578 (17.7) | 409 (17.5) | 169 (18.2) | 450 (17.6) | 128 (17.9) | 426 (17.3) | 152 (18.7) | |||
1–3 times a month | 558 (17.1) | 416 (17.8) | 142 (15.3) | 462 (18.1) | 96 (13.4) | 446 (18.1) | 112 (13.8) | |||
Less often | 782 (23.9) | 560 (23.9) | 222 (23.9) | 616 (24.1) | 166 (23.2) | 613 (24.9) | 169 (20.8) | |||
Never | 607 (18.6) | 431 (18.4) | 176 (19.0) | 463 (18.1) | 144 (20.1) | 463 (18.8) | 144 (17.7) | |||
Physical activity: N (%) | < 0.001 | < 0.01 | 0.16 | |||||||
No physical activity | 843 (25.8) | 603 (25.7) | 240 (25.9) | 652 (25.5) | 191 (26.6) | 640 (26.0) | 203 (25.0) | |||
Less than 1 h per week | 605 (18.5) | 397 (16.9) | 208 (22.4) | 451 (17.7) | 154 (21.5) | 445 (18.1) | 160 (19.7) | |||
Regularly, 1–2 h per week | 813 (24.9) | 574 (24.5) | 239 (25.8) | 624 (24.4) | 189 (26.4) | 594 (24.2) | 219 (27.0) | |||
Regularly, 2–4 h per week | 550 (16.8) | 415 (17.7) | 135 (14.6) | 449 (17.6) | 101 (14.1) | 417 (17.0) | 133 (16.4) | |||
Regularly, more than 4 h per week | 459 (14.0) | 354 (15.1) | 105 (11.3) | 377 (14.8) | 82 (11.4) | 362 (14.7) | 97 (11.9) | |||
Health-conscious diet: N (%) | 0.02 | < 0.01 | < 0.01 | |||||||
Not at all | 351 (10.7) | 239 (10.2) | 112 (12.1) | 255 (10.0) | 96 (13.4) | 240 (9.8) | 111 (13.7) | |||
A little | 1172 (35.8) | 854 (36.4) | 318 (34.3) | 931 (36.5) | 241 (33.6) | 884 (36.0) | 288 (35.5) | |||
Strongly | 1505 (46.0) | 1094 (46.7) | 411 (44.3) | 1191 (46.7) | 314 (43.8) | 1162 (47.3) | 343 (42.2) | |||
Very strongly | 242 (7.4) | 156 (6.7) | 86 (9.3) | 176 (6.9) | 66 (9.2) | 172 (7.0) | 70 (8.6) | |||
Self-rated health: Mean (SD) | 3.6 (0.8) | 3.7 (0.7) | 3.2 (0.9) | < 0.001 | 3.7 (0.8) | 3.2 (1.0) | < 0.001 | 3.6 (0.8) | 3.3 (0.9) | < 0.001 |
Number of chronic conditions: Mean (SD) | 1.7 (1.8) | 1.5 (1.6) | 2.3 (2.1) | < 0.001 | 1.6 (1.7) | 2.3 (2.1) | < 0.001 | 1.6 (1.7) | 2.1 (2.0) | < 0.001 |
Hugging others (average per day): N (%) | < 0.01 | 0.06 | < 0.01 | |||||||
0 | 814 (24.9) | 541 (23.1) | 273 (29.4) | 612 (24.0) | 202 (28.2) | 581 (23.6) | 233 (28.7) | |||
1 | 1251 (38.3) | 915 (39.1) | 336 (36.2) | 994 (38.9) | 257 (35.8) | 972 (39.5) | 279 (34.4) | |||
2–3 | 998 (30.5) | 739 (31.5) | 259 (27.9) | 792 (31.0) | 206 (28.7) | 763 (31.0) | 235 (28.9) | |||
4+ | 207 (6.3) | 148 (6.3) | 59 (6.4) | 155 (6.1) | 52 (7.3) | 142 (5.8) | 65 (8.0) |
Anzeige
Overall, 24.9% of respondents reported no daily hugs, 38.3% hugged one individual, while 30.5% hugged two or three individuals per day. Moreover, 6.3% hugged four or more people each day. In total, 28.3% of the respondents had probable depression, 21.9% had probable anxiety, and 24.8% reported having suicidal thoughts for at least several days in the past 2 weeks. In bivariate analyses there was a significant association of hugging others with probable depression (Cramer’s V = 0.07, p < 0.01) and suicidal ideation (Cramer’s V = 0.07, p < 0.01). The association between hugging others and probable anxiety was of borderline statistical significance: Cramer’s V = 0.05, p = 0.056). Additional details are shown in Table 1.
Regression analysis
Findings from the unadjusted and adjusted regression analyses are presented in Table 2. Hereafter, we mostly focus on describing the fully-adjusted models (i.e., adjusting for sociodemographic, lifestyle-related, and health-related covariates). However, it is worth noting that adding covariates to the regression models attenuated the associations of interest only slightly.
Table 2
Link of hugging frequency with probable depression, probable anxiety, and suicidal ideation. Results based on logistic regression analyses
Independent variables | Probable depression | Probable anxiety | Suicidal ideation | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
Hugging others (average per day): 1 (Reference category: 0) | 0.73** | 0.67*** | 0.65*** | 0.65*** | 0.78* | 0.71** | 0.71** | 0.73* | 0.72** | 0.69** | 0.65*** | 0.66** |
(0.60–0.88) | (0.54–0.83) | (0.52–0.81) | (0.51–0.83) | (0.63–0.97) | (0.56–0.90) | (0.56–0.90) | (0.56–0.95) | (0.58–0.88) | (0.55–0.87) | (0.51–0.82) | (0.52–0.85) | |
2–3 | 0.69*** | 0.57*** | 0.53*** | 0.61*** | 0.79* | 0.63*** | 0.60*** | 0.70* | 0.77* | 0.65*** | 0.56*** | 0.63*** |
(0.57–0.85) | (0.45–0.73) | (0.41–0.68) | (0.47–0.80) | (0.63–0.98) | (0.49–0.82) | (0.46–0.78) | (0.52–0.93) | (0.62–0.95) | (0.51–0.84) | (0.43–0.72) | (0.48–0.83) | |
4+ | 0.79 | 0.62* | 0.55** | 0.58* | 1.02 | 0.77 | 0.71+ | 0.78 | 1.14 | 0.97 | 0.77 | 0.85 |
(0.57–1.10) | (0.43–0.90) | (0.37–0.81) | (0.38–0.89) | (0.71–1.45) | (0.52–1.14) | (0.47–1.06) | (0.51–1.20) | (0.82–1.59) | (0.67–1.40) | (0.52–1.13) | (0.57–1.28) | |
Sociodemographic covariates | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | |||
Lifestyle-related covariates | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ||||||
Health-related covariates | ✓ | ✓ | ✓ | |||||||||
R2 | 3270 | 3270 | 3270 | 3270 | 3270 | 3270 | 3270 | 3270 | 3270 | 3270 | 3270 | 3270 |
Observations | 0.004 | 0.10 | 0.12 | 0.24 | 0.002 | 0.09 | 0.11 | 0.19 | 0.004 | 0.08 | 0.11 | 0.17 |
Hugging others (on average one to three persons; compared to individuals not hugging anyone) on a daily basis was associated with lower odds of probable depression (e.g., hugging one individual: OR 0.65; 95% CI, 0.51 to 0.83), probable anxiety (e.g., hugging one individual: OR: 0.73; 95% CI, 0.56 to 0.95), and suicidal ideation (e.g., hugging one individual: OR 0.66; 95% CI, 0.52 to 0.85). It is of note that the largest significant differences were found between individuals who hugged two to three individuals on average daily and those who did not hug anyone. It is also worth noting that hugging on average 4 or more individuals daily (compared to individuals not hugging others) was significantly associated with lower odds of probable depression (OR 0.58; 95% CI, 0.38 to 0.89), but not with lower odds of probable anxiety or suicidal ideation.
In a robustness check (described in detail in the statistical analysis section), the PHQ-8 was used rather than the PHQ-9 to quantify probable depression (to avoid an overlap in the mental health outcomes). The results of this analysis are presented in Table 3. The results remained nearly the same as those from the main analysis shown in Table 2. In a second robustness check, the main model was extended by adding the covariates on the number of children in the household and living arrangements (see Table 4). The findings remained virtually the same compared to the main model.
Table 3
Link of hugging frequency with probable depression (based on the PHQ-8 rather than the PHQ-9). Results based on logistic regression analyses
Independent variables | Probable depression | |||
|---|---|---|---|---|
Hugging others (average per day): 1 (Reference category: 0) | 0.71*** | 0.64*** | 0.62*** | 0.62*** |
(0.58–0.86) | (0.51–0.80) | (0.50–0.78) | (0.48–0.79) | |
2–3 | 0.69*** | 0.56*** | 0.53*** | 0.61*** |
(0.56–0.85) | (0.44–0.72) | (0.41–0.68) | (0.46–0.80) | |
4+ | 0.74+ | 0.57** | 0.52** | 0.55** |
(0.53–1.05) | (0.39–0.83) | (0.35–0.78) | (0.35–0.84) | |
Sociodemographic covariates | ✓ | ✓ | ✓ | |
Lifestyle-related covariates | ✓ | ✓ | ||
Health-related covariates | ✓ | |||
R2 | 3270 | 3270 | 3270 | 3270 |
Observations | 0.004 | 0.010 | 0.12 | 0.24 |
Table 4
Link of hugging frequency with probable depression, probable anxiety and suicidal ideation. Results based on logistic regression analyses (with additional covariates: number of children in the household, and living arrangement)
Independent variables | Probable depression | Probable anxiety | Suicidal ideation | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
Hugging others (average per day): 1 (Reference category: 0) | 0.73** | 0.68*** | 0.66*** | 0.65*** | 0.78* | 0.71** | 0.71** | 0.72* | 0.72** | 0.69** | 0.64*** | 0.65*** |
(0.60–0.88) | (0.54–0.85) | (0.52–0.83) | (0.50–0.84) | (0.63–0.97) | (0.56–0.91) | (0.55–0.91) | (0.55–0.94) | (0.58–0.88) | (0.55–0.87) | (0.50–0.81) | (0.51–0.83) | |
2–3 | 0.69*** | 0.55*** | 0.51*** | 0.59*** | 0.79* | 0.60*** | 0.57*** | 0.67** | 0.77* | 0.65*** | 0.55*** | 0.62*** |
(0.57–0.85) | (0.43–0.71) | (0.39–0.66) | (0.44–0.77) | (0.63–0.98) | (0.46–0.79) | (0.44–0.76) | (0.50–0.89) | (0.62–0.95) | (0.50–0.83) | (0.42–0.72) | (0.47–0.82) | |
4+ | 0.79 | 0.58** | 0.52** | 0.55** | 1.02 | 0.73 | 0.67+ | 0.75 | 1.14 | 0.96 | 0.77 | 0.85 |
(0.57–1.10) | (0.39–0.85) | (0.35–0.77) | (0.35–0.85) | (0.71–1.45) | (0.49–1.09) | (0.44–1.02) | (0.48–1.16) | (0.82–1.59) | (0.66–1.40) | (0.52–1.14) | (0.56–1.29) | |
Sociodemographic covariates | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | |||
Lifestyle-related covariates | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ||||||
Health-related covariates | ✓ | ✓ | ✓ | |||||||||
R2 | 3270 | 3270 | 3270 | 3270 | 3270 | 3270 | 3270 | 3270 | 3270 | 3270 | 3270 | 3270 |
Observations | 0.004 | 0.10 | 0.12 | 0.24 | 0.002 | 0.09 | 0.11 | 0.20 | 0.004 | 0.09 | 0.12 | 0.17 |
Anzeige
Discussion
The aim of the present study was to examine the association of hugging with probable depression, probable anxiety, and suicidal ideation based on a large, representative sample (in terms of sex, age group, federal state) of the adult population in Germany. We found that hugging others daily—ideally one to three individuals—is associated with reduced odds of probable depression, anxiety, and suicidal thoughts. The strongest differences emerged between those who hug two to three individuals daily and those who do not hug anyone. Hugging four or more individuals daily was significantly associated with lower odds of probable depression, but not with lower odds of probable anxiety or suicidal ideation.
Until now there has been an almost total absence of research on the association of hugging with the aforementioned mental health outcomes, making it difficult to compare our findings with former studies. In fact, we could locate only one previous quantitative study (Rosenberg et al. 2021) that examined the association between kissing or hugging either a family member or partner almost every day (in comparison to not at all) with mental health, which found that such behaviors were associated with a lower prevalence of probable depression based on data from adults in the United States at the beginning of the COVID-19 pandemic. The current study therefore significantly advances understanding in this research field by showing that hugging is not merely linked to probable depression, but also to probable anxiety and suicidal ideation in Germany in the post-pandemic era.
In general, it is probable that the daily frequency of hugs is more likely to be driven by gestures of greeting or affection rather than for comforting purposes. Therefore, we propose that the physical intimacy of a hug may strengthen individuals' awareness of their positive and meaningful social bonds—such as those with friends, romantic partners or family members (Forsell and Åström 2012). This may have a direct and positive impact on mental health (Tunçgenç et al. 2023). Even when individuals experience negative emotions (such as anger or sadness), hugs may help them feel better (Murphy et al. 2018).
Interestingly, an average of two to three hugs per day was sufficient to produce the most benefits in the present study. We believe that this level of hugging may be a marker of the quality of one’s respective relationships. More precisely, we assume that the quality of the hugs in particular, could be important (e.g. valuable hugs with friends, family members or one’s partner) for mental health outcomes (see Berry and Worthington 2001). However, this is a speculative explanation that requires further research (e.g., based on qualitative studies).
Anzeige
Also remarkably, hugging four or more individuals on average per day was significantly associated with lower odds of probable depression, but not with lower odds of probable anxiety or suicidal ideation. We assume that frequent hugs may alleviate depression via social support, security, and the release of oxytocin (Dreisoerner et al. 2021; Field 2010). However, it is also possible that frequent hugs are not necessarily an expression of deep friendships and connections. Thus, individuals having many daily hugs might still feel little depth and serious support which may explain the missing link to probable anxiety and suicidal ideation. This explanation also fits with the result that a lower frequency of hugs (one to three on average daily) was associated with anxiety and suicidal thoughts in our study. It is worth emphasizing that this is only an initial, speculative explanation. Other explanations may refer to the professional environment. Certain jobs may involve frequent hugging (e.g. kindergarten teachers, social workers). Further research in this area is clearly needed. The context of hugging should be further explored. Another potential explanation refers to the limited statistical power.
Some important strengths and limitations are worth bearing in mind when interpreting the present findings. Importantly, data were taken from a quota-based sample (representative in terms of sex, age group, and federal state). Moreover, established screening tools (the PHQ-9 [and PHQ-8 in a robustness check] and GAD-7) were used to quantify probable depression and probable anxiety, respectively. The usefulness of these screening tools has been repeatedly demonstrated (Plummer et al. 2016; Shin et al. 2019; Wu et al. 2020). Further, the item used to measure suicidal ideation has clear face validity. However, future research with multi-item tools is recommended to gain additional insights into the association between the frequency of hugs and suicidal ideation (Na et al. 2018). As the present study was based on cross-sectional data it has inherent limitations regarding the directionality of the observed relationships (e.g., certain depressive symptoms such as anhedonia or anxiety may contribute to a lower likelihood of hugging). Future research based on longitudinal data is therefore recommended. It is also important to note that using data from an online survey may exclude a certain proportion of individuals (e.g., those without access to the internet or older individuals not familiar with it). However, the proportion of this group of people in the general German adult population is extremely low (Statistisches Bundesamt 2025). Moreover, individuals may give more honest responses in an online-survey (e.g., compared to being interviewed face-to-face) (Casler et al. 2013).
The findings of this study indicated an association between the frequency of hugging and mental health (including suicidal ideation). The daily hugging of one to three individuals may assist in preventing poor mental health. However, future longitudinal evidence is needed to confirm this. Studies from other countries are also recommended. Moreover, studies focusing on other forms of hugging (e.g., virtual hugging) are also warranted (Wang et al. 2025).
Declarations
Ethics approval
The Psychological Ethics Committee of the University Medical Center Hamburg-Eppendorf gave its approval for this study (LPEK-0849). The study was conducted in accordance with the Declaration of Helsinki.
Anzeige
Consent to participate
Each individual gave his or her informed consent.
Consent for publication
Not applicable.
Conflicts of interest
The authors have no competing interests to declare that are relevant to the content of this article.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.