Demographic and call characteristics & prevalence of suicidal ideation

Tables 1 and 2 summarize the demographic information and call characteristics of 128,245 callers to Hangzhou’s mental health hotline, exploring the association between these features and suicidal ideation. The data reveals that males constitute 45.9% (n = 58,878) of all callers, while females make up 54.1% (n = 69,367). In terms of gender, 9.4% (n = 5,511) of males and 9.8% (n = 6,774) of females reported suicidal ideation, with this difference reaching statistical significance (P = 0.014). Age distribution shows that the highest proportion is in the 20-29 age group (26.7%, n = 34,183), followed by 30-39 (18.4%, n = 23,598), where the 20-29 and 10-19 age groups have significantly higher numbers reporting suicidal ideation (P < 0.001). Among those who called, 46.7% (n = 59,913) were single, with this group having the highest rate of suicidal ideation at 14.3% (n = 8,594), which was also statistically significant (P < 0.001). In terms of education background, 28.3% (n = 36,306) of callers had a bachelor’s degree or above, indicating higher risk for suicidal ideation in those with lower educational attainment (P < 0.001). In the occupational distribution, the white-collar class accounted for 40.2% (n = 51,527), with 13.5% (n = 6,963) expressing suicidal ideation, while the highest proportion of unemployed callers (20.9%, n = 2,569) showed a significant phenomenon (P < 0.001). Analysis of call times indicates that calls were evenly distributed throughout the day and night (48.8% vs. 51.2%, n = 62,593 vs. n = 65,652, p = 0.074). Regarding call duration, over 62.4% (n = 80,045) of calls lasted between 1-15 minutes, with a higher rate of suicidal ideation observed as call duration increased (p < 0.001). Geographically, 35.2% (n = 45,161) of calls originated from Hangzhou, where the suicidal ideation rate was 5.7%. Callers from Zhejiang province had a suicidal ideation rate of 10.3% (n = 3,322), while callers from other provinces had the highest rate of 12.5% (n = 6,389), with statistically significant differences (p < 0.001). Most calls (87.3%, n = 112,018) were made for personal needs, with those calling for their own needs showing more pronounced signs of suicidal ideation compared to those calling for others (P < 0.001). In terms of emotional state, 95.4% (n = 122,329) of callers reported normal emotions, but moderate to severe depression was closely linked to suicidal ideation (P < 0.001). Concerning suicide behavior, 96.9% (n = 124,252) of callers had no record of suicide attempts, with a significantly higher risk of suicidal ideation among those who had attempted suicide (P < 0.001).The analysis during the pandemic reveals that the rate of suicidal ideation was highest during the mid-pandemic phase (11.95%, n = 5,421), compared to the pre-pandemic (7.68%, n = 5,093) and post-pandemic phases (10.66%, n = 1,771), with statistically significant differences (p < 0.001).Among the consultation themes, mental health issues were the most prevalent (45.1 %, n = 57,872), followed by relationship problems in romantic partnerships (6.5 %, n = 8,312) and marriage and family issues (9.0 %, n = 11,593). Notably, there is a particularly strong association between mental health issues and suicidal ideation (P < 0.001).

Table 1 Demographic and call characteristics of callers to the Hangzhou Mental Health Hotline(Part 1) Table 2 Demographic and call characteristics of callers to the Hangzhou Mental Health Hotline(Part 2)Overall analysis of the association between consultation themes and suicidal ideation

Table 3 presents the results of logistic regression analyses examining the association between consultation themes and suicidal ideation, with “Other issues” as the reference category. The analysis includes multiple models, progressively adjusting for key covariates: Model 1 is unadjusted, providing the raw association between consultation themes and suicidal ideation; Model 2 adjusts for demographic factors, including sex, age, marital status, educational background, and occupation; Model 3 further includes call-related variables such as call duration, region, quarter, call time, as well as clinical factors like mood status and suicide behavior;Model 4 includes adjustments for the pandemic phase, accounting for the potential variation in the association between consultation themes and suicidal ideation during different phases of the COVID-19 pandemic.

Table 3 Logistic regression analysis of the association between consultation themes and suicidal ideation across multiple models

Physical illness consistently showed the strongest association with suicidal ideation across all models. In the unadjusted model (Model 1), the odds ratio (OR) was 9.11 (95% CI: 7.54, 10.9), which decreased to 4.70 (95% CI: 3.71, 5.91) in the fully adjusted model (Model 4). This reduction suggests that, even after accounting for other factors, physical illness remains strongly associated with suicidal ideation. Mental health issues also showed a strong association with suicidal ideation, with an OR of 7.77 (95% CI: 7.20, 8.40) in Model 1, decreasing to 3.35 (95% CI: 3.03, 3.70) after full adjustment in Model 4. The attenuation in ORs after adjusting for covariates suggests that while demographic and clinical factors contribute to this association, mental health issues remain an important factor related to suicidal ideation.

Romantic relationship issues were significantly linked to suicidal ideation, with an OR of 6.90 (95% CI: 6.26, 7.60) in the unadjusted model, which reduced to 2.43 (95% CI: 2.14, 2.76) in Model 4. The decrease across models suggests that the association is moderated by other factors, such as the individual’s emotional state and the context of the call. Marital and family issues had an initial OR of 4.88 (95% CI: 4.44, 5.37) in Model 1, which decreased to 2.15 (95% CI: 1.90, 2.43) in Model 4. This finding suggests that while marital and family issues are relevant, their association with suicidal ideation is influenced by various other factors, including socio-demographic and situational variables.

Sexual issues showed a relatively weak and protective association with suicidal ideation, with the OR reducing from 0.28 (95% CI: 0.16, 0.45) in Model 1 to 0.20 (95% CI: 0.11, 0.33) in Model 4. This pattern suggests that discussing sexual issues might be associated with lower risk, potentially reflecting differences in the type of distress leading to suicidal ideation. Parenting and child education also demonstrated a protective effect, with the OR reducing from 0.70 (95% CI: 0.51, 0.94) in the unadjusted model to 0.65 (95% CI: 0.46, 0.89) in the fully adjusted model, indicating that these concerns might buffer against suicidal ideation when controlling for other factors. Work-related issues had an initial OR of 2.98 (95% CI: 2.56, 3.47), which decreased to 1.56 (95% CI: 1.29, 1.87) after full adjustment. This suggests that the stress associated with work issues is significantly associated with suicidal ideation, though this relationship is moderated by other variables.

Interpersonal relationship issues showed a modest association with suicidal ideation, starting with an OR of 1.31 (95% CI: 1.09, 1.57) in Model 1, which decreased to 0.68 (95% CI: 0.54, 0.83) in Model 4. The reduction indicates that when other factors are considered, interpersonal issues alone are less predictive of suicidal ideation. Learning issues had a strong association with suicidal ideation in the unadjusted model (OR = 4.28, 95% CI: 3.61, 5.05), which remained significant even after full adjustment in Model 4, with an OR of 1.36 (95% CI: 1.10, 1.67). This suggests that learning-related stressors are independently linked to suicidal ideation, highlighting the need for targeted interventions.

Gender-stratified analysis

Table 4 presents a gender-stratified logistic regression analysis of the association between various consultation themes and suicidal ideation among callers to the Hangzhou Mental Health Hotline. For physical illness, the odds ratio (OR) for suicidal ideation was significantly higher in females (OR = 6.86, 95% CI: 4.58, 10.1) compared to males (OR = 3.71, 95% CI: 2.75, 4.94), with a Z-score of -2.45 and a P-value of 0.014, indicating a statistically significant gender difference. This suggests that physical illness is more strongly associated with suicidal ideation among female callers. Mental health issues showed a strong association with suicidal ideation in both genders, but the effect was more pronounced in females (OR = 6.81, 95% CI: 5.75, 8.09) than in males (OR = 2.11, 95% CI: 1.86, 2.40), with a highly significant Z-score of -10.78 (P < 0.001), highlighting that Mental health issues are a critical concern for both genders, with a stronger association observed in females. Romantic relationship issues were associated with an increased risk of suicidal ideation in both males (OR = 1.88, 95% CI: 1.59, 2.23) and females (OR = 4.13, 95% CI: 3.36, 5.07), with a Z-score of -5.79 (P < 0.001) indicating a significant gender difference, with females being more affected by these issues than males. Marital and family issues were significantly associated with suicidal ideation, with a stronger association in females (OR = 4.58, 95% CI: 3.78, 5.57) compared to males (OR = 1.21, 95% CI: 1.01, 1.46), as indicated by a Z-score of -9.76 (P < 0.001).

Table 4 Stratified logistic regression analysis of the association between consultation themes and suicidal ideation by gender

Sexual issues showed a protective effect against suicidal ideation in males (OR = 0.16, 95% CI: 0.09, 0.28), but this effect was not significant in females (OR = 0.12, 95% CI: 0.01, 0.94), with the Z-score of 0.24 (P = 0.809) indicating no significant gender difference. Parenting and child education concerns were associated with a lower likelihood of suicidal ideation, particularly among males (OR = 0.27, 95% CI: 0.10, 0.52), compared to females (OR = 1.52, 95% CI: 1.02, 2.21), with a Z-score of -4.27 (P < 0.001), indicating significant gender differences in how these issues are related to suicidal ideation. Work-related issues were found to increase the risk of suicidal ideation, more so in females (OR = 2.56, 95% CI: 1.89, 3.43), while the association was not significant in males (OR = 1.15, 95% CI: 0.90, 1.46), with a significant Z-score of -4.09 (P < 0.001). Interpersonal relationship issues had a stronger protective effect against suicidal ideation in males (OR = 0.32, 95% CI: 0.22, 0.46) compared to females (OR = 1.54, 95% CI: 1.16, 2.03), as indicated by a Z-score of -6.65 (P < 0.001).Learning issues were significantly associated with suicidal ideation in females (OR = 2.45, 95% CI: 1.84, 3.24), but not in males (OR = 1.02, 95% CI: 0.72, 1.41), with a Z-score of -3.91 (P < 0.001), indicating a significant gender difference in the association between learning issues and suicidal ideation.

Pandemic phase-stratified analysisAssociation between consultation themes and suicidal ideation across different pandemic phases (male callers)

The analysis of male callers reveals notable changes in the association between various consultation themes and suicidal ideation across the three pandemic phases: pre-pandemic, mid-pandemic, and post-pandemic (Tables 5 and 6). Between the pre-pandemic and mid-pandemic phases, there was a significant intensification in the association between several consultation themes and suicidal ideation among male callers. Mental health issues saw a dramatic increase in their association with suicidal ideation, with the odds ratio (OR) rising from a non-significant 1.07 (95% CI: 0.90, 1.28) pre-pandemic to a significant 4.66 (95% CI: 3.79, 5.76) mid-pandemic, confirmed by a highly significant Z-test (Z = -10.54, P < 0.001). Romantic relationship issues became significantly associated with suicidal ideation mid-pandemic, with the OR increasing from a non-significant 1.03 (95% CI: 0.79, 1.33) pre-pandemic to 3.98 (95% CI: 3.04, 5.21) mid-pandemic (Z = -7.07, P < 0.001). Marital and family issues also showed a significant increase, with the OR rising from a significant 0.51 (95% CI: 0.39, 0.67) pre-pandemic to 2.79 (95% CI: 2.08, 3.74) mid-pandemic (Z = -8.34, P < 0.001). Work-related issues saw their OR increase from a non-significant 0.78 (95% CI: 0.54, 1.12) pre-pandemic to 1.89 (95% CI: 1.27, 2.77) mid-pandemic, with a significant change confirmed by the Z-test (Z = -3.25, P = 0.001). Interpersonal relationship issues increased significantly from a protective OR of 0.10 (95% CI: 0.05, 0.15) pre-pandemic to 0.86 (95% CI: 0.48, 1.44) mid-pandemic, with the Z-test showing a significant reduction in this protective effect (Z = -4.99, P < 0.001). Learning issues also became significantly associated with suicidal ideation mid-pandemic, with the OR rising from a protective 0.55 (95% CI: 0.31, 0.94) pre-pandemic to 1.96 (95% CI: 1.16, 3.17) mid-pandemic (Z = -3.33, P < 0.001). These results suggest that the mid-pandemic period intensified the association between these themes and suicidal ideation among male callers, as stressors related to mental health, relationships, work, and education became more pronounced.

Table 5 Stratified logistic regression analysis of the association between consultation themes and suicidal ideation among males across different pandemic phases Table 6 Z-Test Comparison of odds ratios for consultation themes associated with suicidal ideation among males across different pandemic phases

In the transition from mid-pandemic to post-pandemic, there was a general trend of decreasing associations between certain consultation themes and suicidal ideation. Mental health issues saw their OR decrease from 4.66 (95% CI: 3.79, 5.76) mid-pandemic to 2.77 (95% CI: 1.83, 4.28) post-pandemic, with the Z-test confirming a significant decrease (Z = 2.15, P = 0.031). This indicates some recovery or adaptation as the pandemic progressed, though the risk remained elevated compared to pre-pandemic levels. Romantic relationship issues also showed a decrease in their OR from 3.98 mid-pandemic to 2.46 post-pandemic, but this change was not statistically significant. Marital and family issues saw a decrease from 2.79 mid-pandemic to 2.10 post-pandemic, with no significant change indicated by the Z-test. Learning issues decreased in significance, with the OR falling from 1.96 mid-pandemic to 1.17 post-pandemic, and the Z-test indicating no significant change. While some themes, like mental health issues, showed a significant decrease in association post-pandemic, the risk for suicidal ideation generally remained higher than pre-pandemic levels, reflecting the lingering effects of the pandemic.

When comparing the pre-pandemic and post-pandemic periods, several themes maintained a significantly higher association with suicidal ideation. Mental health issues remained significantly associated with suicidal ideation post-pandemic (OR = 2.77, 95% CI: 1.83, 4.28) compared to pre-pandemic (OR = 1.07), with a significant increase confirmed by the Z-test (Z = -4.05, P < 0.001). Romantic relationship issues also remained significantly higher post-pandemic (OR = 2.46) compared to pre-pandemic, as indicated by a significant Z-score (Z = -2.99, P = 0.003). Marital and family issues continued to be more strongly associated with suicidal ideation post-pandemic (OR = 2.10) compared to pre-pandemic levels, with a significant Z-score (Z = -4.75, P < 0.001). These findings highlight that while the immediate changes related to the pandemic may have eased post-pandemic, the association between certain themes-especially mental health and relationship issues-and suicidal ideation remained significantly stronger than in the pre-pandemic period.

Association between consultation themes and suicidal ideation across different pandemic phases (female callers)

The analysis of female callers reveals significant changes in the association between various consultation themes and suicidal ideation across the pre-pandemic, mid-pandemic, and post-pandemic phases (Tables 7 and 8). From the pre-pandemic to mid-pandemic phases, several consultation themes showed an increase in their association with suicidal ideation among female callers. Mental health issues saw an increase in the odds ratio (OR) from 5.14 (95% CI: 3.97, 6.73) pre-pandemic to 7.36 (95% CI: 5.76, 9.50) mid-pandemic, although this increase was not statistically significant (Z = -1.94, P = 0.053). Romantic relationship issues also showed a rise in OR from 3.19 (95% CI: 2.29, 4.47) pre-pandemic to 4.70 (95% CI: 3.50, 6.34) mid-pandemic, though the change was not statistically significant (Z = -1.70, P = 0.089). Marital and family issues increased in OR from 3.69 (95% CI: 2.73, 5.04) pre-pandemic to 4.80 (95% CI: 3.62, 6.41) mid-pandemic, without statistical significance (Z = -1.23, P = 0.219). Work-related issues saw the OR rise from 1.68 (95% CI: 0.97, 2.82) pre-pandemic to 3.12 (95% CI: 2.02, 4.74) mid-pandemic, nearing statistical significance (Z = -1.78, P = 0.076). Learning issues slightly decreased in OR from 2.34 (95% CI: 1.46, 3.71) pre-pandemic to 1.93 (95% CI: 1.24, 2.96) mid-pandemic, with no statistical significance (Z = 0.59, P = 0.553). Despite these increases, none of these changes reached statistical significance, indicating a trend toward stronger associations with suicidal ideation but without strong evidence.

Table 7 Stratified logistic regression analysis of the association between consultation themes and suicidal ideation among females across different pandemic phases Table 8 Z-Test Comparison of odds ratios for consultation themes associated with suicidal ideation among females across different pandemic phases

During the transition from mid-pandemic to post-pandemic, the associations between certain consultation themes and suicidal ideation continued to increase. Mental health issues showed a significant rise in OR from 7.36 (95% CI: 5.76, 9.50) mid-pandemic to 15.5 (95% CI: 8.74, 30.1) post-pandemic (Z = -2.19, P = 0.028). Romantic relationship issues increased from an OR of 4.70 (95% CI: 3.50, 6.34) mid-pandemic to 6.88 (95% CI: 3.59, 14.2) post-pandemic, with the Z-test confirming a significant increase (Z = -1.97, P = 0.049). Marital and family issues also saw a significant increase in OR from 4.80 (95% CI: 3.62, 6.41) mid-pandemic to 10.2 (95% CI: 5.52, 20.6) post-pandemic (Z = -2.74, P = 0.006). Parenting and child education issues exhibited a significant rise in OR from 1.93 (95% CI: 1.24, 2.96) mid-pandemic to 5.18 (95% CI: 2.14, 12.4) post-pandemic (Z = -2.82, P = 0.005). Learning issues also showed a significant increase from an OR of 1.93 (95% CI: 1.24, 2.96) mid-pandemic to 5.00 (95% CI: 2.22, 11.5) post-pandemic (Z = -2.01, P = 0.045). Work-related issues saw an OR increase from 3.12 (95% CI: 2.02, 4.74) mid-pandemic to 4.49 (95% CI: 1.98, 10.4) post-pandemic, with the Z-test indicating a near-significant change (Z = -1.95, P = 0.051). These findings suggest that themes such as mental health, marital and family issues, romantic relationships, parenting/child education, learning issues, and work-related issues became more strongly associated with suicidal ideation as the pandemic progressed.

Comparing the pre-pandemic and post-pandemic phases highlights significant long-term increases in associations. Mental health issues showed a substantial increase in OR from 5.14 (95% CI: 3.97, 6.73) pre-pandemic to 15.5 (95% CI: 8.74, 30.1) post-pandemic, with a significant Z-test result (Z = -3.22, P = 0.001). Romantic relationship issues also increased significantly from an OR of 3.19 (95% CI: 2.29, 4.47) pre-pandemic to 6.88 (95% CI: 3.59, 14.2) post-pandemic, as confirmed by the Z-test (Z = -1.97, P = 0.049). Marital and family issues saw a significant rise in OR from 3.69 (95% CI: 2.73, 5.04) pre-pandemic to 10.2 (95% CI: 5.52, 20.6) post-pandemic (Z = -2.74, P = 0.006). Parenting and child education issues showed a significant increase in OR from 1.33 (95% CI: 0.62, 2.57) pre-pandemic to 5.18 (95% CI: 2.14, 12.4) post-pandemic, confirmed by the Z-test (Z = -2.36, P = 0.018). Work-related issues also increased significantly from an OR of 1.68 (95% CI: 0.97, 2.82) pre-pandemic to 4.49 (95% CI: 1.98, 10.4) post-pandemic, with the Z-test indicating a near-significant change (Z = -1.95, P = 0.051).

In summary, the pandemic led to significant increases in the association between suicidal ideation and various consultation themes among female callers. While the changes from the pre-pandemic to mid-pandemic phases were not always statistically significant, indicating a more gradual shift, the associations continued to strengthen throughout the pandemic. Unlike what was observed in male callers, the associations for female callers did not weaken post-pandemic; instead, they persisted and, in many cases, further intensified. This trend highlights a continuous and escalating association between the pandemic and female mental health, particularly in areas such as mental health issues, marital and family issues, parenting/child education issues, and romantic relationship issues, underscoring the critical need for sustained and targeted mental health support and intervention even as the pandemic’s immediate changes subside.