Association of demographics and psychological barriers with late presentation of HIV among men who have sex with men in Chengdu, China Authors: Jing Zhang, Junguo Xin, Dezhi Guo, Lan Chen, Lingli He, Jing Tang, Yuhen Li, Jiachuan Wang, Xiaohong Yang Published: 2025 ABSTRACT AbstractIn China, despite the improved availability of diagnostic services, men who have sex with men (MSM) still face high HIV risk and barriers to timely diagnosis due to internalized homophobia, self-stigma, and other potential factors. This study aimed to explore these factors and how they associated to late presentation (LP) of HIV among MSM in Chengdu, China. Characteristics of participants were collected by face-to-face interviews, including demographics, sexual behaviors, mental health, HIV knowledge, self-stigma, and internalized homophobia. Baseline CD4 at first presentation was obtained from the national integrated AIDS prevention and control information system. Logistic regression analysis was applied to analyze factors associated with LP (defined as a baseline CD4 count < 350 cells/µL). A total of 285 HIV-positive MSM participants were included, 56.5% were classified as late presenters (LPs). Multivariate regression analyses suggested that higher levels of internalized homophobia and self-stigma are positively associated with LP of HIV (both p < 0.001). Other linked factors included lower educational attainment (p = 0.039) and inadequate HIV-related knowledge (p < 0.001). Additionally, using severe anxiety as the reference group, individuals without anxiety symptoms showed an 86% lower likelihood of LP (p = 0.045, 95% CI: 0.02–0.96). Psychological barriers and a lack of related education are the potential barriers for LP of HIV among MSM in Chengdu, China. Interventions should focus on reducing homophobia and stigma while enhancing mental health support and HIV education to promote timely HIV testing in MSM.Keywords: HIV infection, Men who have sex with men, Late presentation, Internalized homophobia, Self-stigma, Mental healthSubject terms: Diseases, Health care, Psychology, Psychology, Risk factors Introduction Late presentation (LP) of HIV refers to a diagnosis when an AIDS-defining event (ADE) occurs or/and a CD4 cell count below 350 cells/ml1, which poses a threat to both individual health and public health systems2. Studies consistently indicate that LP is linked to higher mortality3,4 and elevated healthcare costs5 due to delayed initiation of antiretroviral therapy (ART) and missed opportunities for reducing HIV transmission through early intervention6. Globally, particularly in Asia, LP remains a persistent and significant public health problem. Despite expanded testing and treatment, an estimated 34%–72% of newly diagnosed individuals in Asia still meet LP criteria7. Late diagnosis remains common among key populations such as MSM; for instance, South Korea continue to report substantial LP burdens8. Evidence from Malaysia further suggests that delayed testing is often linked to fear of disclosure, and low risk perception9. Between 2005 and 2022, over half of newly diagnosed HIV/AIDS patients in China were presented late10,11. Among men who have sex with men (MSM), the average proportion of late presenters (LPs) was 56.53%, with a cumulative growth of 9.5% from 2015 to 202010. Despite efforts to improve HIV education and testing services targeting MSM. They still face barriers to HIV testing and treatment, such as reluctance to disclose their sexual orientation due to the fear of shame12. Evidence13 has suggested that stigma, homophobia, and the fear of disclosing sexual behavior significantly hinder MSM from seeking timely HIV prevention and care services, leading to worse outcomes. Internalized homophobia14 is an emotional experience and cognitive processing pattern formed in which individuals adopt negative attitudes or stereotypes about their sexual orientation, shaped by the predominantly heterosexual norms of society. Research15 has proposed that homosexuality and internalizing discrimination are prevalent in MSM regardless of HIV infection, especially in the elderly and married population. This emotional and cognitive distress can result in strained doctor-patient relationships, delayed medical care, and severe consequences for the health of MSM16. Self-stigma17 is also a severe psychosocial concern among Chinese MSM18. It refers to the internalization of negative beliefs, including discriminatory attitudes, stereotypes, or prejudices, about one’s characteristics. Various forms of stigma may reduce their willingness to undergo HIV or other sexually transmitted infection (STI) testing and diagnosis10. In addition, studies conducted in China19, India20 and the USA21,22 have shown a strong correlation between self-stigma and elevated HIV risk. Similar findings were also found in Internalized homophobia23,24. However, the associations between them and LP in MSM are still poorly understood. Chengdu, Sichuan province, has a relatively high rate of HIV infection25,26 and mobility27, making it a key area for HIV detection and prevention. In light of this, this study was conducted to explore the factors related to LP of HIV/AIDS in Chengdu, particularly the psychosocial factors of internalized homophobia and self-stigma among MSM. Understanding these barriers and whether they relate to LP is urgent for designing effective interventions to promote early diagnosis and improve health among MSM28. Methods Study design and population This facility-based cross-sectional study was performed at the Nuclear Industry 416 Hospital (The 2nd Affiliated Hospital of Chengdu Medical College), one of the largest HIV treatment centers in Chengdu. From 2009 to 2023, a total of 1,082 MSM living with HIV received care at this hospital. Considering the funding constraints and feasibility, we used a pre-designed simple random sampling method to represent the entire HIV/AIDS-diagnosed MSM patient population in the hospital. The sample size was estimated using the following equation: The parameter n represents the required sample size for the study. Where P (prevalence of LP) was taken as 28% which is based on a prior study conducted in China29. We took α as 0.05, as 1.96, and 5% of as an allowable error, the calculated sample size was 307. The inclusion criteria were: (1) participants > 18 years of age or older; (2) had ever engaged in same-sex sexual behavior; (3) having no cognitive impairments. Patients with incomplete data, such as missing CD4 test results or other required information were excluded. After excluding participants for refusal to participate or incomplete information (n = 22), a total of 285 patients were included. This research strictly adhered to the principles of the “WMA Declaration of Helsinki—Ethical Principles for Medical Research Involving Human Participants” and relevant Chinese laws and regulations. All participants provided voluntary informed consent after being fully informed of the research’s purpose and procedures. All personal information was anonymized throughout the research process. The research protocol was approved by the Ethics Committee of Nuclear Industry 416 Hospital (registration number: R-5M1-23-028718). Data collection and definition Medical records of HIV-positive MSM between January 2009 and December 2023 were reviewed. The total number of new diagnoses during this period was summarized and categorized into NLPs and LPs. Among them, selected participants were surveyed by trained interviewers and completed a questionnaire face-to-face. Demographic data (age, education, marital status, occupation, income level, etc.), HIV-related knowledge, depression, anxiety, internalized homophobia and self-stigma were collected. Baseline CD4 is obtained from the National Integrated AIDS Prevention and Control Information System. The criteria for defining LP of HIV/AIDS is a baseline CD4 cell count of less than 350 cells/µL, an ADE reported or progression to AIDS within one year. Knowledge about HIV/AIDS was assessed using eight questions from the HIV/AIDS core knowledge and questionnaire for HIV education and awareness, published by the China Center for Disease Control and Prevention (China CDC) in 201611. Responses were scored as follows: 1 for correct answers, and 0 for both incorrect and ‘don’t know’ responses, with a maximum possible score of 8. A total score of 6 or higher was considered to have adequate HIV-related knowledge, otherwise defined as inadequate. The Patient Health Questionnaire-9 (PHQ-9)30 was applied to evaluate symptoms of depression, with the full version consisting of nine items and a total score range from 0 to 27. PHQ-9 scores of 5–9, 10–14, 15–19, and 20–27 indicate mild, moderate, moderately severe, and severe depressive symptoms, respectively. The Generalized Anxiety Disorder-7 (GAD-7)31 was used to evaluate symptoms of anxiety. The scale includes seven items, with responses ranging from 0 (not at all) to 3 (almost every day). Scores of 5–9, 10–14, and 15–21 representing mild, moderate, and severe anxiety symptoms, respectively. Internalized homophobia was assessed using the Short Internalized Homonegativity Scale (SIHS)11, a seven-item scale that evaluates the negative emotions participants may have due to their sexual orientation. Each item is rated on a scale from 1 (strongly disagree) to 7 (strongly agree), with item 5 being reverse scored. The scale comprises three dimensions of self-stigma. The scale comprises three dimensions of self-stigma including personal comfort with a gay identity, social comfort with gay men and public identification as gay. The mean score serves as the measurement indicator, with lower scores indicating more serious internalized homophobia. Self-stigma was assessed through the Self-Stigma Scale-Short Form (SSS-S)32,33, an established tool for assessing the internalized stigma associated with being homosexual in China. A 4-point Likert scale was applied to rate the level of agreement with each question, ranging from 1 (strongly disagree) to 4 (strongly agree). The scale comprises nine items that assess the affective, behavioral, and cognitive dimensions of self-stigma. A higher score indicates stronger self-stigma. Statistical analysis Characteristics of participants were summarized as frequencies or percentages for categorical variables, and between-group comparisons were performed using the chi-square test; Fisher’s exact test was used when the expected cell count was < 5. Continuous quantitative variables with a normal distribution were summarized as means ± standard deviations (SDs); differences between two groups were assessed using the t test, and differences among multiple groups were examined using analysis of variance (ANOVA). Continuous quantitative variables with a skewed distribution were summarized as medians and interquartile ranges (IQRs); differences between two groups were assessed using the Mann–Whitney U test, and comparisons among multiple groups were conducted using the Kruskal–Wallis test. Univariate and multivariate analyses were performed to calculate odds ratios (OR) using logistic regression. In the multivariate analysis, variables with p < 0.10 in univariable analyses and variables that may related to LP were included to further explore the associated factors. Both crude and adjusted odds ratios (aOR), 95% confidence interval (CI), and the corresponding p-values were reported. A two-tailed p-value below 0.05 is considered significant. Data were analyzed in R (version 4.3.1; R Core Team, 2023; R Foundation for Statistical Computing, Vienna, Austria). Results The characteristics of the study population were summarized in (Table 1). Overall, the mean age of the participants was 33.98 years (SD = 8.69), and a large proportion (70.9%) had attained an education of associate or bachelor’s degree. Most participants (84.9%) were unmarried, and 13.1% reported being married or divorced. Of all participants, more than one-half of the participants (61.4%) were employed in white-collar jobs, and 72.3% reported a monthly income below 10,000 yuan. Additionally, 213 (74.7%) of the patients had adequate HIV-related knowledge. When comparing the two groups, 56.5% of the participants were classified as LPs, with a median CD4 counts of 222 cells/µL, and 43.5% were classified as NLPs, with a median CD4 counts of 450 cells/µL. While patients with low education (high school or below) often presented late (72.9%, p = 0.014). Moreover, the proportion of MSM without adequate knowledge was significantly higher among LPs (38.5%) compared to NLPs (8.1%) (p < 0.001).In our studies, 50.9% of the participants exhibited symptoms of depression (PHQ > 9), and 43.5% showed signs of anxiety (GAD > 7). More anxious symptoms were reported in LPs, especially the moderate and severe anxiety symptoms (17.4% vs. 4.8%, p < 0.001). Table 1Sociodemographic and psychological characteristics of the MSM overall, non-late presenters and late presenters.CharacteristicsTotalNLPsLPsp-value(NLPs vs. LPs)n (%)n (%)n (%)All285 (100%)124 (43.5%)161 (56.5%)Age0.60118–2998 (34.4%)45 (36.3%)53 (32.9%)30–39124 (43.5%)55 (44.4%)69 (42.9%)40 and above63 (22.1%)24 (19.4%)39 (24.2%)Education level 0.014 High school or below59 (20.7%)16 (12.9%)43 (26.7%)*Associate or bachelor’s degree202 (70.9%)94 (75.8%)108 (67.1%)Master’s degree or above24 (8.4%)14 (11.3%)10 (6.2%)Marital status0.135Unmarried248 (84.9%)109 (88.7%)133 (85.7%)Married18 (6.3%)4 (3.2%)14 (8.7%)Divorced19 (6.8%)10 (8.1%)9 (5.6%)Occupation0.239White collar175 (61.4%)81 (65.3%)94 (58.4%)Blue collar110 (38.6%)43 (34.7%)67 (41.6%)Income levelNo income19 (6.7%)6 (4.8%)13 (8.1%)0.072Less than 2000 yuan16 (5.6%)7 (5.6%)9 (5.6%)2000–5000 yuan57 (20%)24 (19.4%)33 (20.5%)5000–10,000 yuan114 (40%)61 (49.2%)53 (32.9%)10,000–20,000 yuan65 (22.8%)23 (18.5%)42 (26.1%)More than 20,000 yuan14 (4.9%)3 (2.4%)11 (6.8%)HIV-related knowledge < 0.001 Adequate72 (25.3%)10 (8.1%)62 (38.5%)*Inadequate213 (74.7%)114 (91.9%)99 (61.5%)*Depression symptoms0.716None140 (49.1%)60 (31.5%)80 (27.3%)Mild83 (29.1%)39 (48.4%)44 (49.7%)Moderate39 (13.7%)18 (14.5%)21 (13%)Moderately severe16 (5.6%)5 (4.0%)11 (6.8%)Severe7 (2.5%)2 (1.6%)5 (3.1%)Anxiety symptoms 0.013 None161 (56.5%)78 (62.9%)83 (51.6%)Mild90 (31.6%)40 (32.3%)50 (31.1%)Moderate21 (7.4%)4 (3.2%)17 (10.6%)*Severe13 (4.6%)2 (1.6%)11 (6.8%)*CD4 count, median (IQR)324(207–439)450 (399–559)*222 (139–291)* < 0.001 SIHS, means ± SD4.04 ± 1.514.49 ± 1.36*3.69 ± 1.45* < 0.001 SSS-S, means ± SD2.22 ± 0.871.90 ± 0.74*2.47 ± 0.85* < 0.001 Bold values and *superscripts indicate significant difference between groups.LPs, late presenters; NLPs, non-late presenters; n, number of MSM; SIHS, the short internalized homonegativity scale; SSS-S, the self-stigma scale-short form. Sociodemographic and psychological characteristics of the MSM overall, non-late presenters and late presenters. 324 (207–439) Bold values and *superscripts indicate significant difference between groups. LPs, late presenters; NLPs, non-late presenters; n, number of MSM; SIHS, the short internalized homonegativity scale; SSS-S, the self-stigma scale-short form. The mean score for the SIHS and SSS-S was 4.04 (SD = 1.51) and 2.22 (SD = 0.87), respectively, with LP being more common among MSM who had higher levels of internalized homophobia and self-stigma (both p < 0.001). “Social situations with gay men make me feel uncomfortable” on the SIHS and “I fear that others would know that I am an MSM” on the SSS-S were the most concerning items for participants. Tables 2 and 3 summarized the details of SIHS and SSS-S. Table 2Descriptive statistics of short internalized homonegativity scale among MSM in Chengdu.Domain and itemMean ± SDPersonal comfort with a gay identity (PC)Even if I could change my sexual orientation, I wouldn’t.4.32 ± 1.50I feel comfortable being a homosexual man.3.75 ± 1.56Homosexuality is morally acceptable to me.4.07 ± 1.77Social comfort with gay men (SC)I feel comfortable in gay bars.4.19 ± 1.47Social situations with gay men make me feel uncomfortable.3.61 ± 1.24Public identification as gay (PUBID)I feel comfortable discussing homosexuality in a public situation.4.09 ± 1.52I feel comfortable being seen in public with an obviously gay person.4.25 ± 1.51Mean Internalized homophobia4.04 ± 1.51MSM: men who have sex with men; SD: standard deviation. Descriptive statistics of short internalized homonegativity scale among MSM in Chengdu. MSM: men who have sex with men; SD: standard deviation. Table 3Descriptive statistics of self-stigma scale-short form among MSM in Chengdu.Domain and itemMean ± SDAffective dimensionI fear that others would know that I am a MSM.2.75 ± 1.09I feel helpless because I am a MSM.2.46 ± 0.97I feel uncomfortable because I am a MSM.2.48 ± 0.93Behavioral dimensionI dare not to make new friends lest they find out that I am a MSM.2.00 ± 0.79I estrange myself from others because I am a MSM.2.34 ± 0.85I avoid interacting with others because I am a MSM.1.78 ± 0.79Cognitive dimensionThe identity of being a man who has sex with men taints my life.2.02 ± 0.77My identity as a MSM incurs inconvenience in my daily life.1.90 ± 0.79My identity as a MSM is a burden to me.2.28 ± 0.87Mean Self-stigma2.13 ± 0.85MSM: men who have sex with men; SD: standard deviation. Descriptive statistics of self-stigma scale-short form among MSM in Chengdu. MSM: men who have sex with men; SD: standard deviation. Multicollinearity was assessed using variance inflation factors (VIFs) and tolerance. All predictors had VIFs < 5 and tolerance values > 0.2, indicating no evidence of significant multicollinearity. Table 4 summarizes the factors associated with LP of HIV. Univariate and multivariate analyses consistently identified education level, HIV-related knowledge, anxiety symptoms, internalized homophobia, and self-stigma as associated factors for LP. In the multivariate analysis, individuals with an education of high school or below were more likely to experience LP compared to those who had a master’s degree or higher [ p = 0.039, (aOR: 4.14, 95% CI: 1.08–15.9)]. A lack of HIV-related knowledge was significantly linked to risk of LP [ p < 0.001, (aOR: 7.19, 95% CI: 3.20-16.15)]. Participants without anxiety symptoms were less likely to present late compared to those with severe symptoms [ p = 0.045, (aOR: 0.14, 95% CI: 0.02–0.96)], but we found no association between depression and LP. Lower scores on the SIHS were identified as a risk factor for LP [ p < 0.001, (aOR: 0.32, 95% CI: 0.20–0.52)], while lower scores on the SSS-S were associated with a decreased risk of LP. [ p < 0.001, (aOR: 3.46, 95% CI: 1.86–6.44)]. Table 4Univariate and multivariate logistic regression analysis of factors associated with HIV late presentation.CharacteristicUnivariable analysisMultivariable analysisCrude OR (95% CI)p-valueaOR (95% CI)p-valueAge18–290.73 (0.38–1.38)0.32830–390.77 (0.42–1.44)0.41340 and above1.00 (ref)-Education levelHigh school or below3.76 (1.39–10.17) 0.009 4.14 (1.08–15.9) 0.039 Associate or bachelor’s degree1.61 (0.68–3.79)0.2771.75 (0.57–5.44)0.251Master’s degree or above1.00 (ref)-1.00 (ref)Marital statusUnmarried0.26 (0.06–1.08)0.063Married0.36 (0.12–1.12)0.078Divorced1.00 (ref)-OccupationWhite collar0.75 (0.46–1.21)0.234Blue collar1.00 (ref)-Income levelNo income1.00 (ref)-Less than 2000 yuan0.59 (0.15–2.37)0.4592000–5000 yuan0.64 (0.21–1.91)0.4185000–10,000 yuan0.40 (0.14–1.13)0.08410,000–20,000 yuan0.84 (0.28–2.51)0.759More than 20,000 yuan1.69 (0.34–8.40)0.520HIV-related knowledgeAdequate1.00 (ref)-1.00 (ref)-Inadequate7.14 (3.48–14.67) < 0.001 7.19 (3.20-16.15) < 0.001 Anxiety symptomsNone0.19 (0.04–0.90) 0.036 0.14 (0.02–0.96) 0.045 Mild0.23 (0.05–1.09)0.0630.25 (0.04–1.70)0.162Moderate0.77 (0.12–4.96)0.7860.63 (0.07–5.93)0.685Severe1.00 (ref)-1.00 (ref)-SIHS0.29 (0.20–0.42) < 0.001 0.32 (0.20–0.52) <0.001 SSS-S 4.68 (2.95–7.44) < 0.001 3.46 (1.86–6.44) <0.001 Bold values indicate statistical significance.aOR, adjusted odds ratio; CI, confidence interval; OR, odds ratio; SIHS, the short internalized homonegativity scale; SSS-S, the self-stigma scale-short form. Univariate and multivariate logistic regression analysis of factors associated with HIV late presentation. Bold values indicate statistical significance. aOR, adjusted odds ratio; CI, confidence interval; OR, odds ratio; SIHS, the short internalized homonegativity scale; SSS-S, the self-stigma scale-short form. Discussion This study examined the prevalence of LP of HIV among MSM in Chengdu, China. The results showed a high proportion of LP cases, which is similar to recent studies34,35 from China but higher than the average risk reported in the general population36. Our findings suggested that educational level, HIV-related knowledge, anxiety, internalized homophobia, and self-stigma were associated with LP of HIV among MSM in Chengdu. Notably, we did not find a significant association between LP and socio-demographic factors including age, marital status, or income. Consistent with studies from Spain3 and Norway37, our study found that low levels of education were associated with a higher likelihood of LP among MSM. One of the main factors associated with late presentation in this study was limited HIV-related knowledge. This finding aligns with a study conducted in Switzerland38, which reported insufficient knowledge about HIV transmission and symptoms as the primary factor linked with delayed testing. Our findings link poor mental health to LP of HIV, this has been the case in a previous study39. The anxiety rate among MSM in our study was 43.5%, which is higher than the anxiety rate of 21.8% in the general Chinese population40. This study found that higher levels of internalized homophobia and self-stigma are independently associated with increased odds of LP. Evidence24,41,42 has promoted that internalized homophobia and stigma significantly reduce testing behaviors and adherence to treatment among MSM. Other research from Ethiopia43 also found that fear of stigma and discrimination was more common among individuals who were late for HIV diagnosis. This result may explained by the fact that lower educational attainment often limits health literacy, restricts access to accurate HIV-related information, and reduces engagement with preventive services. Previous research44 has reported that individuals with lower health literacy are more likely to delay HIV testing. Conversely, higher education levels tend to foster greater engagement with healthcare services, leading to earlier HIV diagnoses45. Without adequate knowledge, patients may fail to recognize early symptoms, leading to reduced risk perception and delays in testing37,46. These findings underline the importance of targeted interventions aimed at addressing the potential knowledge gap, particularly in high-risk populations such as MSM, and it has been proven effective in the USA47. Mental health is highly prevalent among MSM in China48. Depression and anxiety have been identified as key risk factors for Unprotected Anal Intercourse (UAI) and poor HIV care49–51, which increases the risk of HIV infection and LP in MSM. Considering that HIV-infected individuals with poor mental health related to have higher mortality rates52, comprehensive medical services are needed to address the psychological burden of LP. These findings reflected their pervasive impact on healthcare access among MSM. Internalized homophobia and self-stigma create psychological barriers53, indirectly affect social functioning54 and self-esteem55, reduce the likelihood of seeking health-promoting behaviors56, and hinder MSM from accessing HIV testing services57. Addressing these issues through stigma-reduction interventions and fostering supportive environments are essential for improving timely HIV diagnosis. Interventions such as community-based peer support and education programs have proven effective in reducing psychological barriers and promoting earlier healthcare engagement58. This lack of association may reflect the homogeneity of the management. MSM in the city are managed by regional disease control and community health service centers, ensuring high availability of HIV testing, which may mitigate the impact of demographic variables and equalize access across different groups59,60. Moreover, stigma associated with sexual orientation and poor mental health may have a greater impact on LP, suggesting that integrating mental health support into HIV care services may be more urgent than focusing solely on demographic predictors61. This study has several disadvantages. First, the cross-sectional design limits the ability to draw causal conclusions about the relationships between variables and LP of HIV. Second, Because the social–psychological variables were measured at the time of the questionnaire survey, whereas LP reflects diagnostic status at the initial visit or confirmation, we cannot rule out the possibility that individuals diagnosed at a late stage may subsequently report higher levels of stigma or anxiety after diagnosis, rather than these factors leading to LP. Third, the study population was drawn from a single tertiary hospital in Chengdu, which may restrict the ability to generalize the findings to other regions or healthcare environments. Conclusions These findings underscore the need for multifaceted interventions including enhancing awareness about HIV/AIDS, developing targeted strategies for internalized homophobia and self-stigma reduction, especially among individuals with low educational attainment. Given the high prevalent of LP among MSM, more epidemiological studies are needed to confirm these associations and further improve early diagnosis among MSM living with HIV. Acknowledgements The authors gratefully acknowledge for the staff of the Department of Infectious Diseases, Nuclear Industry 416 Hospital for their cooperation during the data collection phase and extend our heartfelt gratitude to them. Author contributions Jing Zhang, Junguo Xin and Dezhi Guo was responsible for statistical analysis and writing the manuscript. Lan Chen and Lingli He, Jing Tang and Yuhen Li were involved in the data collection and interpretation. Jing Zhang, Junguo Xin, Jiachuan Wang and Xiaohong Yang were responsible for study design and the critical revision of the manuscript. All authors have contributed to and have approved the final manuscript. Funding This work was supported by the Chengdu Medical College Clinical Science Research Fund Project [grant numbers 2022LHFSZD-01] and Graduate Student Innovation Fund of Chengdu Medical College [Grant Number YCX2024-01-81]. Data availability The dataset analyzed during the current study is not publicly available to protect the privacy of study participants given the small size of the recruited sample. However, de-identified personal data can be obtained from the corresponding author upon reasonable request. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Jing Zhang, Junguo Xin and Dezhi Guo contributed equally to this work. Contributor Information Jiachuan Wang, Email: wjc416@126.com. Xiaohong Yang, Email: 21376686@qq.com. References Associated Data Data Availability Statement The dataset analyzed during the current study is not publicly available to protect the privacy of study participants given the small size of the recruited sample. However, de-identified personal data can be obtained from the corresponding author upon reasonable request.