Abstract

Background

Methamphetamine is one of the most abused drugs worldwide. Methamphetamine use disorders (MAUD) not only affect the physical and mental health of addicts but also increase the risk of suicide. Although the suicide rate of MAUD patients is high, there is little research on the clinical characteristics and risk factors of suicide risk in MAUD patients. So, it is necessary to investigate the related risk factors of suicidal ideation of MAUD.

Methods

Using response-driven sampling (RDS), a total of 10,225 MAUDs were randomly selected from April to May 2023 in Guangdong, China. MAUD was evaluated by the Beck Scale for Suicide Ideation (BSSI) and divided into a case (n鈥�=鈥�1228) and a control group (n鈥�=鈥�8997). Multivariate generalized linear mixed model analysis was used to analyze the risk factors of suicidal ideation among MAUDs.

Results

The result of multivariate generalized linear mixed model analysis showed that the dosage of each use of MA (OR鈥�=鈥�5.52, 95%CI鈥�=鈥�4.93锝�6.17), age (OR鈥�=鈥�1.11, 95%CI鈥�=鈥�1.03锝�1.20), childhood physical abuse(OR鈥�=鈥�3.43, 95%CI鈥�=鈥�3.17锝�3.71), childhood emotional neglect(OR鈥�=鈥�2.77, 95%CI鈥�=鈥�2.60锝�2.94), introverted and extroverted personality(OR鈥�=鈥�2.85, 95%CI鈥�=鈥�2.64锝�3.08), and neurotic personality(OR鈥�=鈥�1.15, 95%CI鈥�=鈥�1.02锝�1.30) were the risk factors of suicidal ideation among MAUDs in Guangdong, China.

Conclusion

Suicidal ideation among MAUDs is influenced by family, society, stress, and other factors, so it is necessary to take comprehensive intervention measures.

Introduction

Methamphetamine is one of the most abused drugs worldwide, and it poses a widespread threat to human health and life safety [1,2,3]. According to the 2022 United Nations World Drug Report, 34听million people aged 15鈥�64 in the world used amphetamine-type stimulants in 2020 [4]. The addiction problems caused by the abuse of methamphetamine, such as methamphetamine use disorder (MAUD), have become a public health issue of great concern worldwide. Methamphetamine use disorder (MAUD) is a cognitive, behavioral, and physiological syndrome caused using methamphetamine. It is a serious public health problem characterized by compulsive drug-seeking behavior and significant neurotoxicity, which imposes a heavy burden on individuals and society. According to the classification and diagnosis of substance-related and addictive disorders in the Diagnostic and Statistical Manual of Mental Disorders 5th Edition (DSM-5) [5], stimulant use disorders lead to clinically significant damage or pain, manifested as increased intake, tolerance, difficulty in limiting intake, severe withdrawal symptoms and negative effects during the restriction process Symptoms such as drug seeking and craving.

MAUD not only affects the physical and mental health of addicts [6], causes the spread of AIDS and other infectious diseases, but also increases the risk of suicide [7]. Related studies have shown that compared to non-methamphetamine users, methamphetamine users have a 4.4 times higher risk of attempting suicide [8]. In contrast, patients with methamphetamine use disorders have a standardized suicide mortality rate as high as 17 times [9]. A meta-analysis of the mortality rate of methamphetamine users showed that those who used methamphetamine had a 12-fold increased risk of suicide compared to the general population [10]. Lee et al. [11] recently conducted a large local cohort study in Taiwan, China. The results showed that the standardized mortality rate of suicide in MAUD patients was the highest, 16 times higher than that in the general population. One of the most common substances found in blood or urine samples of suicide victims is methamphetamine, which also means that methamphetamine is a proximal risk factor for suicide [12]. Methamphetamine users have reduced dopamine neurotransmission in their bodies, which plays an important role in suicide.

There is no consensus on the definition of suicidal ideation. Some scholars contend that it merely encompasses an individual鈥檚 inclination to end their own life, while others argue that it signifies a strong desire and motivation to terminate one鈥檚 existence, along with the contemplation of suicide methods. Additionally, some experts believe that the formulation of suicide plans should also fall under the umbrella of suicidal ideation [13]. Suicidal ideation serves as a risk factor for suicide, potentially exerting long-lasting effects [14], and thus, it can be characterized by both state and trait attributes. Two prominent theories can elucidate the emergence of suicidal ideation. The first is the quality stress model proposed by Schotte and Clum [15], which posits that stress alone is not sufficient to trigger suicidal ideation. Even under identical stress conditions, many individuals do not develop suicidal thoughts or behaviors. The risk of suicidal ideation and behavior escalates only when stressed individuals already possess certain susceptibility traits. Another iteration of the quality stress model, introduced by Mann et al. [16], clarifies the concepts of stress and quality. Stress specifically pertains to mental and psychological disturbances, whereas quality represents trait-based psychological structures, such as an impulsive personality. The second theory is the Integrated Motivation-Volitional Model (IMV) [17], which categorizes the occurrence of suicidal behavior into three distinct stages: pre-motivational, motivational, and volitional. The pre-motivational stage encompasses background factors like personal 鈥渜ualities,鈥� external environmental influences, and life events. The motivational stage is marked by feelings of frustration, shame, and entrapment, which progress to the emergence of suicidal ideation. The volitional stage, building upon suicidal thoughts, involves acquiring methods or means for suicide, exposure to suicide-related information, or impulsive actions, ultimately leading to the act of suicide.

Both theories offer valuable frameworks for researchers to examine the development of suicidal thoughts and behaviors. However, to gain a deeper understanding of the root causes of suicidal ideation, it is essential to consider not only the interplay of internal risk factors but also potential underlying mechanisms. Given that the human mind is a complex and multifaceted system, its psychological processes are best explained and interpreted through dynamic models rather than traditional static theories. Furthermore, when studying the specific population of drug-dependent individuals, investigating the distribution and determinants of suicidal ideation necessitates a focus on their unique personality traits and specific environmental factors to unravel the mechanisms at play.

Thus, this work aimed to obtain the influencing factors of suicidal ideation among methamphetamine addicts in Guangdong Province, China. This study explores risk factors for suicidal ideation based on sociodemographic characteristics, MA usage characteristics, childhood trauma, and personality traits to provide a scientific basis for relevant departments to adopt targeted prevention and comprehensive intervention measures.

Methods

Study population

This study is a cross-sectional survey design. From April to May 2023, the researchers used a response-driven sampling (RDS) method to conduct continuous sampling of methamphetamine users in four sampled cities in Guangdong Province and conducted semi-structured interviews and questionnaire surveys.

Using a multi-stage stratified random sampling method stratified by region and economic level, one city (Guangzhou, Shantou, Qingyuan, and Maoming) was randomly selected from the eastern, western, northern, and Pearl River Delta regions of Guangdong Province from April to May 2023. Three districts (counties) were randomly selected from each selected city, and three townships (streets) were selected from each district (county). 145 MA use populations were randomly selected from each township (street). The MA use population was sampled using response-driven sampling (RDS). This sampling method is developed from the long-chain recommendation method, which is driven by peer response. Firstly, the local drug user dynamic control system will be used to recruit MAUD individuals as seeds, and face-to-face questionnaire surveys will be conducted by trained investigators. After completing the survey, the seed will receive recruitment cards and souvenirs (Junior Incentives). Afterward, the seed is responsible for recruiting peers who reside in the same township (street) and handing them a peer recruitment card. Each seed can only recommend a maximum of 5 peers. Companions bring recruitment cards to the designated survey location, and the investigator screens whether they meet the research conditions. If they meet the conditions, they obtain informed consent from the visitors and then conduct a questionnaire survey. At the same time, peers receive the same recruitment cards and primary incentives. Recruiters will receive souvenirs (secondary incentives) for each successful recruitment of a companion residing in the same township (street) to participate in the survey. In this cycle, after multiple rounds of recruitment, the characteristic variable composition of the obtained sample tends to stabilize. Recruitment will be stopped after obtaining a sample that can represent the overall population and is independent of the seed.

RDS is a type of chain nomination sampling, but its recruitment process allows for the calculation of selection probabilities, making it an approximate probability sampling method. First, RDS differs from traditional snowball sampling by using a bidirectional incentive mechanism, where both the recruiter and the recruited individual receive a certain amount of material rewards or cash, thereby improving recruitment efficiency. Second, RDS requires that the survey subjects be recruited by peers rather than having peers provide the names of other members to the investigators, thus reducing the proportion of refusals during recruitment. RDS stipulates that everyone can only nominate a certain number of peers (typically 3鈥�5) while also analyzing the number of peers each nominator knows, thereby enabling the calculation of the probability of each peer being selected. This helps to compensate for the over-sampling of respondents with large social networks and reduces recruitment bias. The limitation on the number of peers an individual can nominate extends the recruitment chain, allowing peers with different characteristics to be selected and avoiding bias caused by similarity in population characteristics. As an approximate probability sampling method, RDS allows for point estimates and interval estimates of the sampled population. The target population members (seeds) who participate in the survey elect other members (peers) to participate, and the sample is composed of continuous recommendations from the participants.

All MA users underwent urine toxicology screening and self-reporting to confirm MA use upon admission. A diagnosis of MAUD was established through semi-structured clinical interviews for DSM-V, Axis I, Patient Version, and SCID-I/P (Chinese version) by a professional psychiatrist with at least five years of clinical experience. Objects in the study must meet the following inclusion criteria: (1) All survey subjects met the diagnostic criteria for methamphetamine use disorders in the International Classification of Diseases, Tenth Version (ICD-10) and the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-V); (2) Residents with the household registration of Guangdong or residents unmoved more than six months in the geographical scope of Guangdong; (3) The drug abusers who have been deprived of drug addiction for less than one years as of May 30, 2023, consisting of the ones still in social activities, the ones who have been forced into drug rehabilitation centers or voluntarily detoxified; (4) The age is ranging from 18 to 60 years; (5) No history of diabetes and special nervous system diseases, including head trauma and epilepsy. Exclusion criteria: Patients are unwilling to continue participating.

The Beck Scale for Suicide Ideation (BSSI) [18] was used to evaluate the suicidal ideation of the MAUD population, who were divided into a suicidal ideation group and a non-suicidal ideation group based on the total weighted score of the BSSI. The final investigation involved 10,225 cases of MAUD, including 1228 cases in the suicidal ideation group and 8997 cases in the non-suicidal ideation group. All participating patients have been approved by the Medical Ethics Committee and have signed informed consent forms.

Study tools

(1) The Beck Scale for Suicide Ideation (BSSI) [18] was used to evaluate the suicidal ideation and suicidal tendencies of MAUD patients. The scale was developed by Beck in 1979 to assess the intensity of specific attitudes, behaviors, and plans related to suicide and is the most widely used suicide ideation standard assessment tool. BSSI was originally designed as a self-assessment tool for psychiatric patients evaluated by interviewers. It has since been widely used in other non-psychiatric populations and can also serve as a self-assessment tool to screen for suicide risk [19].

BSSI has a total of 19 items, evaluating the following two dimensions: suicide ideation (SI: items 1鈥�5) and suicide tendency (ST: items 6鈥�19). Each item is graded based on the intensity of suicidal ideation using a 3-level Likert rating system: 0鈥�=鈥塶o thoughts, 1鈥�=鈥墂eak thoughts, 2鈥�=鈥塵oderate to strong thoughts. Then, these item scores are added together to obtain a total score ranging from 0 to 38 points. Specific items assess specific characteristics related to suicide, such as the desire to die, the desire for active or passive suicide attempts, the duration and frequency of suicidal ideation, a sense of control over suicide attempts, and the amount of preparation work done for suicide. Most participants only need to complete the first five items, namely the suicide ideation subscale. If items 4 and 5 score 0, they are considered non-suicidal individuals and do not need to evaluate the second dimension, namely suicidal tendencies. If the score for item 4 or item 5 is not 0, it indicates suicidal ideation, and further evaluation of the suicidal tendency dimension is required. This scale showed good internal consistency in this study鈥檚 MAUD case questionnaire group, with the overall scale Cronbach鈥檚 伪 coefficient of 0.93 and the subscale Cronbach鈥檚 伪 coefficient scores ranging from 0.84 to 0.90.

(2) A self-designed general situation survey questionnaire includes gender, age, education level, marital status, occupation, whether they are only children, whether they were mainly cared for by their parents before reaching adulthood, and family history of mental illness, among other things. The Only child refers to the only child born to a couple. The family history of mental illness refers to whether a person鈥檚 paternal, maternal, direct, or collateral lineage, within three generations, has ever suffered from mental illnesses such as anxiety, obsessive-compulsive disorder, depression, bipolar disorder, schizophrenia, etc. The caring by parents before reaching adulthood refers to parents personally raising their children until they reach the age of 18.

(3) A self-designed MA usage feature information table was used to collect MA user feature information, including substances and frequency of use (excluding MA, tobacco, and alcohol, those who have used other substances 1 or more times are not included in this study, such as heroin, ketamine, marijuana, etc.); Age of first use of MA (continuous variable, divided into two groups with a cutoff value of 30 years, group under 30 years and group over 30 years), first use of MA (with/without), years of use of MA (continuous variable, divided into two groups with a cutoff value of 10 years, group under 10 years and group over 10 years), dosage of each use of MA (continuous variable, divided into two groups with a cutoff value of 0.5听g, group under 0.5听g and group over 0.5听g), frequency of use of MA in the past month (continuous variable, score range of 0鈥�5 points, 0 is greater than once a month, 1 is once a month, 2 is once a 10鈥�15 days, 3 is once every 4鈥�9 days, 4 is once every 2鈥�3 days, 5 is at least once a day, with 0鈥�2 points classified as no more than 3 times/month group and 3鈥�5 points classified as more than 3 times/month group), Longest integrity period of MA (continuous variable, divided into two groups with a cutoff value of 12 months, the group below 12 months and the group above 12 months), while MA currently craves (continuous variable, scored from 0 to 10 points, and divided into craving group and non-craving group based on whether the score is 0).

(4) The Alcohol Use Disorders Identification Test (AUDIT) [20] is used to screen for alcohol use disorders in MAUD patients in the year before investigation. This scale is a standard tool used to evaluate an individual鈥檚 level of alcohol addiction. The entries in this scale help distinguish between 鈥渉azardous鈥� drinkers (those at high risk of encountering alcohol-related physiological or psychological hazards) and 鈥渉armful鈥� drinkers (those who have already experienced such problems). AUDIT is suitable for alcohol users of different ages, genders, and regions, and its Chinese version is also reliable and valid. AUDIT has a total of 10 items, using a Likert 5 scale, corresponding to 0 to 4 points, to evaluate the following three types of alcohol use issues: alcohol intake (items 1鈥�3), alcohol dependence (items 4鈥�6), and adverse consequences of alcohol (items 7鈥�10). The setting of the items also prioritizes questions that can lead to possible answers as a basis for short-term treatment interventions. AUDIT only takes 2听min to complete, and its scoring rules are simple and easy to operate. Add up the scores of all items, and the final total score range is 0鈥�40 points. The higher the score, the stronger the individual鈥檚 dependence on alcohol.

(5) The Fagerstr 枚 m Test for Nicotine Dependence (FTND) [21] was used to screen for tobacco use issues in MAUD patients in the year before investigation. This scale is a standard tool for evaluating an individual鈥檚 level of nicotine addiction. FTND is a modified version of the Fagerstr 枚 m Tolerance Questionnaire designed and developed by Todd Heatherton et al. The FTND consists of 6 items that evaluate the following three types of cigarettes use issues: cigarette consumption, smoking impulsiveness, and smoking dependence, namely the number of cigarettes smoked per day, the time of smoking the first cigarette per day, difficulty in controlling when smoking is prohibited, continuing to smoke during bed rest due to illness, the most difficult situation to abandon smoking, and the frequency of smoking. The scoring rules for each item are as follows: scores for non-items are 0 to 1, scores for multiple-choice items are 0 to 3, and finally, the total score of the scale is obtained by adding up the scores of all items, ranging from 0 to 10 points. The higher the score, the stronger the individual鈥檚 dependence on nicotine, usually divided into five levels: very low (0鈥�2 points), low (2鈥�4 points), medium (5 points), high (6鈥�7 points), and very high (8鈥�10 points). Some studies use different classification methods, such as low (0鈥�3 points), medium (4鈥�5 points), and high (6鈥�10 points). In this study, MAUD patients with an FTND score of 6 or above were classified as the comorbid tobacco use disorder (TUD) group, and those with a score of less than six were classified as the non-comorbid TUD group, using a cutoff value of 6.

(6) The Childhood Trauma Questionnaire Short Form (CTQ-SF) [22] was used to evaluate the experiences of abuse or neglect experienced by MAUD patients in their childhood before the age of 16. The questionnaire was developed by renowned clinical psychologist Bernstein in New York, USA, and is a globally recognized standard tool for measuring childhood trauma experiences. The CTQ-SF is a self-assessment scale with 28 items, measuring the following five dimensions: emotional abuse (EA: items 3, 8, 14, 18, 25), physical abuse (PA: items 9, 11, 12, 15, 17), sexual abuse (SA: items 20, 21, 23, 24, 27), emotional neglect (EN: items 5, 7, 13, 19, 28), and physical neglect (PN: items 1, 2, 4, 6, 26), with reverse scoring items included. For 2, 5, 7, 13, 19, 26, 28. Each dimension has 5 items, and an additional 3 items (items 10, 16, 22) are not included in the validity evaluation score. Each item uses the following 5-level Likert scoring system based on the frequency of occurrence of the experience: 1 point鈥�=鈥塶ever, 2 points鈥�=鈥塷ccasional, 3 points鈥�=鈥塷ccasional, 4 points鈥�=鈥塮requent, 5 points鈥�=鈥塧lways. Each subscale score ranges from 5 to 25 points, and the scale鈥檚 total score ranges from 25 to 125 points. The higher the score, the more severe the childhood trauma experience.

According to the retrospective self-assessment questionnaire manual developed by Bernstein [23], when emotional abuse is 鈮モ��3 points, physical abuse is 鈮モ��10 points, sexual abuse is 鈮モ��8 points, emotional neglect is 鈮モ��15 points, and physical neglect is 鈮モ��10 points in the subscale, as long as one subscale score meets the above conditions, it is classified as a person with childhood trauma; Those who simultaneously meet the criteria of emotional abuse鈥�<鈥�13 points, physical abuse鈥�<鈥�10 points, sexual abuse鈥�<鈥�8 points, emotional neglect鈥�<鈥�15 points, and physical neglect鈥�<鈥�10 points on the subscale are considered as those without any form of childhood trauma. The scale showed good internal consistency in the MAUD case and the healthy control questionnaire groups in this study, with an overall scale of Cronbach鈥檚 伪 coefficient of 0.88 and the subscale of Cronbach鈥檚 伪 coefficient scores ranging from 0.63 to 0.84.

(7) The Eysenck Personality Questionnaire Short Scale (EPQ-RS) was used to evaluate the personality characteristics of MAUD patients. The scale was jointly developed by Professor H. J. Eysenck and S. B. G. Eysenck and is a globally recognized standard tool for measuring personality traits and types [24]. The EPQ-RS is a self-evaluation scale with 48 items, measuring personality traits in three dimensions: extraversion (E-scale, positive scores: 3, 7, 11, 15, 19, 23, 32, 36, 41, 44, 48; reverse scores: 27); Neuroticism (N-scale, positive scores: 1, 5, 9, 13, 17, 21, 25, 30, 34, 38, 42, 46; negative scores: none); Psychoticism (P-scale, positive scores: 10, 14, 22, 31, 39; negative scores: 2, 6, 18, 26, 28, 35, 43). In addition, there is another dimension called the validity scale, which measures concealment (lie, L scale, positive scores: 4, 16, 45; reverse scores: 8, 12, 20, 24, 29, 33, 37, 40, 47), originally used as a validity scale for separate valid or invalid responses. A high L score indicates that the answer is not truthful and is invalid. However, later experience has shown that its score is related to many factors, not just authenticity.

The Chinese version of EPQ-RS was translated and revised by Chinese psychologists Qian Mingyi and Wu Guocheng [25], making it more in line with the psychological characteristics of Chinese people and forming the Eysenck Personality Questionnaire-Revised Short Scale for Chinese (EPQ-RSC), which includes 48 items (12 items for each subscale) and is suitable for adults aged 16 and above, with good reliability and validity. Each item is a true or false question, with 1 point for matching the reference answer and 0 points for not matching. The total score of each scale is calculated as the original score, which is then converted into a standard T score based on the national norm of age and gender: 50鈥�+鈥�10 x (original score - average score)鈥壝封�塻tandard deviation. The conversion score of each scale is between 0 and 100 points. A score below 38.5 on the E-scale indicates a typical introverted personality, a score above 61.5 indicates a typical extroverted personality, and a score between 38.5 and 61.5 indicates an intermediate and propensity type. A score below 38.5 on the N-scale indicates a typically stable personality, a score above 61.5 indicates a typically unstable personality, and a score between 38.5 and 61.5 indicates an intermediate and propensity type. A P-scale score of 61.5 or below indicates a normal personality, while a score of 61.5 or above indicates a psychotic personality. Each dimension can be explained separately and can also be combined with other dimensions for collaborative interpretation. EPQ-RSC showed good internal consistency in the MAUD case group of this study, as shown in Cronbach鈥檚 overall table 伪 coefficient, which is 0.80, and the subscale Cronbach鈥檚 伪 coefficient score range is 0.36鈥�0.78.

All survey questionnaires are filled out by the respondents themselves after the uniformly trained investigators explain the filling methods. The questionnaires are collected on the spot after completion.

Data quality control

First, before the formal survey, we conducted a pre-survey by separately selecting 20 MAUD patients and 20 healthy individuals to assess the applicability and acceptability of the questionnaire among the survey subjects. Based on the feedback from the pre-survey, necessary adjustments were made to the selection of the questionnaire. Second, during the formal survey, after each MAUD patient completed the questionnaire on-site, the investigators immediately verified and conducted quality control of the answers, checking whether the questionnaire was fully completed and whether the logic was consistent. Any issues were promptly addressed by supplementation or modification. Third, in terms of data processing, after the survey was completed, two quality control officers conducted uniform numbering and inspection. The data entry software used was Epidata 3.1, with double data entry and real-time proofreading to detect errors. If discrepancies were found between the two entries, the questionnaire was immediately returned for verification and correction. After data entry was completed, logical checks were performed on the variables to identify outliers, which were then promptly addressed.

Statistical analysis

Statistical analysis was conducted using SPSS 25.0 software. If the measurement data conforms to a normal distribution, use the x鈥壜扁�塻 description, t-test, and diversity comparison using univariate analysis of variance; For example, description of counting data, using 蠂²-test. Further adjust other covariates with statistical significance in univariate analysis and establish a multivariate generalized linear mixed model (inclusion criterion 伪鈥�=鈥�0.05, exclusion criterion 伪鈥�=鈥�0.10). Using multivariate generalized linear mixed model analysis to explore the risk factors for suicidal ideation in the research group, P鈥�<鈥�0.05 indicates a statistically significant difference.

Result

The situation of suicide ideation score of MAUD

In this study, 10,225 MAUD patients were measured for their suicidal ideation and scores in various dimensions using the Beck Scale for Suicide ideation (BSSI). The minimum score of the total scale is 0, the maximum score is 26, the average score is 2.59鈥壜扁��4.82, and the median is 0. The two dimensions of suicidal ideation are ranked in descending order based on the average score: suicidal tendency (1.49鈥壜扁��3.12) and suicidal ideation (1.11鈥壜扁��1.94). Because most patients have a total score of 0 for suicidal ideation, we further divided MAUD patients into suicidal ideation group and non-suicidal ideation group based on whether the sum of scores for items 4 and 5 is 0. Finally, 1228 MAUDs with suicidal ideation were included in the suicidal ideation group, and 8997 MAUDs without suicidal ideation were included in the non-suicidal ideation group. The detection rate of suicidal ideation was 12.01%.

Comparison of demographic characteristics between the suicidal ideation group and non-suicidal ideation group of MAUD

Using 蠂²-tests were conducted to compare the differences in general demographic data between the suicidal ideation group and non-suicidal ideation group of MAUD (Table听1). The results showed statistically significant differences (P鈥�<鈥�0.05) in age, per capita monthly income status, and family history of mental illness between the two groups. At the same time, there were no statistically significant differences (P鈥夆墺鈥�0.05) in gender, marital status, occupation, education level, only child, and caring by parents before adulthood between the two groups.

Table 1 Comparison of demographic characteristics between the suicidal ideation group and non-suicidal ideation group of MAUD

Comparison of characteristics of methamphetamine use between the suicidal ideation group and non-suicidal ideation group of MAUD

Using 蠂²-tests were conducted to compare the differences in characteristics of methamphetamine use between the suicidal ideation group and non-suicidal ideation group of MAUD (Table听2). The results showed statistically significant differences (P鈥�<鈥�0.05) in the dosage of each use of MA (g), frequency of use of MA in the last month, and co-occurring AUD between the two groups. At the same time, there were no statistically significant differences (P鈥夆墺鈥�0.05) in the age of onset of MAUD, first time use of MA brings a sense of pleasure, years of use of MA, Longest integrity period of MA (month), current craving of MA, severity of MAUD, and co-occurring TUD between the two groups.

Table 2 Comparison of characteristics of methamphetamine use between the suicidal ideation and non-suicidal ideation of MAUD

Comparison of family situation and childhood trauma experience between the suicidal ideation group and non-suicidal ideation group of MAUD

Using 蠂²-tests were conducted to compare the differences in family situation and childhood trauma experience between the suicidal ideation group and non-suicidal ideation group of MAUD (Table听3). The results showed statistically significant differences (P鈥�<鈥�0.05) in the parents committing crimes, physical abuse, and emotional neglect between the two groups. At the same time, there were no statistically significant differences (P鈥夆墺鈥�0.05) in family relationships, parents addicted to drugs and alcohol, emotional abuse, sexual abuse, and physical neglect between the two groups.

Table 3 Comparison of family situation and childhood trauma experience between the suicidal ideation group and non-suicidal ideation group of MAUD

Comparison of personality characteristics between the suicidal ideation group and non-suicidal ideation group of MAUD

Using 蠂²-tests were conducted to compare the differences in personality characteristics between the suicidal ideation group and non-suicidal ideation group of MAUD (Table听4; Fig.听1). The results showed statistically significant differences (P鈥�<鈥�0.05) in internal and external E and neuroticism N between the two groups. At the same time, there were no statistically significant differences (P鈥夆墺鈥�0.05) in Psychosis P between the two groups.

Table 4 Comparison of personality characteristics between the suicidal ideation group and non-suicidal ideation group of MAUD

Fig. 1

Comparison of f personality characteristics between the suicidal ideation group and non-suicidal ideation group of MAUD

Multivariate generalized linear mixed model analysis of risk factors for suicidal ideation among MAUD

Multivariate generalized linear mixed model analysis was conducted with MAUD having suicidal ideation as the dependent variable (0鈥�=鈥塶o, 1鈥�=鈥墆es). Further adjusting for other covariates with statistical significance in the univariate analysis, a multivariate generalized linear mixed model was established, including 30 independent variables (inclusion criterion 伪鈥�=鈥�0.05, exclusion criterion 伪鈥�=鈥�0.10). The results showed that the dosage of each use of MA鈥�>鈥�0.5, age鈥�>鈥�20, childhood physical abuse, childhood emotional neglect, the introversion or middle type, and neuroticism N dimensions in Eysenck personality characteristics were risk factors for the occurrence of suicidal ideation in MAUD. These are shown in Table听5; Fig.听2.

Table 5 Multivariate generalized linear mixed model analysis of risk factors for suicidal ideation among MAUD

Fig. 2

Multivariate generalized linear mixed model analysis of risk factors for suicidal ideation among MAUD

Discussion

In recent years, the problem of methamphetamine abuse has become increasingly serious, posing a serious threat to public health and safety in China. Relevant research has confirmed that long-term or high-dose use of methamphetamine can affect the social, cognitive, and executive functions of the human prefrontal lobe [26], weaken the brain鈥檚 ability to control, and lead to suicidal behavior by abusers.

The average total score of suicidal ideations measured by the Beck Scale for Suicidal Ideation (BSSI) among 10,225 MAUD patients in this study was 2.59鈥壜扁��4.82, with an average score of 1.49鈥壜扁��3.12 for the dimension of suicidal ideation and 1.11鈥壜扁��1.94 for the dimension of suicidal ideation. In addition, the detection rate of suicidal ideation was 12.01%. This research result is consistent with previous reports that substance abuse increases the risk of suicide [27]. The suicide risk caused by different substance abusers also varies [27], and MA users have a higher suicide risk than other substance abusers, which has also attracted the attention of researchers [7]. Compared to non-MA users, MA users have a 4.4 times higher risk of attempting suicide [8], while the standardized suicide mortality rate for MA-dependent patients is as high as 17 times [9]. A meta-analysis conducted by Stockings [10] on the mortality rate of MA users showed that compared to the general population, the all-cause standardized mortality rate of MAUD patients was 6.8, with the three main causes of death and standardized mortality rates being poisoning (24.7), suicide (12.2), and homicide (11.9), respectively. The risk of death from suicide in MA users increased by 12 times compared to the general population. One of the most found substances in blood or urine samples of suicide victims is MA, which also means that MA is a proximal risk factor for suicide [12]. MA users have reduced dopamine neurotransmission in their bodies [28], which plays an important role in suicidal behavior [29].

This study further explored and analyzed the factors influencing the total score of suicidal ideations. The results showed that the following six factors could independently affect the overall suicidal ideation risk of MAUD patients: the dosage of each use of MA, age, childhood physical abuse, childhood emotional neglect, the intrinsic and extrinsic E, and neuroticism N dimensions in Eysenck personality characteristics. The higher the dose of MA used each time, the higher the total score of suicidal ideations, which is reflected in the dimension of suicidal tendency. This is related to the fact that high-dose use of MA is more likely to lead to cognitive impairment in patients, resulting in mental and behavioral abnormalities caused by MA, which can induce suicidal ideation and even behavior. The higher the score in the internal and external dimensions of personality traits, the more inclined they are towards an extroverted personality, and the lower the risk of suicidal ideation. This is reflected in the dimension of suicidal ideation, indicating that an extroverted personality is a protective factor for suicidal ideation. The characteristics of an extroverted personality are confidence and being outgoing, energetic, talkative, and sociable. People with these traits are more likely to have good social networks and interpersonal relationships and, therefore, have more social and psychological resources to cope with adverse life events and are less likely to develop suicidal tendencies. The higher the score of neurotic personality in personality traits, the more inclined they are towards an unstable personality and the stronger their overall suicidal ideation. This is reflected in the dimension of suicidal ideation, which is consistent with the correlation between neurotic personality and various psychological and behavioral problems reported in previous studies [30]. Neurotic individuals are prone to emotional fluctuations and behavioral instability, making them more prone to excessive thinking, such as suicidal ideation, and even developing into suicidal behavior when encountering life events.

In this study, childhood physical abuse and childhood emotional neglect are also risk factors for higher suicide risk among MAUD patients. There is ample evidence to support the association between childhood trauma and suicide, with 33 -50% of suicide attempts attributed to childhood trauma experiences [31]. A recent meta-analysis suggests that childhood trauma experiences can increase the risk of adults attempting suicide by 2鈥�3 times [32]. The study by Lee [33] showed a dose-response relationship between childhood trauma events and suicide rates among MA users. MA plays a significant mediating role in the relationship between childhood trauma experiences and suicide, with indirect effects of 16% (2 childhood trauma experiences) and 42% (3 or more childhood trauma experiences), respectively. In addition, age is significantly positively correlated with suicidal ideation. The older the age, the higher the suicidal tendencies, which may be related to the fact that older individuals are more likely to experience suicidal tendencies due to experiencing greater life stress.

Limitations of this study. Firstly, representativeness of the sample. This study employed a peer-driven sampling method, selecting only MAUD patients from the drug abuse population in four sampled cities in Guangdong Province as research subjects. The study results may not be generalizable to MAUD patients in other regions. Future studies could consider conducting systematic sampling nationwide and carrying out multicenter investigations in different institutions to obtain a more representative sample while also making comparisons between different regions or centers. Secondly, this study is designed as a cross-sectional study, so in exploring the influencing factors, only the correlation between the independent variable and the dependent variable can be established, and its causal relationship cannot be established. In the future, longitudinal studies are needed to clarify further the causal relationship between the influencing factors and the outcome variable. Thirdly, when exploring the case-control study of MAUD suicidal ideation, this study was limited by the difficulty in collecting clinical samples and only included populations with methamphetamine abuse, lacking a normal control group. Future research should include a normal control group to assess more accurately the impact of various influencing factors on suicidal ideation in MAUD.

Conclusion

In summary, the results of this study suggest that MAUD patients generally have a higher risk of suicidal ideation. Therefore, when conducting suicide risk assessment and management for MAUD patients, special attention should be paid to those who take high doses of MA, have introverted and unstable personalities, have experienced childhood emotional neglect, and have been subjected to childhood physical abuse. Timely detection and intervention measures should be taken to prevent suicide risks.