The influence of parenting style and coping behavior on nonsuicidal self-injury behavior in different genders based on path analysis

Subjects

A cross-sectional survey was carried out in Ningbo City, China from September to October 2021. The study population was derived from three junior middle schools and three senior high schools in four districts of Ningbo City (Haishu, Jiangbei, Gaoxin, and Yinzhou). Specific investigation process: students from grade 1 to grade 3 of junior middle school and grade 1 to grade 2 of senior high school were randomly selected. After a trained psychiatrist reads out the instructions, all students complete the questionnaire on-site within 40 min.

We obtained the written informed consent of all participants and their parents before the study, including the potential trigger nature of all relevant issues specifically mentioned, as well as confidentiality restrictions related to mandatory reporting concerns. The study was approved by the ethics committee of Ningbo Kangning hospital (Ethical Application Ref. No.: nbknyy-2020-Lc-52), which was approved in October 2020 and is valid until October 2023. All research investigation methods carried out in this work were conducted in accordance with the relevant guidelines and regulations in the declaration of Helsinki.

Survey content

Data on demographic characteristics and conditions considered to be related to NSSI were collected (Lin et al., 2017; Wan et al., 2015). This included age, gender, physical condition, academic achievement, peer relationship, family relationship, family economic situation, parental education level, and other factors. In particular, the revised coping style scale for middle school students and the parental rearing style evaluation scale (EMBU) jointly prepared by C. Perris, Department of Psychiatry, Umea University, Sweden, were used to effectively evaluate the individual and interpersonal factors of the subjects.

Quality control

To ensure the reliability and consistency of the questionnaire results, psychiatrists administered the test to participants. A pre-test was conducted before the survey. The purpose of the survey was explained on-site. After the test, the doctors of the main test will collect the questionnaire, enter and screen the questionnaires from two people, and then conduct a logical check on the questionnaire to ensure the quality of the answers. A total of 2,784 questionnaires were administered and 2,753 questionnaires were recovered, with a recovery rate of 98.89%. Among them, 64 questionnaires had >30% of missing data and logical errors; thus were recorded as invalid questionnaires. There were 2,689 valid questionnaires, and the effective rate was 96.59%.

Data analysis method

After logic checking and proofreading, Epidata 3.1 Software was used for data entry. The scores of NSSI behavior, parental rearing style, and coping style of male and female students were compared by independent sample t-test; The count data were described by the number of cases (constituent ratio) method. Comparisons between groups were made using the ${\chi }^2$ test. All analyses were conducted with SPSS software, version 26.0.

Single-factor analysis

A total of 2,689 participants were investigated, among which 435 participants had NSSI behavior, and the positive rate was 15.16%. The numbers of participants by age group were: 11–13 years old: 817 (97 in the NSSI group, 11.87%); 14–16 years old: 1,872 (338 in the NSSI group, 18.06%). Gender characteristics of the sample: male: 1,157 (78 in NSSI group, 6.74%); female: 1,532 (357 in the NSSI group, 23.30%). It was found that the age and sex of middle school students were statistically significant (P < 0.05).

Construction of path model and the mediating effects of NSSI behavior

In our previous article, based on the basic characteristics of the sample, parental rearing style, and middle school students’ coping behavior, we carried out a single factor analysis with the NSSI behavior/gender of the sample respectively. The screening excluded the variables with weak statistical significance between parents’ investment in children and parents’ expectations for children and NSSI, as well as the variables with weak statistical significance between seeking social support and patience and gender differences. However, the single factor analysis of parameters is a relatively simple causal relationship model. The relationship between independent variables and dependent variables is one-to-one corresponding, and there is no form of interaction between variables. NSSI behavior is a complex individual behavior under the cross-influence of multiple factors, such as social relations, individual psychology, and behavioral coping style. The single factor analysis method is difficult to be used in deeply exploring the interaction and intermediary effects between the influencing factors of NSSI.

Therefore, in this work, we propose a path distribution network of NSSI influencing factors based on a path analysis model. At the same time, the construction of the intermediary effect of NSSI endogenous derivative factors at the level of genders will be discussed in depth. To effectively understand the complex transmission process between variables, we further explored the relationship between external derivative variables and endogenous derivative variables for this complex behavior under the influencing factor network of NSSI.

Construction of path model for NSSI behavior

Although the direct path between exogenous variables and NSSI is significant, there is also evidence that exogenous variables affect the occurrence of NSSI through endogenous variables, such as stressful life events, negative self-evaluation, drug use, and suicide attempts. In the previous section, we completed a series of screening on the influencing factors with significant correlation and obtained the hint of a mediating effect between endogenous derivative variables. The above results were then incorporated into the subsequent path model construction to further explore the complex network of the intersection of external derivative variables and endogenous derivative variables of NSSI behavior. The construction results of the path network model are shown in Fig. 1. According to Hu & Bentler (1999), several fitting indexes were used to evaluate the fitting effect of the model, including the comparative fitting index (CFI), the goodness of fit index (GFI), approximate root mean square error (SRMR) and weighted root mean square residual (RMSEA). Good model fitting is represented by the CFI value greater than 0.9, GFI value greater than 0.9, SRMR value less than 0.1, and RMSEA value less than 0.1. The model data fit well; the indicators of female and male part are: CFI = 0.913/0.923, GFI = 0.964/0.977, SRMR = 0.055/0.047, RMSEA = 0.097/0.069 with 90% confidence interval (CI) [0.084–0.111)/[0.054–0.084].

As shown in Fig. 1, the path relationship of each factor in the model is drawn accordingly. For females, (1) when radical education, peer relationship, academic achievement and physical condition affect negative coping style respectively, the standardized path coefficient values are 0.510 (z = 17.454, P = 0.000 < 0.001), −0.170 (z = −7.814, P < 0.001), −0.096 (z = −4.425, P < 0.001), −0.053 (z = −2.436, P < 0.01), which shows that radical education has a significant positive impact on negative coping style. Additionally, peer relationship, academic achievement, and physical condition have a significant negative impact on negative coping style. (2) When family harmony and family conditions have an impact on radical education, the standardized path coefficient values are −0.173 (z = −9.542, P < 0.001) and −0.138 (z = 7.554, P < 0.001), respectively. Therefore, it shows that family harmony will have a significant negative impact on radical education, and family conditions will have a significant positive impact on radical education. (3) When parents’ education level affects family conditions, peer relationship, academic performance, and physical condition, the standardized path coefficient values are 0.148 (z = 5.862, P < 0.001), 0.236 (z = 9.518, P < 0.001), −0.068 (z = −3.227, P < 0.001) and 0.166 (z = 6.588, P < 0.001) respectively, indicating that parents’ education level affects family conditions, peer relationship, and physical conditions significantly. Additionally, parents’ education level will have a significant negative impact on learning. (4) When negative coping style and radical education affect NSSI respectively, the standardized path coefficient values are 0.481 (z = 20.636, P < 0.001) and 0.195 (z = 6.524, P < 0.01) respectively, which indicates that negative coping style and radical education have a significant positive impact on NSSI.

For males, (1) when radical education, peer relationship, academic achievement, and physical condition affect negative coping style, the standardized path coefficient values are 0.473 (z = 18.326, P < 0.001), −0.008 (z = −0.291, P = 0.771 > 0.05), −0.051 (z = −1.965, P = 0.049 < 0.05), and −0.048 (z = −1.847, P = 0.065 > 0.05) respectively, which shows that radical education will have a significant positive impact on negative coping style, learning will have a significant negative impact on negative coping style, while peer relationship and physical condition will not have an impact on negative copying style. (2) When family harmony and family conditions affect radical education respectively, the paths are not significant (z = 0.979, P = 0.328 > 0.05), (z = −0.798, P = 0.425 > 0.05). Therefore, it shows that family harmony and family conditions do not affect radical education. (3) When parents’ education level affects family conditions, peer relationship, academic performance and physical condition, the standardized path coefficient values are 0.289 (z = 10.280, P < 0.001), 0.123 (z = 4.219, P < 0.001), −0.175 (z = −6.048, P = 0.000 < 0.01), and 0.079 (z = 2.679, P = 0.007 < 0.01) respectively, indicating that parents’ education level will affect family conditions respectively, peer relationship and physical condition have a significant positive impact, and parents’ education level will have a significant negative impact on learning. (4) When negative coping style and radical education affect NSSI respectively, the standardized path coefficient values are 0.225 (z = 7.057, P < 0.001) and 0.104 (z = 3.262, P < 0.001), indicating that both negative coping style and radical education will have a significant positive impact on NSSI.

Mediating effect of coping style in parental rearing style and NSSI

In the relationship between the independent variable and dependent variable, if the independent variable indirectly affects the dependent variable through a third variable, this behavior becomes an intermediary effect behavior. Baron & Kenny (1986) put forward the complete conceptual system and test procedure of an intermediary effect from the perspective of statistical methodology, which is generally accepted by researchers in the field of social sciences.

Based on the path model constructed above, this work will further discuss the mediating effects of endogenous variables such as parental radical education and negative coping styles, on NSSI behavior. Figure 2 shows a schematic diagram of the mediation model of NCS in the relationship between RE and NSSI. As shown in Fig. 2, the mediation model with negative coping methods as the intermediary variable, parental radical education as the independent variable, and NSSI as the dependent variable component are significant for the girls in the study. The indirect impact of parental radical education is 0.109 (SE = 0.016, 95% CI [0.172–0.210]). In addition, before using negative coping methods as media, the regression coefficient of the direct relationship between parental radical education and NSSI was significant (B = 0.191, 95% CI [0.172–0.210]). When negative coping methods were added to the analysis, the correlation between parental radical education and NSSI was still significant (B = 0.082, 95% CI [0.062–0.102]). This shows that negative coping methods play a partial mediating role in the relationship between parental radical education and NSSI. Negative coping methods accounted for 57.2% of the total impact of radical education on NSSI. See Table 2 for specific parameters. The intermediary model is also significant for the boys in the study. The indirect impact of parental radical education is 0.027 (SE = 0.016, 95% CI [0.076–0.140]). In addition, before using negative coping methods as media, the regression coefficient of the direct relationship between parental radical education and NSSI was significant (B = 0.053, 95% CI [0.039–0.067]). When negative coping methods were added to the analysis, the correlation between parental radical education and NSSI was still significant (B = 0.026, 95% CI [0.010–0.042]), indicating that negative coping methods played a partial intermediary role in the relationship between radical education and NSSI. Negative coping methods account for 50.872% of the total impact of parental radical education on NSSI. See Table 5 for specific parameters.

Table 5:

Summary of mediation test results.

	Test conclusion	c	a	b	a*b	c′	a*b (95% BootCI)	Effect proportion (a*b/c)
Female
RE→ NCS→NSSI	Partial intermediary	0.191 ***	0.534 ***	0.205 ***	0.109	0.082 ***	0.229–0.290	57.191%
Male
RE→ NCS→NSSI	Partial intermediary	0.053 ***	0.477 ***	0.057 ***	0.027	0.026 **	0.076–0.140	50.879%

DOI: 10.7717/peerj.14507/table-5

Note:

** p < 0.01.

*** p < 0.001.

In Table 5, c represents the regression coefficient of independent variable x to dependent variable y (when there is no intermediary variable m in the model), that is, the total effect. a represents the regression coefficient of x to m, b represents the regression coefficient of m to y, and a * b is the product of a and b, that is, the intermediary effect. c′ represents the regression coefficient of x to y (when there is an intermediary variable m in the model), that is, the direct effect. 95% BootCI refers to the 95% confidence interval calculated by bootstrap sampling. If the interval does not include 0, it indicates that the intermediate effect is significant.

As shown in Table 5, regulated by gender differences, after introducing the intermediary variable m (i.e., negative coping style), the direct effect (i.e., regression coefficient) of radical education on NSSI shows a downward trend (female: 0.191, 0.082; male: 0.053, 0.026), which shows that the relationship between RE and NSSI is further constructed through the intermediary variable negative coping style. The effects of women and men accounted for 57.191% and 50.879%, both of which were partial mediating effects. At the same time, the 95% confidence intervals calculated by bootstrap sampling do not cover 0, which verifies that our model has a significant mediating effect.

Parenting style and NSSI

Significance test analysis was conducted using an independent sample t-test on the parenting style of the participants on NSSI behavior and gender respectively (Table 2). It was found that the parenting styles of the adolescents were extreme and contradictory: parents treat their children with excessive severity on the one hand and excessive care on the other, which is consistent with the study by Zuo, Xi & Sang (2004). Therefore, this type of parenting style is classified as “radical education” in this study. The boundaries of “radical education” are unclear, extreme, and full of contradictions, which on the one hand directly increases children’s stress and makes them more likely to develop tension and on the other hand, will impair children’s emotional control, which will make children have a negative tendency to cope with stress. Together, these two pathways put children at increased risk of developing NSSI. In contrast, emotional warmth and understanding shown by parents can increase their children’s psychological and emotional resilience (Guerreiro et al., 2013), thus reducing the risk of their children developing NSSI. At the same time, studies combining family environment factors (see Table 1) observed that families with NSSI also generally have higher levels of conflict and lower levels of relaxed recreation (Domínguez-Baleón et al., 2018; Liu et al., 2020). In the traditional Chinese parenting concept, parents have a relatively absolute say in the family, and their parenting style has an inevitable and profound impact on the physical and mental health of their children (Gu, Ma & Xia, 2020; Serafini et al., 2017). When a child’s behavior or emotions are problematic, the response of the parents often determines the outcome. If parents adopt a punitive parenting style and emotionally deny and reject, disruptions in the child’s cognitive ability may happen (Masud et al., 2019). At the same time, from a physiological point of view: the disruption of the level of secretion of transmitters such as norepinephrine in the adolescent’s body during the reception of negative information input makes it impossible to inhibit the response to the stimulus until it is adequately evaluated, i.e., the rapid response impulse, which eventually leads to risky behavior (Swann et al., 2013; Francis et al., 1999) and even to the development of NSSI. In particular, we found an interesting phenomenon: gender differences present a clear difference in the significance level of the effects of parenting styles. Both positive parenting style factor scores and negative parenting behavior factor scores were greater for boys than for girls, implying that compared to girls, boys faced more positive parenting behaviors in family relationships while suffering more negative parenting styles, including punitive severity and denial. In China, there is still a certain degree of “son preference” in education, and boys are often expected to grow up with more responsibility, so they are more involved in family education. In the initial stage of family education, positive inductive education such as emotional warmth and understanding is often reflected (Liu, Su & Yin, 2022) but when there is a certain gap between expectations and reality or when the child shows resistant behavior, such caring education will gradually transform into radical education in the form of severe punishment, denial, and excessive interference, to further realize the parents’ expectations. At the same time, in some Chinese families with strong traditional values, the parenting style of girls does not have this contradictory situation of “high expectations and high strictness” (Hannum, Kong & Zhang, 2009; Ong et al., 2018). In conclusion, the gender salience of parenting styles is closely related to the unique Chinese social context. An in-depth analysis of this phenomenon provides a theoretical basis for effectively avoiding gender-specific NSSI behaviors. In conclusion, the gender salience of parenting styles is closely related to the unique Chinese social context.

Coping style and NSSI

As shown in Table 3, we found that positive problem-oriented coping styles (seeking social support, positive rationalization) were protective factors for NSSI while negative emotion-oriented coping styles (tolerance, avoidance, fantasy denial) were risk factors for NSSI. In a review, Fu et al. (2018) similarly found that people with NSSI behaviors were more inclined to fantasy avoidance, self-blame, venting, patience, substance abuse, and other negative coping styles directed toward emotions. In this study, NSSI behaviors were highly correlated with the coping styles of secondary school students, and the NSSI group tended to adopt patience, avoidance, and denial in coping with individual and external stressors, which tended to cause the accumulation of negative emotions within and ultimately lead to poor outcomes.

Interestingly, there are obvious differences in the handling of stress as well as negative emotions among different genders: boys are more inclined to adopt a positive confrontation and reasonable solution, while girls are more inclined to make coping behaviors that are more emotional such as avoidance, venting emotions, and fantasy denial. For a long time, with the development of social productivity and the emergence of male prominence in the production sector, men began to dominate, i.e., patriarchal society gradually became the mainstream in the development of human history nowadays. The patriarchal culture of “male superiority and female inferiority” has had a profound impact on the difference in social status between men and women, resulting in men being given or expected to have personality qualities such as strength, independence, and extroversion, while women are usually given or expected to have personality qualities such as demureness and introversion (Zosuls et al., 2011). Established research on personality and gender roles has shown that masculinity scores have a strong positive correlation with the extraversion dimension, and femininity scores have a strong positive correlation with the interpersonal dimension, with cross-cultural consistency (Schmitt et al., 2017). Therefore, in terms of emotion regulation strategies, girls adopt emotion-focused strategies while boys tend to use cognitive reappraisal strategies. The difference in strategies may lead to less effective regulation in girls than in boys, which can easily trigger adverse emotions such as anxiety and depression and lead to more psychological problems and more self-injurious behaviors.

Exogenous and endogenous derivative variables on NSSI

NSSI is in essence a maladaptive coping strategy that can be interpreted as a way of regulating tension. Studies have shown (Xu et al., 2021) that NSSI behaviors are not directly caused by a single factor, but are often influenced by the interaction of social and environmental factors, family factors, and personal factors in the process of their formation. In particular, there seems to be an interesting cross-influence phenomenon among the influencing factors, and this cross-influence relationship is accompanied by either indirect or direct extrapolation and endogeneity (i.e., the role of exogenous and endogenous variables). Ecosystem theory suggests that personality traits, endogenous variables, and exogenous environmental variables jointly influence adolescents’ regulation (Panlilio, Jones Harden & Harring, 2018). That is, emotion regulation is the result of the interaction between the individual and the external environment. Parenting styles and adolescents’ coping styles do not seem to be independent events: children with positive parenting styles receive more support and trust when facing problems, have more courage to face life challenges, apply resources positively, solve problems, and thus are better able to regulate stressful emotions under pressure. On the contrary, children with a negative parenting style often have a sense of mistrust, are not confident enough to face problems, and often exhibit negative coping behaviors such as procrastination and social detachment, which lead to psychological problems such as obvious NSSI symptoms.

To this end, this work constructs a pathway analysis model of multiple influences on NSSI behavior to effectively explore the extrapolative and endogenous variables among the influences on this complex behavior. In terms of the influence of endogenous variables on NSSI, we found that negative coping styles mediated the association between radical education and NSSI behaviors in the pathway model. Positive parenting with warm and understanding parents is a major contributor to the development of high self-esteem, effective emotion regulation management, and problem-oriented positive coping styles in adolescents, while negative and radical parenting styles such as harsh punishment, denial, and excessive control are responsible for the development of immature and negative coping styles in adolescents, thus making them more likely to trigger NSSI behaviors in the face of external stimuli (Liu et al., 2021; Peng, Fu & Zhang, 2020). On the other hand, adolescents who are exposed to long-term negative parenting styles often have difficulties in receiving effective communication and help from their parents, and tend to adopt more negative coping styles that gravitate towards emotions (patience, avoidance, venting emotions, fantasy/denial) when dealing with stress alone, which gradually evolve into some kind of self-harm phenomenon (i.e., NSSI behavior) when these coping styles are not effective. Therefore, it is necessary to adopt positive (warmth, understanding) parenting styles and suppress negative (punishment, harshness, rejection, denial) parenting styles to reduce the immature coping styles of middle school students.

As shown in Fig. 2, the contribution factor of the mediating effect in gender differences showed a significant difference: in the face of radical family education, females were more likely to adopt a negative coping style for diversion (female effect coefficient: 0.534, P < 0.05; male effect coefficient: 0.477, P < 0.05); at the same time, negative coping style showed a more significant influence on females’ NSSI behavior compared to that of males (female effect coefficient: 0.205, P < 0.05; male effect coefficient: 0.057, P < 0.05). The role of negative coping styles showed a more significant effect on women’s NSSI behavior compared to men’s mediation effect (female effect coefficient: 0.205, P < 0.05; male effect coefficient: 0.057, P < 0.05). Although the current findings advanced the understanding that the existence of gender differences in NSSI, the mechanism of gender differences’ influence on NSSI behavior is currently less studied. It has been found that the gendered socialization of emotions, e.g., shame and anger, may influence the types of emotions experienced by men and women, leading to women being more likely to engage in NSSI (Domes et al., 2010; Buchanan et al., 2010), and that this gendered socialization of emotions is often closely related due to the differential levels of perceptions of the status roles of men and women in society at large. At the same time, biological factors such as hormonal differences between males and females (e.g., testosterone and estradiol) also seem to be involved in gender-influenced relationships in NSSI.

Here, a noteworthy finding was that females used emotionally directed negative coping more frequently than boys in conflict resolution and scored higher on all behaviors of negative coping, while good peer relationships had a significant protective effect on girls. The effect of individual factors on boys’ coping styles was not significant. In addition, girls’ negative coping behaviors were more influenced by parenting styles than were boys’. These results suggest that girls may be more prone to mood swings in the face of stressful conflicts, leading them to consistently use more negative coping styles than boys to the point of developing more mental health problems, which may imply a greater risk of developing NSSI. Another noteworthy finding is that the way girls are educated is significantly influenced by their family environment; a harmonious family atmosphere has a significant negative effect on aggressive parenting styles, and good family conditions have a positive effect, presumably related to a higher per capita monthly household income and more intrusive parental control over children.

From the perspective of extrapolation quantity, the parental education level influences the NSSI phenomenon in various ways, and it affects both individual and environmental factors of the child, and ultimately affects the generation of NSSI behavior from different paths, respectively. Therefore, in general, only secondary school students whose parents’ education level is elementary school have the highest incidence of NSSI, and as the parental education level increases, the incidence of NSSI decreases. However, the incidence of NSSI increases for parents with university education level and above; parents with low education level (elementary school) are likely to use more negative parenting styles such as harsh punishment and denial, while parents with higher education put high demands on their children and use more negative parenting styles such as excessive interference and control (Zhang, Cheng & Xiao, 2014), both of which can impair children’s development of mature, positive coping styles; socioemotional development, and emotional management skills. In addition, there is a positive relationship between parental education and family income, and family socioeconomic status is a reflection of the level of resources available to the individual and the family. Individuals with low family socioeconomic status not only have higher levels of economic stress but also face more uncontrollable stress. They also face more uncontrollable negative life events and persistent stressful experiences. Society also has negative perceptions and judgments of individuals with low family socioeconomic status or low income (e.g., they are generally perceived as lazy and unmotivated), and these negative stereotypes or prejudices are risk factors for individuals’ self-esteem levels. At the same time, these stereotypes or prejudices can also cause individuals of low family socioeconomic status to have more experiences of social exclusion (i.e., being ignored or rejected by others), which is not conducive to children forming good peer relationships. However, parental education showed an inverse relationship with academic achievement, which is different from the conventional perception that highly qualified parents place more importance on their children’s education but is consistent with the phenomenon of “poor students” in China. Students whose families have a weaker financial base are likely to make more efforts to change their status quo, and studying is the best way for them to do so. However, in this state, they are also prone to develop a focused and withdrawn personality, which is not conducive to emotional regulation and stress release, thus increasing the risk of NSSI.