Introduction

Mental health problems are as prevalent among university students as they are in their non-university-attending same age peers ( ^{Blanco et al., 2008} ). In a study of surveys carried out in 21 countries, an estimated 20% of university students met criteria for a mental disorder ( ^{Auerbach et al., 2016} ). In the case of the Mexican university population, ^{Melo-Carrillo, Van Oudenhove and López-Avila (2012)} found that among Mexican medical students 36.29% of the students scored positive for depressive symptoms. In another study, there was a significant increase in the consumption of both alcohol and tobacco in transitioning from secondary school to university ( ^{Alvear-Galindo et al., 2015} ). Other mental health problems that have been researched among Mexican university students are eating disorders, suicide ideation, drinking problems, and post-traumatic stress disorder ( ^{Chávez-Rosales, Camacho, Maya, & Márquez, 2012} ; ^{Díaz et al., 2008} ; ^{Mendoza, Márquez, Guadarrama, & Ramos, 2013} ; ^{Moreno & Ortiz, 2009} ; ^{Rosales, Córdova, & Ramos, 2012} ). However, as far as we know, non-suicidal self-injuries (NSSI) have not been studied in the Mexican university population.

NSSI, conceptualized as the deliberate destruction or alteration of body tissue without the intent to die ( ^{Klonsky, 2007} ), like cutting or burning oneself, is a growing public health problem worldwide. NSSI are used for a variety of functions or motivations, like affect-regulation or self-punishment ( ^{Klonsky, 2007} ). It has also been hypothesized that the brain structures involved in processing the cognitive and emotional components of pain are altered in those that engage in NSSI indicated by an anomaly in the integration of pain as a conscious experience conceived of as unpleasant ( ^{Mendoza & Pellicer, 2002} ). Most research in this area has been conducted with adolescents ( ^{Jutengren, Kerr, & Stattin, 2011} ; ^{Whitlock, Eckenrode, & Silverman, 2006} ; ^{You, Leung, & Fu, 2011} ) and university students ( ^{Tsypes, Lane, Paul, & Whitlock, 2016} ; ^{Whitlock et al., 2011} ; ^{Yurkowski et al., 2015} ). The international prevalence in young adults (from Asia, Australia/New Zealand, Canada, Europe, United Kingdom, and the USA) is 13.4% ( ^{Swannell, Martin, Page, Hasking, & St John, 2014} ). NSSI can have negative consequences, like scars and infections, guilt, shame, and possible rejection of parents and peers ( ^{Wilkinson & Goodyer, 2011} ). Also, high rates of NSSI have been associated with severe psychopathology, such as depression, anxiety, borderline personality, drug abuse, and eating disorders as well as an increased risk of suicidal attempts ( ^{Bresin & Gordon, 2013} ; ^{Klonsky & Olino, 2008} ). Some studies have reported that few people seek professional help for NSSI, which limits the possibility of intervention ( ^{Whitlock et al., 2006} ).

Some of the research on college students has found relations between NSSI and different psychological variables. ^{Yurkowski et al. (2015)} found that the quality of parent-child relationships had a more relevant impact on prediction of NSSI than the quality of peer relations, and that the feelings of alienation of both parents and peers had indirect effects on NSSI through deficits of emotion regulation. Another study in university students found that there is a continuum between NSSI and suicidal self-injury, and the differences are in degree rather than in kind ( ^{Orlando, Broman-Fulks, Whitlock, Curtin, & Michael, 2015} ). ^{Whitlock et al. (2011)} report that in a sample of college students the lifetime prevalence of NSSI was 15.3%. Females were more likely to self-injure because they were upset or in the hopes that somebody would notice them, whereas males were more likely to report anger or intoxication as a motivation. Also, disclosure of NSSI was low among both sexes, because only 8.9% disclosed their self-injuries to a mental health professional. ^{Kharsati and Bhola (2014)} found that the most common method of NSSI was self-hitting (15.2%), followed by cutting or carving skin (13.2%). Self-efficacy has been found to be related to depressive and anxiety symptoms and in one study to mediate the relation of stressful life events with depression and anxiety symptoms ( ^{Maciejewski, Prigerson, & Mazure, 2000} ; ^{Muris, 2002} ). We therefore consider that self-efficacy could be a relevant variable in the study of NSSI.

Currently, there is limited data from developing countries, which hinders the efforts to have a broader picture of the problem and coordinate efforts to address this health issue ( ^{Muehlenkamp, Claes, Havertape, & Plener, 2012} ). There is scarce information from Latin America in general and Mexico in particular; in Mexican adolescents some studies have found prevalences between 5.6 to 17.1% ( ^{Albores-Gallo et al., 2014} ; ^{González-Forteza et al., 2005} ). One study in a clinical sample with adolescents found that self-injury was present in high frequencies in female patients, with low socioeconomic status and a history of family violence ( ^{Ulloa, Contreras, Paniagua, & Figueroa, 2013} ). Currently we do not have any estimates of the prevalence of NSSI in Mexico in the general population of young adults, or in the college population in particular. It is important to have more information about how NSSI are present in the Mexican population in order to guide relevant strategies of intervention.

Therefore, the objective of the current investigation is to evaluate a sample of university students in Mexico City, a group for which there is no prior data that we are aware of with regards to NSSI, in terms of frequency, types of self-injury, gender differences, and correlations with psychological variables like depressive symptoms, anxiety symptoms, impulsivity, self-efficacy, and emotional regulation.

Method

Results

As shown in Table 1, the sample contained more female subjects (76.5%), most were psychology students (85.4%), and were largely young adults between 20 and 26 years of age (56.4%).

Table 1. Socio-demographic characteristics of the sample, 2012-2013 

Note: There were 583 students, but 19 students’ didn´t report their sex.

With respect to NSSI, during the prior 12 months, 10.9% of the sample self-injured between 1 and 50 times. The age of onset for NSSI among those with at least one self-injury was: 18.5% before the age of 11, 74.1% between 11 and 17 years of age, and 7.4% after 17 years of age. The frequencies of lifetime NSSI are shown in Table 2. Of the total sample, 42.1% never had any self-injury, whereas 30.9% experimented with NSSI at some point in their life and 26.9% engaged in the behavior recurrently. The most frequent NSSI behaviors in the sample among those who self-injure (Table 3) were self-cutting (48.0%), carving words into skin (44.6%), severe scratching (39.1%), punching self (39.1%), sticking sharp objects into skin (37.6%), and carving pictures into skin (33%). While there were no gender differences in overall self-injurious behaviors, more females engaged in self-cutting than males (χ² = 7.13; p < .01) and more males engaged in burning themselves with cigarettes than females (χ² = 6.65; p < .01).

Table 2. Frequencies of NSSI in the total sample and by sex, 2012-2013 

Table 3. Frequencies of specific NSSI behaviors among those who self-injure by sex, 2012-2013  

Table 4 shows the means of the psychological characteristics (depressive symptoms, anxious symptoms, impulsivity, self-efficacy, cognitive reappraisal, and expressive suppression) by NSSI group and sex. The recurrent group had the highest levels of depressive symptomatology, anxiety, and impulsivity. Females had higher levels of depression and anxiety than males in all groups. Results of the ANOVAs are also presented on Table 4 showing a significant group effect for depressive symptoms, anxiety, impulsivity, and self-efficacy, a gender effect for depressive symptoms, anxiety, and self-efficacy, and an interaction effect for depressive symptoms only. Post hoc analyses reveal that those with recurrent NSSI had more elevated levels of anxiety, depression, and impulsivity and lower self-efficacy compared with those with experimental NSSI and those without NSSI. There were no significant differences between the experimental and non-NSSI groups for these characteristics. With regards to the interaction effect for depressive symptoms, only females, but not males in the recurrent group had higher levels of depressive symptoms. The three groups exhibited no significant differences in emotional regulation (either cognitive reappraisal or expressive suppression).

Table 4. Means of psychological factors by sex and  self-injury category, 2012-2013 

To estimate the odds of being in the experimental or recurrent groups of NSSI given high rates in the psychological correlates, a multivariate logistic regression analysis was conducted (Table 5). Those without NSSI were the comparison group for odds ratios (ORs). Those with elevated levels of depressive symptomatology had 1.7 times the odds of experimental self-injury (95% CI [.09, 2.76]) and 2.2 times of recurrent self-injury (95% CI [1.34, 3.60]). Subjects with high impulsivity were 2.78 times more likely to have engaged in recurrent self-injury (95% CI [1.74, 4.45]), while those with high self-efficacy had half the odds of recurrent self-injury (95% CI [.30, .78]). No significant associations for emotional regulation and anxiety were found.

Table 5. Results of a multivariate logistic regression analysis to evaluate the odds of experimental and recurrent self-injury given high levels of depression, anxiety, impulsivity, expressive suppression and cognitive reappraisal 

Note: OR = Odds Ratio; 95% CI = 95% Confidence Interval.

Finally, we evaluated, the perception of need for psychological or psychiatric attention, actually seeking psychological or psychiatric attention, and hospitalization or medical attention for NSSI (Table 6). While 54.4% of those who engaged in self-injury perceived the need for psychological or psychiatric help, only 18.1% sought psychological or psychiatric attention, and 53.1% required hospitalization or medical attention. A greater proportion of those with recurrent NSSI sought psychological or psychiatric attention than those with experimental NSSI, whereas there was no difference between the recurrent and experimental groups with regards to perceived need or required hospitalization or medical attention.

Table 6. Perceived need, help seeking and medical attention in experimental, recurrent and all self-injurers 

Note: a) In this table only the yes answer is shown. Of those who self-injured, 88 did not answer these questions; b) In this table only the yes answer is shown. Of those who self-injured, 84 did not answer this question.

Discussion and conclusion

We found high lifetime frequencies of NSSI in this sample of Mexican university students, with 57.8% having reported at least one occasion of self-injury. This lifetime estimate is higher than others reported in college samples from other countries that range from 7-35% ( ^{Duggan, Toste, & Heath, 2012} ; ^{Gratz, 2001} ; ^{Muehlenkamp et al., 2012} ; ^{Whitlock & Knox, 2009} ). These studies in general have few or no Mexicans in their samples.

Given the emphasis on adolescent samples in much of the prior literature, it is interesting to note that the proportion of these emerging adults with NSSI in the prior 12 months, only 10.9%, was much lower than the lifetime proportion. These 12-month estimates are in the range found in other countries of 7.3% to 45.5% during the prior 12 months ( ^{Duggan et al., 2012} ; ^{Muehlenkamp et al., 2012} ; ^{Whitlock et al., 2006} ). This suggests that while a greater proportion of adolescents engage in NSSI than young adults, this continues to be a problem in emerging adulthood. This is further evidenced by the reported ages of onset. While most university students with a history of NSSI (74.1%) began to self-injure in adolescence (between 11 and 17 years of age), less (18.5%) reported initiating in childhood (before age 11) and even fewer (7.4%) initiated as young adults (18 years or older). As was the case in other studies, the most common self-injury behavior was cutting ( ^{Claes, Klonsky, Muehlenkamp, Kuppens, & Vandereycken, 2010} ; ^{Klonsky, 2011} ).

Some have suggested that the number of persons that self-injure has risen in the last decade, especially among adolescents and young adults, although this may be due to greater attention on this behavior from the media and researchers ( ^{Auerbach et al., 2016} ; ^{Klonsky, 2007} ). Multiple factors may account for the elevated rates of NSSI among young adults. In Western societies, there is an increase in the acceptance of tattoos and piercings, and even if these behaviors are not considered NSSI, body modification, along with chats and web sites dedicated to self-injury, may have an impact upon the “normalization” of self-injury behavior ( ^{Favazza, 2009} ). Peer learning and the media can be relevant for the increase of NSSI, because the number of references to self-injury has increased in songs, movies, press, and internet during the last decade ( ^{Favazza, 2009} ). Internet sites that share or promote experiences of NSSI have risen and this could influence emotionally vulnerable young adults ( ^{Favazza, 2009} ; ^{Whitlock et al., 2006} ). The increased exposure to NSSI may increase the probability for self-injury in such a way that social influences in college students should be taken into account ( ^{Muehlenkamp, Hoff, Licht, Azure, & Hasenzahl, 2008} ). Multiple factors can affect the vulnerability of young adults, from emotional regulation problems, body image, depression, impulsivity, child abuse, anxiety, substance abuse, borderline personality disorder, and suicide attempts among other factors ( ^{Brown, 2009} ; ^{Duggan et al., 2012} ; ^{Klonsky, 2007} ; ^{Klonsky & Olino, 2008} ; ^{MacLaren & Best, 2010} ).

Adolescence and young adulthood are stages that involve new responsibilities, more demands and stressors, and physical, cognitive, and neurological changes, so it is possible that young people are particularly vulnerable to the development of NSSI when they cannot cope with environmental demands. NSSI appear to fulfill different functions such as the emotional regulation of affective states, like depression, because when they self-injure many people report a decrease of negative mood, and a temporary relief, which influences the repetition of the behavior ( ^{Klonsky, 2007} ). Persons with NSSI appear to have more risk factors and difficulties responding in more adaptive ways to problems ( ^{Favazza, 2009} ), presenting more psychological disorders, or problems than those without self-injuries ( ^{Brown, 2009} ; ^{Cerutti, Manca, Presaghi, & Gratz, 2011} ; ^{Klonsky & Olino, 2008} ; ^{Whitlock et al., 2006} ).

In this research, we considered depressive symptoms, anxious symptoms, impulsivity, self-efficacy, and emotional regulation as relevant factors to understand NSSI. From the risk factors analyzed in this investigation, depression had the greatest association with experimental NSSI. Self-injuries among experimenters may be attempts to modify and find temporary relief to negative mood ( ^{Klonsky, 2009} ). Among those who engage in recurrent NSSI, depression, impulsivity, and low self-efficacy are significantly associated with self-injury. Contrary to what we expected, emotional regulation (cognitive reappraisal and expressive suppression) was not significant in any analyses. This is different from some studies ( ^{Klonsky, 2007} ; ^{Nock, 2009} ) where deficits in emotional regulation were associated with NSSI. It may be that the emotional regulation areas of this scale ( ^{Gross & John, 2003} ) that we used are not particularly relevant for the study of self-injury in this sample. It is possible that the emerging adults in our sample who self-injure can think of adaptive strategies to solve problems, but tend to choose more often maladaptive strategies ( ^{Nock & Mendes, 2008} ), in such a way that the differences between those without self-injury and those with NSSI are in other areas of emotional regulation. These areas might be impulse control difficulties, awareness of one’s own emotions, or the difficulty to perform tasks when negative emotions are felt ( ^{Gratz & Roemer, 2004} ).