Sleep disturbance (SD) is prevalent among adolescents and young adults, with main symptoms of insomnia, poor sleep quality, and other sleep complaints. College students, one of the most sleep deprived student groups, are particularly vulnerable to disturbed sleep. One epidemiological survey showed that more than 60% of college students were suffering from SD [on the Pittsburgh Sleep Quality Index (PSQI)] (1). Gaultney, in a sample of 1845 college students, found that 27% have clinically significant symptoms of SD (on the SLEEP-50 questionnaire) (2). Recent systematic review, involving 250 studies, estimated that global prevalence of SD among college students was 41.16% during the COVID-19 pandemic (3). Adolescents entering college is a major life transition, which means leaving home and living independently, completing demanding course and trying to earn academic degrees, facing a new social environment, and adopting new social roles (4–6). This transition contributes to increased stress levels, which, in turn, can have a profound and lasting effect on sleep (7). Moreover, studies have also shown that excessive media use by college students contributes to poor sleep hygiene and a series of sleep disturbances (8).

A growing number of research have focused on SD of college students and its symptoms, such as insomnia (difficulty initiating sleep, difficulty maintaining sleep, or early morning awakening) (9), poor subjective sleep quality (10, 11), daytime sleepiness (12), salient nightmares (13), and other general sleep problems (14). The majority of extant research on SD has relied on variable-centered analyses, which provides insight into the average relations among symptoms or diagnoses of SD within a specific group. However, the methodological approaches fail to reveal different patterns between individuals and may therefore draw over-generalized conclusions based on overall samples (15). On the contrary, the use of person-centered approaches such as latent class analysis (LCA) and latent profile analysis (LPA), to evaluate the structure of psychopathology, enables a more refined understanding of symptom presentations and is ideal for psychopathology research (16, 17).

Till now, the heterogeneity of SD for college students has only been scarcely explored. Based on the available literature, only three studies have been conducted using a person-centered methodological approach to explore the heterogeneous patterns of sleep-related impairment among college students. For instance, Carpi et al. (18) identified a five-class solution among a small sample of college students (N = 490) by using LPA, which were termed as the “severe insomnia” subtype, “moderate insomnia with medication use” subtype, “subthreshold insomnia” subtype, “subthreshold insomnia with sleep latency complaints” subtype, and “moderate insomnia with sleep duration complaints” subtype (18). Zhou et al. (19) applied LCA to discern heterogeneous patterns of sleep behaviors among 1,288 Chinese college students, and summarized four subtypes: good sleep, prolonged sleep latency, sleep disturbances–daytime dysfunction, and multiple poor sleep behavior (19). Another study in 312 at-risk college drinkers also indicted that the four profiles can clearly explain sleep and sleep-related consequences (20). Therefore, further studies with a large sample are needed to confirm heterogeneous patterns of SD among general college students.

This study intended to explore the heterogeneity of SD among the college students group with a large sample web-based survey for college students. As suggested by prior research, the risks of occurrence and persistence of SD are associated with numerous sample characteristic factors, including age (21, 22), sex (23, 24), parental marital status (25), and family economic status (26). Additionally, solid evidence showed that SD is also associated with a wide variety of adverse mental health outcomes, such as depression (27) and psychotic-like experiences (PLEs) (28), as well as inhibits the development of positive mental qualities, such as resilience (29). Collectively, there are three main aims of this study. First, the study aims to identify the heterogeneity differences in SD among college students, employing LPA. Second, it explores what sample characteristic factors (e.g., sex, age) are significant predictors of distinct profiles of SD in college students. Finally, it examines the associations among these different subtypes of SD and mental health outcomes. Based on previous research, we speculated that the college students’ SD feature obvious latent profiles, and several sample characteristic factors, such as sex, parental marital status, were significant predictors of distinct profiles for SD. We also anticipated that there are robust associations between SD subtypes and depression, PLEs and resilience.

The present study adopted LPA to explore subtypes of SD and their related factors in college students. The findings are beneficial to education administrators and clinicians better understand the heterogeneous cluster classification characteristics of SD in college students to formulate more specific and targeted intervention measures.

First, three latent profiles of SD among college students were identified in this study: “high SD,” “mild SD,” and “no SD.” The finding of the presence of three potential SD profiles is similar to previous studies in other groups (49, 50). For example, Tejada and colleagues explored the heterogeneity of SD in 1,331 chemotherapy patients using LPA, pointing out three SD profiles: “low,” “high,” and “very high” (50). Our data showed that approximately half of the samples (51.9%) were categorized as “no SD” profile, and only 10.6% have high SD. We compared the results with previous studies of similar university students. Zhou and colleagues used LCA to analyze the distinct classes of sleep behavior (on the Dysfunction Beliefs and Attitudes about Sleep [DBAS] and PSQI), and concluded that 6.8% of college students was in the “sleep disturbances and daytime dysfunction” subtype, while 31.8% was in the “good sleep” subtype (19). Another study (on the Insomnia Severity Index [ISI] and PSQI) adapted LPA and pointed out that 8.8% college students have severe insomnia, 66.9% reported subthreshold insomnia or subthreshold insomnia with sleep latency complaints (composed by students with relatively low insomnia severity) (18). Due to instruments and sampling, the findings differ to some extent. However, these studies have reached consistent conclusions, that is, most individuals could cope well to stressor of life events and maintain good sleep function, while only a small percentage exhibit clinical levels of SD. In addition, 37.5% of the samples were categorized as “mild SD” profile. Although the YSIS scores of these college students did not meet the criteria for clinically SD (cut-off value = 26), they had higher than average levels of SD. College students in “mild SD” profile still exhibited low frequency (the average score of the 8 YSIS item was 2.18) of SD related symptoms. Therefore, it is necessary to employ targeted interventions for those in this subtype to prevent them from sliding into sleep-related impairment and SD.

Second, multivariate logistic regression identified several risk factors associated with increased likelihood of experiencing moderate and high SD. Our data showed that male students were more likely to be found in the “high SD” profile than the female. Although previous studies have examined SD in the college students group, findings concerning the effects of sex on SD have been inconsistent. Several studies found that male students exhibiting greater SD (51–53). However, some studies suggested that SD among female students was even more severe than that of the male (23, 54, 55), while others reported no sex differences in SD (56–58). We speculated that male college students have more sleep-adverse lifestyles, such as Internet addiction, alcohol and energy drinks consumption, and cigarette smoking (59, 60). Poor parental marital status was also one of the risk factors for SD, in line with previous research (25, 61). Stressful changes in the parents’ relationship lead to disruptions in family processes, which, in turn, negatively impact the individual’s physiologic functioning (62, 63), and physiologic changes may interfere with sleep (64, 65). The results also indicate the possible impacts of the sophomores on the development of SD among college students. The finding is similar with previous study (66), in which sophomores had more professional course loads compared with freshmen, and these stress may have affected their sleep quality to some extent.

Third, college students in the “mild SD” and “high SD” profile report greater depressive symptoms and PLEs in comparison to those in the no SD profile. A great volume of literature has consistently asserted that SD as a risk factor for depressive symptoms and PLEs (28, 67, 68). Recent evidence has also suggested that college students with SD are prone to maladaptive emotion regulation (69) and impaired social functioning (70), resulting in the development of depressive symptoms and PLEs. Additionally, our study also found SD negatively associated with resilience. Studies in children have reported that greater SD reduced resilience (71). Another longitudinal study also found that higher sleep rhythmicity and fewer SD in early age predicted higher behavioral control in adolescence, which, in turn, predicted resilience in young adulthood (72).

Accordingly, therapeutic interventions targeting SD may benefit from a tailored approach that takes individual symptom patterns of SD into great consideration. For example, college students in profile 1 (high SD) have poor sleep quality and more adverse psychological outcomes, therefore require prompt clinical treatment to help alleviate their sleep disorders. Meanwhile, although college students in profile 1 (mild SD), who did not have particularly severe SD related symptoms, were still at higher risk for mental health problems relative to those in the no SD profile. Therefore, it is necessary to optimize the sleep habits of these students and improve their sleep quality. Several interventions can be carried out (e.g., cognitive behavioral therapy, mindfulness and hypnotherapy (73)) to improve sleep quality for college students in the mild SD and high SD profile. Also, influential factors should also be taken into consideration for effective psychosocial intervention for college students. Some students, especially those with male gender, sophomores and poor parental marital status, exhibit moderate or high SD. Thus, the need for individualized intervention with these students is indicated.

Despite all the relevant findings, several limitations of the current study should be noted. First, measures of SD, depressive symptom, PLEs, and other psychological factors relied on self-report questionnaires, which might be influenced by reporting bias caused by recollection inaccuracy and individuals’ own psychiatric states. Second, the data were collected in the Shanxi Province of China, which is uncertain whether the findings could be generalized to all college students to other regions of China. Thus, future studies would benefit from increasing the sampling range and sample size. Finally, this was a cross-sectional study, which limits the ability to make causal inference. In clinical practice, SD and mental health may do form a bidirectional relationships (74–76). Therefore, future research is necessary to design longitudinal studies to further explore the effects of different SD subtypes on mental health outcomes.

These findings indicate that SD in Chinese college students has obvious classification characteristics. We identified three profiles of SD in college students. Furthermore, we suggest that education or clinical workers should pay particular attention to college students classified into the mild SD and high SD profile and especially male college students, sophomores and those with poor parental marital status.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.

The studies involving human participants were reviewed and approved by Ethics Committees of South China Normal University. The patients/participants provided their written informed consent to participate in this study.

DW: conceptualization. CC and DW: methodology, formal analysis, and writing—original draft. ZH, BX, and JS: data curation. BX and DW: writing–review and editing. All authors contributed to the article and approved the submitted version.

The present study was funded by Graduate Research and Innovation Project of School of Psychology, South China Normal University (PSY-SCNU202017).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.