Abstract
Background
A variety of unhealthy lifestyle behaviors (ULBs) in adolescence often exist in the form of “clustering”, and anxiety and depression often appear as comorbidities. We aimed to explore the prospective relationship between the clustered unhealthy lifestyle behaviors and comorbidity of anxiety and depression in adolescents.
Methods
Data were collected from the Chinese Adolescent Health Surveillance and Cohort Study with a sample of 6 656 adolescents. LCA was used to identify clustering patterns of multiple unhealthy lifestyle behaviors. Anxiety and depression symptoms were evaluated by GAD-7 and PHQ-9. Multivariate logistic regression model was established to analyze the association between the clustering patterns and clustering quantity of unhealthy lifestyle behaviors and the comorbidity of anxiety and depression symptoms.
Results
Multivariate logistic regression results showed that the sedentary/screen-time group (OR = 1.51, 95%CI: 1.19–1.90) and the Multiple unhealthy behaviors group (OR = 3.65, 95%CI: 2.90–4.61) significantly increased the risk of comorbidity of anxiety and depression symptoms. 4–6 (OR = 1.42, 95%CI: 1.14–1.80) and ≥ 7 (OR = 2.76, 95%CI: 2.18–3.52) unhealthy lifestyle behaviors clustering quantity significantly increased the risk of comorbidity of anxiety and depression symptoms.
Conclusion
Clustered unhealthy lifestyle behaviors are common among adolescents and strongly associated with an increased risk of subsequent anxiety-depression comorbidity. These results emphasize the necessity of monitoring adolescents with high-risk behavioral profiles. Targeted lifestyle modifications may be effective strategies for the early prevention and intervention of comorbid mental health disorders in youth.
Clinical trial number
Not applicable. (This is a cohort observational study and does not involve a clinical trial requiring prospective registration.)
Supplementary Information
The online version contains supplementary material available at 10.1186/s12888-026-07886-5.
Keywords: Adolescents, Unhealthy lifestyle behaviors, Comorbidity, Anxiety, Depression
Background
Adolescents are in a critical period of physical and mental development, and adopting healthy behaviors and lifestyles is critical to physical and mental health in adulthood. Although adolescents are generally considered to be in good health, the morbidity and mortality that can occur during this period are often triggered by ULBs (ULBs) [1]. ULBs is the behaviors that causes direct or indirect damage to adolescent health, good condition and life quality [2]. ULBs such as unhealthy eating behaviors, smoking, excessive use of smartphones and the Internet, insufficient physical activity, and injury-or violence-related behaviors are common among adolescents [3]. It is worth noting that a variety of ULBs in adolescence often exist in the form of “clustering”, that is, a variety of ULBs gather in one type of person and affect each other [4]. Globally, more than one-third of children and adolescents report co-existence of at least three ULBs, increasing with age [5]. Previous studies have shown that the aggregation of adolescent ULBs such as smoking, unhealthy eating behaviors and insufficient physical activity may affect the physical and mental health at the present stage or even in adulthood [6, 7], leading to a significant increase in the risk of mental health problems, including anxiety, depression and anxiety and depression comorbidity [8, 9].
The increasing prevalence of anxiety and depression in adolescents has become an important public health problem worldwide [10]. In addition, because anxiety and depression share common clinical manifestations, behavioral characteristics, and genetic and environmental etiological mechanisms, they are usually comorbidities [11]. Adolescents with anxiety and depression comorbidities typically have more severe symptoms, more suicidal behavior, and a higher risk of relapse than adolescents with single anxiety or depression [12]. Therefore, it is urgent to explore the etiology and possible mechanism of anxiety and depression comorbidity. Previous studies have shown that ULBs such as smoking, drinking, unhealthy eating behaviors and insufficient physical activity are associated with an increased risk of anxiety and depression in adolescents [13–16]. A meta-analysis that included 15 prospective studies showed that individuals with higher levels of physical activity had a 17% lower risk of depression compared to those with lower levels of physical activity [16]. However, to our knowledge, very few studies have explored the association between clusters of ULBs and comorbidities of anxiety and depression in adolescents. A cross-sectional study of 731 college students in China showed that college students with clusters of ULBs had a higher incidence of anxiety and depression comorbidities than those without clusters of ULBs [17]. Therefore, in order to compensate for the evidence deficiency in the etiology of anxiety and depression comorbidity, we explored the association between multiple ULBs clusters and the comorbidity of anxiety and depression symptoms (ADS) in adolescents in a large population-based longitudinal cohort study, so as to provide more clues for the etiology research of anxiety and depression comorbidity as much as possible.
Based on these considerations, we hypothesize that specific clustering patterns and the accumulating quantity of unhealthy lifestyle behaviors (ULBs) at baseline are prospectively associated with an increased risk of subsequent anxiety and depression comorbidity among adolescents.
The purpose of this study was to utilize a longitudinal cohort of 6,656 adolescents to identify the latent clustering patterns and clustering quantity of 15 common ULBs. Furthermore, we aimed to analyze the prospective associations between these behavioral clusters and the comorbidity of anxiety and depression symptoms (ADS), including sex-stratified analyses to determine potential gender-specific variations. Ultimately, this research seeks to provide a scientific foundation for understanding the behavioral etiology of anxiety-depression comorbidity and to inform targeted health interventions for the adolescent population.
Methods
Participants and procedures
The subjects of this study were participants of the Chinese Adolescent Health Surveillance and Cohort Study. At baseline (T1), conducted from October to December 2021, a stratified cluster sampling method was employed to ensure geographic and economic representativeness. Eight regions across China were selected, encompassing four first-tier cities (Shenzhen, Chongqing, Zhengzhou, and Shenyang) and four second-tier cities (Kunming, Xuzhou, Taiyuan, and Nanchang). In each region, four schools were chosen via convenience sampling: one junior high and one senior high school in both urban and rural areas. Subsequently, all students from 3 to 4 randomly selected classes in each grade (ensuring at least 100 students per grade) were invited to participate. A one-year follow-up survey (T2) was conducted from October to December 2022. At both waves, data collection was performed in classrooms using electronic questionnaires administered via smartphones, a process that typically required 10–20 min to complete. The temporal sequence of assessments is illustrated in the schematic timetable provided in Fig. S4. This study followed the basic principles of “informed consent and privacy protection” and was reviewed by the Ethics Committee of Anhui Medical University (No.20200965). Before the on-site investigation, schools, parents and students have been informed of the investigation content and implementation process, and informed consent has been obtained.
Adolescents were excluded from the study if they met any of the following criteria: (1) current severe mental illness or a history of mental illness; (2) chronic organic diseases; (3) long-term use of hormone medications; or (4) refusal to participate in the survey. Finally, a total of 6,656 participants who completed the questionnaires at both the baseline and follow-up surveys were included in the final longitudinal analysis.
Measures
Demographic factors
Sociodemographic information was collected using self-assessment questionnaires. Including gender (male or female), the economic level of the city (first-tier city or second-tier city), the current grade (junior high school or high school), the family residence (rural, urban or city), the only child (yes or no), the family economic status (poor, moderate or good), the number of friends (≤ 2 or ≥ 3), the father’s education level (primary school and below, junior high school or high school and above), mother’s education level (primary school and below, junior high school or high school and above), learning burden (light, moderate or heavy), family history of depression (yes or no).
Unhealthy lifestyle behaviors (ULBs)
In this study, We included 15 common ULBs among adolescents: smoking, alcohol consumption, physical inactivity, sleep insufficiency/hypersomnia, evening chronotypes, long screen time on weekdays, long screen time on weekend, insufficient vegetable intake, insufficient fruit intake, take out fast food consumption, skip breakfast, sugar-sweetened beverages intake, suicidal ideation, suicidal plan, suicidal behavior [18–20], their specific evaluation criteria can be found in Supplementary Table 1 (Table S1).
The 15 kinds of ULBs in the above two categories were scored cumulatively. “Yes” means that there is one health risk behavior, and 1 point is recorded. “No” means that there is no certain health risk behavior, and 0 points are scored. The total score ranges from 0 to 15 points, and the scores represent the clustering quantity of ULBs. Latent Class Analysis (LCA) was used to identify clustering patterns of the above 15 kinds of ULBs in adolescents.
Comorbid symptoms of anxiety and depression
Anxiety symptom
Generalized Anxiety Disorder Questionnaire-7 (GAD-7) was used to evaluate the participants’ anxiety symptoms in the past two weeks [21]. The scale consists of 7 items of 4-level self-rating scale, each item is set with 4 options according to the frequency from “not at all” to “almost every day”, and the total score ranges from 0 to 21 points. The higher the score, the more serious the anxiety symptoms of the subjects. Scores ≤ 4 were classified as no anxiety symptoms, ≥ 5 was classified as anxiety symptoms.
Depression symptom
Patient Health Questionnaire-9 (PHQ-9) was used to evaluate the participants’ depressive symptoms in the past two weeks [22]. The scale consists of 9 items of 4-level self-rating scale, each item has 4 options according to the frequency from “not at all” to “almost every day”, covering emotion, energy state, appetite, feeling, attention, speaking speed, suicide or self-injury thoughts, and the total score ranges from 0 to 27 points. The higher the score, the more serious the depressive symptoms of the subjects. Scores ≤ 4 were classified as no depressive symptoms, ≥ 5 was classified as depressive symptoms.
The comorbidity of anxiety and depression symptoms
If the subjects have both ADS, in other words, the GAD-7 score is ≥ 5, while the PHQ-9 score is ≥ 5, it is considered to be the comorbidity of ADS.
Confounding
Based on previous research evidence, we adjusted for a range of potential confounders that were associated with health risk behaviors and ADS, but did not have causal associations [23–25]. The confounding factors included in this study mainly included local economic level, family economic status, number of friends, learning burden and family history of depression, which could be obtained in the self-reported questionnaire.
Statistical analyses
All statistical tests were two-sided and significance was set at P < 0.05. All data analyses were performed in SPSS 23.0 and Mplus 8.3. The cumulative score method was used to describe the clustering quantity of ULBs. LCA was used to identify clustering patterns of the above 15 kinds of ULBs in adolescents. Chi-square test was used to compare the rates of comorbidity of ADS in different sociodemographic characteristics and different health risk behaviors. In addition, the rates of comorbidity of ADS in different clustering patterns and clustering quantity of ULBs were also compared. A multiple Logistic regression model was established to analyze the association between the clustering patterns and clustering quantity of ULBs at baseline and comorbidities of ADS at follow-up. Stratified analysis was performed according to sex to determine whether the above associations were sex-specific.
LCA is a form of finite mixture modelling wherein a series of models are fit to data under the assumption that the observed multivariate distribution is the result of a mixture of distributions. LCA has the advantage that it defines subgroups by considering multiple variables concurrently without considering outcome. A fit model with the optimal number of classes is determined according to the following principles: (1) the smaller the absolute values of Bayesian information criterion (BIC), Akaike information criterion (AIC) and adjusted Bayesian information criterion (aBIC), the better the model fits; (2) the Entropy values of each trajectory grouping reflect the fitting of the model. The Entropy value ranges from 0 to 1, and the closer the value is to 1, the more accurate the model classification is; (3) the data provided by the Lo-Mendell-Rubin likelihood ratio test (LMR LRT) helps to identify whether the class k model is more suitable for data than class (K-1) model. A low p value for an LMR LRT indicates that a model of class (K-1) must be rejected in favor of a model with at least class K [26].
Sensitivity analysis
To evaluate the robustness of our primary findings and the independent impact of lifestyle behaviors, we performed a sensitivity analysis by excluding three suicide-related items (suicidal ideation, planning, and behavior) from the LCA and regression models. This was done to account for the inherent overlap between suicidal behaviors and the clinical severity of anxiety and depression. A new 12-item LCA was conducted to identify clustering patterns, and their prospective associations with comorbid ADS were re-assessed using the same covariate-adjusted logistic regression models.
Results
A total of 6,656 subjects (48.0% male and 52% female) were included in the study, with a mean age of 14.2 ± 1.6 years. Among them, junior high school students accounted for 45.5%, high school students accounted for 54.5%; 47.3% of the participants come from cities with first-tier economic development; 26.9% of the participants were only children; 4.0% of participants had a family history of depression; 27.9% of adolescents have no more than 2 friends; 6.5% of adolescents think their learning burden is heavy. In addition, most of the subjects included in this study were adolescents from cities, whose family economic status was moderate, and whose parents’ educational level was junior high school (Table 1).
Table 1.
Comparison of the detection rate of comorbid anxiety-depression symptoms with different demographic characteristics
| Variable | N(%) | Comorbid symptoms of anxiety and depression |
χ2 value | P value | |
|---|---|---|---|---|---|
| Yes | No | ||||
| Gender | |||||
| Male | 3197(48.0) | 365(11.4) | 2832(88.6) | 24.245 | <0.001 |
| Female | 3459(52.0) | 538(15.6) | 2921(84.4) | ||
| Economic level | |||||
| First-tier city | 3147(47.3) | 474(15.1) | 2673(84.9) | 11.382 | 0.001 |
| Second-tier city | 3509(52.7) | 429(12.2) | 3080(87.8) | ||
| Grade | |||||
| Junior high school | 3026(45.5) | 393(13.0) | 2633(87.0) | 1.588 | 0.208 |
| Senior high school | 3630(54.5) | 510(14.0) | 3120(86.0) | ||
| Registered residence | |||||
| Rural | 1154(17.3) | 164(14.2) | 990(85.8) | 0.495 | 0.482 |
| Urban | 5502(82.7) | 739(13.4) | 4 763(86.6) | ||
| Family economic status | |||||
| Poor | 801(12.0) | 157(19.6) | 644(80.4) | 28.730 | <0.001 |
| Moderate | 5719(87.0) | 726(12.7) | 4993(87.3) | ||
| Good | 136(2.0) | 20(14.7) | 116(85.3) | ||
| Number of friends | |||||
| ≤ 2 | 1858(27.9) | 330(17.8) | 1528(82.2) | 38.669 | <0.001 |
| ≥ 3 | 4798(72.1) | 573(11.9) | 4225(88.1) | ||
| Father’s education | |||||
| Primary school and below | 800(12.0) | 121(15.1) | 679(84.9) | 1.886 | 0.390 |
| Junior high school | 4632(69.6) | 618(13.3) | 4014(86.7) | ||
| High school and above | 1224(18.4) | 164(13.4) | 1060(86.6) | ||
| Maternal education | |||||
| Primary school and below | 1322(19.9) | 177(13.4) | 1145(86.6) | 0.296 | 0.862 |
| Junior high school | 4209(63.2) | 578(13.7) | 3631(86.3) | ||
| High school and above | 1125(16.9) | 148(13.2) | 977(86.8) | ||
| Learning burden | |||||
| Light | 395(5.9) | 49(12.4) | 346(87.6) | 66.648 | <0.001 |
| Moderate | 5833(87.6) | 740(12.7) | 5093(87.3) | ||
| High | 428(6.5) | 114(26.6) | 314(73.4) | ||
| Family history of depression | |||||
| Yes | 281(4.0) | 62(22.1) | 219(77.9) | 18.066 | <0.001 |
| No | 6375(96.0) | 841(13.2) | 5534(86.8) | ||
Differences in the distribution of comorbidity of anxiety and depression symptoms
In this study, the prevalence of comorbidities of ADS was 13.6%. The comorbidities of ADS in adolescents were statistically significant in gender, local economic level, registered residence, family economic status, number of friends, learning burden and family history of depression (all P < 0.01). Specifically, the prevalence of comorbidities of ADS was significantly increased among women (15.6%), living in first-tier cities (15.1%), poor family economic status (19.6%), number of friends ≤ 2 (17.8%), heavy learning burden (26.6%), and family history of depression (22.1%) (Table 1).
Clustering patterns and clustering quantity of ULBs
Clustering patterns of ULBs
The LCA model was used to describe the clustering pattern of adolescent ULBs. Based on the fitting data and interpretability of various models, the best fitting model was classified into 4 categories, with low AIC, BIC and aBIC values, and a significant P-value for the LMR LRT, high entropy, and class probability within the interpretable range (Table S2).
The response probabilities of 15 kinds of ULBs are shown in Figure S1. According to the behavioral response probabilities, the participants were categorized into four distinct patterns. Category 1 is named the Low-risk group (24.0%), characterized by a consistently low probability of engaging in any of the 15 ULBs.Category 2 was identified as the Poor dietary habit group (40.1%), these individuals were primarily characterized by high probabilities of insufficient vegetable and fruit intake, frequent consumption of take-out fast food, and skipping breakfast.Category 3 was named the sedentary/screen-time group (22.3%), defined by high levels of physical inactivity combined with excessive screen time during both weekdays and weekends.Category 4 was designated the multiple unhealthy behaviors group (13.6%), which exhibited high response probabilities across nearly all 15 ULBs, including suicidal ideation, planning, and behavior.
Clustering quantity of ULBs
The distribution of ULB clustering quantity among the 6,656 participants followed an inverted U-shaped pattern. Based on the cumulative scores of 15 behaviors, participants were categorized into three distinct groups: 0–3 ULBs (n = 1,304, 19.6%), 4–6 ULBs (n = 3,606, 54.2%), and ≥ 7 ULBs (n = 1,746, 26.2%).
The UpSet plot (Figure S2) illustrates how these ULBs co-occur. The most frequent co-occurrence pattern involved five specific behaviours: insufficient vegetable and fruit intake, physical inactivity, long screen time at weekends and sleep insufficiency/hypersomnia. Other significant intersections also emphasised the significant overlap between dietary deficiencies and sedentary lifestyle factors.
Association between clusters of ULBs and comorbidities of anxiety and depressive symptoms
In different clustering patterns of ULBs, the highest prevalence rate of ADS comorbidity was found in multiple unhealthy behaviors group, up to 29.4%, followed by 13.7% in sedentary/screen-time group and 11.0% in poor dietary habit group, and the lowest prevalence rate of anxiety and depression comorbidity was 8.7% in low-risk group. The difference was statistically significant (P < 0.001). Among different clustering quantity of ULBs, the prevalence rates of ADS comorbidity in adolescents with 0–3 ULBs clusters, 4–6 ULBs clusters and ≥ 7 ULBs clusters increased successively, which were 7.8%, 11.7% and 21.7%, respectively, and the differences were statistically significant (P < 0.001) (Table S3).
Table 2 shows the results of multiple Logistic regression analysis on the clustering patterns of ULBs and comorbidity of ADS. In Model 1 without adjusting for any confounder, adolescents in the poor dietary habit groups (OR = 1.30, 95%CI: 1.06–1.61), sedentary/screen-time groups (OR = 1.67, 95%CI: 1.33–2.10)and multiple unhealthy behaviors groups (OR = 4.37, 95%CI: 3.49–5.48) were significantly positively associated with comorbidities of anxiety and depressive symptoms compared with adolescents in the low-risk group. Stratified analysis by sex showed that these associations were present in both men and women, and the risk of comorbidities of ADS was higher in women than in men. In Model 2, which adjusted for confounding factors such as local economic level, family economic status, number of friends, learning burden, and family history of depression, After adjusting for potential confounders, the positive associations between the sedentary/screen-time and multiple unhealthy behaviors patterns and the risk of comorbid ADS remained statistically significant (all P < 0.001).
Table 2.
Association analysis of clustering unhealthy lifestyle behaviors pattern and comorbid anxiety-depression symptoms
| Unhealthy lifestyle behaviors patterns | Model 1(OR, 95%CI) | Model 2(OR, 95%CI) |
|---|---|---|
| Total | ||
| Low-risk | 1.00 | 1.00 |
| Poor dietary habit | 1.30 (1.06–1.61)* | 1.20 (0.97–1.49) |
| Sedentary/Screen-Time Group | 1.67 (1.33–2.10)** | 1.51 (1.19–1.90)** |
| Multiple unhealthy behaviors | 4.37 (3.49–5.48)** | 3.65 (2.90–4.61)** |
| Males | ||
| Low-risk | 1.00 | 1.00 |
| Poor dietary habit | 1.26 (0.93–1.73) | 1.17 (0.86–1.61) |
| Sedentary/Screen-Time Group | 1.66 (1.20–2.31)** | 1.46 (1.05–2.05)* |
| Multiple unhealthy behaviors | 4.13 (2.89–5.91)** | 3.38 (2.34–4.89)** |
| Females | ||
| Low-risk | 1.00 | 1.00 |
| Poor dietary habit | 1.30 (0.98–1.75) | 1.21 (0.90–1.63) |
| Sedentary/Screen-Time Group | 1.68 (1.22–2.31)** | 1.56 (1.13–2.16)** |
| Multiple unhealthy behaviors | 4.21 (3.14–5.71)** | 3.60 (2.66–4.91)** |
Table 3 shows the results of multiple Logistic regression analysis on the clustering quantity of ULBs and comorbidities of ADS. In Model 1, without adjusting for any confounding factors, compared with adolescents with 0–3 ULBs clusters, adolescents with 4–6 ULBs clusters (OR = 1.56, 95%CI: 1.25–1.97) and ≥ 7 ULBs clusters (OR = 3.27, 95%CI: 2.60–4.14) were positively associated with the risk of comorbidities of ADS. The risk increased gradually, and the differences were statistically significant (P < 0.05). Stratified analysis by sex showed that these associations were found in both men and women. In Model 2, which adjusted for confounding factors such as local economic level, family economic status, number of friends, learning burden, and family history of depression, the results showed that the positive association between the risk of comorbidities of ADS remained stable in adolescents with 4 to 6 or ≥ 7 ULBs clusters (P < 0.05).
Table 3.
Association analysis of clustering unhealthy lifestyle behaviors quantity and comorbid anxiety-depression symptoms
| Unhealthy lifestyle behaviors quantity | Model 1(OR, 95%CI) | Model 2(OR, 95%CI) |
|---|---|---|
| Total | ||
| 0–3 | 1.00 | 1.00 |
| 4–6 | 1.56 (1.25–1.97)** | 1.42 (1.14–1.80)** |
| ≥ 7 | 3.27 (2.60–4.14)** | 2.76 (2.18–3.52)** |
| Males | ||
| 0–3 | 1.00 | 1.00 |
| 4–6 | 1.55 (1.13–2.15)** | 1.42 (1.03–1.98)* |
| ≥ 7 | 3.11 (2.23–4.41)** | 2.60 (1.83–3.73)** |
| Females | ||
| 0–3 | 1.00 | 1.00 |
| 4–6 | 1.51 (1.10–2.10)* | 1.38 (1.00-1.93) |
| ≥ 7 | 3.12 (2.28–4.35)** | 2.67 (1.93–3.76)** |
Sensitivity analysis based on 12 unhealthy lifestyle behaviors
The sensitivity analysis based on 12 ULBs yielded clustering patterns and associations that were highly consistent with the primary 15-item analysis (Table S4–S7 and Figure S3). In the 12-item LCA, four similar clusters were identified: Low-risk lifestyle, Poor dietary habit, Late sleep and little exercise, and multiple unhealthy behaviors. Compared to the Low-risk group, the multiple unhealthy behaviors group remained the strongest predictor of comorbid ADS (Adjusted OR = 1.95, 95%CI: 1.48–2.57). Furthermore, a significant dose-response relationship was maintained between the quantity of 12-item ULBs and the risk of comorbid symptoms (Adjusted OR per 1 additional ULB = 1.13, 95%CI: 1.09–1.18).These results confirm that the observed associations between behavioral clustering and mental health outcomes are robust and not solely driven by the inclusion of suicide-related indicators.
Discussion
To the best of our knowledge, while previous studies have explored associations between ULBs and anxiety or depression, this study is the first to explore associations between clustering patterns and clustering quantity of ULBs and comorbidities of anxiety and depressive symptoms in a youth-based cohort study. First, our results show that the comorbidity of anxiety and depressive symptoms in adolescents is 13.6%. A new finding of this study is that the current clustering patterns of ULBs in adolescents can be divided into four categories, namely, low-risk group, poor dietary habit group, sedentary/screen-time group, and multiple unhealthy behaviors group. Compared with 0–3 ULBs clusters, more adolescents had 4 or more ULBs clusters at the same time. Secondly, compared with adolescents with 0–3 ULBs clusters, adolescents with 4–6 ULBs clusters and ≥ 7 ULBs clusters were positively associated with comorbidities of ADS. Another new finding of this study is that compared with adolescents in the low-risk group with the clustering pattern of ULBs, adolescents in the poor dietary habit group, sedentary/screen-time group, and multiple unhealthy behaviors group. are significantly positively associated with the risk of comorbidity of ADS. Multiple ULBs during adolescence often exist in the form of “aggregation”, that is, multiple ULBs gather in one type of person and affect each other [4]. To our knowledge, this study is the first to describe clustering patterns of such a large range of ULBs in an adolescent population, covering nearly all common adolescent ULBs. In this study, the clustering patterns of adolescent ULBs were divided into four categories. The detection rates of low-risk group, poor dietary habit group, sedentary/screen-time group, and multiple unhealthy behaviors group. were 24.0%, 40.1%, 22.3% and 13.6%, respectively. A cross-sectional study of 687 female adolescents in Australia divided the clustering patterns of ULBs in Australian women into three groups, namely, the active exercise healthy diet group, the moderate risk group and the excessive video viewing group, accounting for 19%, 76% and 5%, respectively [27]. The results of a cross-sectional study in Thailand, which included 25 566 students from 197 schools, showed that the clustering patterns of ULBs among Thai adolescents were mainly divided into low-risk group, drinking and fighting group and high-risk group, with detection rates of 87.7%, 11.7% and 0.6%, respectively [28]. The classification and proportion of ULBs in this study are slightly different from those in domestic and foreign studies, which may be due to the interaction between ULBs, and the change of one behavior will affect the incidence of other behaviors [29]. Another possible reason is that differences between the natural and social environments in different regions may also influence the clustering patterns of ULBs. In addition, this study also found that the presence of multiple ULBs clusters in the adolescent population is quite common. We found that the distribution of ULBs clusters among adolescents was 19.6% with 0–3 ULBs clusters, 54.2% with 4–6 ULBs clusters, and 26.2% with 7 or more ULBs clusters.
This study found that the comorbidity of ADS in adolescents was 13.6%. A longitudinal study of 1 110 college students in central China showed that the prevalence of comorbidity of anxiety and depression among college students was 18.4% [30]. In 2008, a health survey of 439 adolescents aged 12 to 17 in the United States showed that the comorbidity of anxiety and depression in males and females was 17.1% and 28.4%, respectively [31]. A cross-sectional study of Australian university students showed that the comorbidity of anxiety and depression was 33.2% [32]. The comorbidities of anxiety and depression in adolescents are different in different countries, which may be due to the different social and cultural background and economic conditions in different study areas, the different sample size, and the different evaluation tools for anxiety and depression. However, the existing research evidence indicates that the comorbidity of anxiety and depression is generally high in adolescents.
This study explored the association between the clustering patterns and clustering quantity of ULBs and comorbidities of anxiety and depressive symptoms in adolescents. We found that compared with adolescents with 0–3 ULBs clusters, adolescents with 4–6 ULBs clusters and ≥ 7 ULBs clusters were positively correlated with the risk of comorbidity of ADS, and the risk of comorbidity of ADS gradually increased with the increase of the number of ULBs clusters. This is consistent with existing research evidence on the association between ULBs and comorbidities of anxiety and depressive symptoms [9, 33]. At the same time, we also have a new finding, that is, in different clustering patterns of ULBs, compared with the low-risk group, adolescents in the sedentary/screen-time groups and multiple unhealthy behaviors groups are significantly positively correlated with the risk of ADS of comorbidity, and adolescents in the multiple unhealthy behaviors group have the highest risk of ADS of comorbidity. The response probability of high suicide ideation, high suicide planning and high suicide behavior, as well as 12 other ULBs, was higher in the multiple unhealthy behaviors group. The inclusion of suicide-related behaviors in the ‘Multiple unhealthy behaviors group’ is supported by neurobiological evidence suggesting that such behaviors are closely linked to systemic inflammation. As highlighted in a recent review, elevated levels of pro-inflammatory cytokines may drive behavioral changes associated with suicidality by altering neurotransmitter systems [34]. This finding not only provides new evidence support for the association between ULBs clustering and anxiety and depression comorbidity, but also is consistent with previous research evidence to a certain extent. A cross-sectional study of 28 298 college students in China showed a positive dose-response relationship between sedentary behavior and symptoms of anxiety and depression. The longer the time spent sitting, the higher the incidence of ADS in college students [35]. A cross-sectional study of 1 274 middle school students in Zhejiang Province, China, showed that middle school students in the high-risk behavior and poor diet groups had a higher risk of developing depressive symptoms than those in the behavioral health group [36]. Results from a cross-sectional study of 339 nurses in Spain showed that moderate physical activity was associated with a reduced risk of depressive symptoms, while low levels of physical activity, unhealthy eating habits, and alcohol abuse were associated with an increased risk of anxiety symptoms [37]. In addition, gender did not change the association between the clustering patterns and clustering quantity of ULBs in adolescents and the comorbidities of ADS, suggesting that the clustering of ULBs has a general effect on the risk of ADS comorbidities, regardless of gender.
Notably, the stability of our findings across the primary and sensitivity analyses underscores the significant role of clustered lifestyle factors. By excluding suicide-related items in the sensitivity analysis, we demonstrated that the clustering of dietary, sleep, and physical activity behaviors independently predicts a higher risk of subsequent anxiety-depression comorbidity. This further validates that comprehensive lifestyle interventions targeting these behavioral profiles are essential for adolescent mental health prevention.
From a transdiagnostic and developmental psychopathology perspective, the identified clustering of ULBs likely represents a shared behavioral substrate that contributes to the onset and maintenance of internalizing conditions during adolescence. Rather than acting as specific triggers for either anxiety or depression in isolation, these behavioral profiles may serve as common risk factors that exacerbate overall psychological vulnerability. This framing is crucial for public health, as it suggests that targeting these ‘shared substrates’ through comprehensive lifestyle interventions could simultaneously mitigate the risk for multiple comorbid psychiatric disorders, shifting the intervention focus from individual symptoms to broader developmental health trajectories.
As a cohort study in adolescents, this study included a wide range of 15 common ULBs, and described in detail the clustering patterns and clustering quantity of ULBs, providing new research evidence for the clustering of ULBs in adolescents. In addition, this study further found that the clustering of ULBs is associated with an increased risk of comorbidities of ADS in adolescents, suggesting that adolescents with multiple ULBs should be considered as the key population for the prevention and intervention of ADS. Schools, parents and society should work closely together to promote the healthy physical and mental development of young people. However, there are several limitations to this study: First of all, the adolescents included in this study are currently enrolled in school, and teenagers who are not enrolled in school for other reasons, such as dropping out of school, are not included, so the extrapolation of the results to the overall youth is limited. Compared with those who drop out of school, adolescents who are attending school may have a relatively low incidence of health risk behaviors due to school management, parental constraints, and other reasons. Second, the evaluation of ADS in this study used self-rated questionnaires rather than standard clinical diagnostic assessments, so reporting bias may not be avoided. Finally, although some confounding factors have been adjusted during the analysis, the influence of unmeasured confounding factors on the results, such as early life factors and adverse childhood experiences, cannot be excluded.
Conclusions
In this population-based longitudinal cohort study of 6,656 adolescents, we found that clusters of ULBs are common and are prospectively associated with a significantly increased risk of comorbid anxiety and depression symptoms. These associations remained robust in sensitivity analyses specifically excluding suicide-related items. Our findings suggest that adolescents with high-risk behavioral profiles should be more closely monitored to facilitate the prevention and early diagnosis of psychiatric disorders. Future large-scale longitudinal studies are warranted to explore whether targeted, cluster-based lifestyle interventions can effectively mitigate the development of comorbid mental health conditions in the youth population.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
The authors sincerely thank the consent of data access from Chinese Adolescent Health Surveillance and Cohort Study, carried out by the Institute of Anhui Medical University, and all staff andstudents involved in this study for their full support.
Abbreviations
- ULBs
Unhealthy lifestyle behaviors
- ADS
Anxiety and depression symptoms
- MCTQ-SF
Munich Chronotype Questionnaire Short Form
- MEQ-5
Morningness-Eveningness Questionnaire-5
- LCA
Latent Class Analysis
- GAD-7
Generalized Anxiety Disorder Questionnaire-7
- PHQ-9
Patient Health Questionnaire-9
- CV
Coefficient of variation
- BIC
Bayesian information criterion
- AIC
Akaike information criterion
- LMR LRT
Lo-Mendell-Rubin likelihood ratio test
- OR
Odds ratio
- CI
Confidence interval
Author contributions
W. C. contributed to the writing of this article. Z. R. and M. W. were responsible for the data analysis of the study. X. W. and F. T. proposed the research design and contributed to the design of the study, conceived the survey. S. T. and L. Z. were involved in the fieldwork. T. L., Y. C., Y. N., Y. C. and T. J. played a role in participant recruitment and data collection. All authors reviewed the manuscript.
Funding
This study was supported by National Natural Science Foundation of China (82373592, 82173542) and the Scientific Research Level Improvement Plan from Anhui Medical University (2022xkjT004).
Data availability
No datasets were generated or analysed during the current study.
Declarations
Ethics approval and consent to participate
The design and data collection were reviewed and approved by the Ethics Committee of Anhui Medical University, China (No.20200965). Before the on-site investigation, schools, parents and students have been informed of the investigation content and implementation process, and informed consent has been obtained. All participants provided written informed consent prior to the survey. All methods were carried out in accordance with the Declaration of Helsinki and relevant guidelines and regulations.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Wanyu Che and Ziyan Ruan contributed equally to this work.
References
- 1.Patton GC, Sawyer SM, Santelli JS, Ross DA, Afifi R, Allen NB, et al. Our future: a lancet commission on adolescent health and wellbeing. Lancet Lond Engl. 2016;387:2423–78. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.de Moura LR, Torres LM, Cadete MMM, de Cunha C. Factors associated with health risk behaviors among Brazilian adolescents: an integrative review. Rev Esc Enferm U P. 2018;52:e03304. [DOI] [PubMed] [Google Scholar]
- 3.Song X, Zhou B, Baird S, Lu C, Song Z, Zhang Y, et al. Trends and predictions to 2030 in demographic structures and metabolic health for children and adolescents in china: analysis of National school health surveys from 2000 to 2019. Lancet Child Adolesc Health. 2025;9:530–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Ames ME, Leadbeater BJ. Depressive symptom trajectories and physical health: persistence of problems from adolescence to young adulthood. J Affect Disord. 2018;240:121–9. [DOI] [PubMed] [Google Scholar]
- 5.Uddin R, Lee EY, Khan SR, Tremblay MS, Khan A. Clustering of lifestyle risk factors for non-communicable diseases in 304,779 adolescents from 89 countries: A global perspective. Prev Med. 2020;131:105955. [DOI] [PubMed] [Google Scholar]
- 6.Champion KE, Mather M, Spring B, Kay-Lambkin F, Teesson M, Newton NC. Clustering of multiple risk behaviors among a sample of 18-Year-Old Australians and associations with mental health outcomes: A latent class analysis. Front Public Health. 2018;6:135. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Wijbenga L, de Winter AF, Almansa J, Vollebergh WAM, Korevaar EL, Hofstra J, et al. Multiple health risk behaviors and mental health from a life course perspective: the Dutch TRAILS study. Prev Med. 2022;154:106870. [DOI] [PubMed] [Google Scholar]
- 8.Firth J, Solmi M, Wootton RE, Vancampfort D, Schuch FB, Hoare E, et al. A meta-review of ‘lifestyle psychiatry’: the role of exercise, smoking, diet and sleep in the prevention and treatment of mental disorders. World Psychiatry Off J World Psychiatr Assoc WPA. 2020;19:360–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Zhang J, Liu MW, Yu HJ, Chen QT, Tang BW, Yuan S, et al. Associations of health-risk behaviors with mental health among Chinese children. Psychol Health Med. 2022;27:528–36. [DOI] [PubMed] [Google Scholar]
- 10.Liu Q, He H, Yang J, Feng X, Zhao F, Lyu J. Changes in the global burden of depression from 1990 to 2017: findings from the global burden of disease study. J Psychiatr Res. 2020;126:134–40. [DOI] [PubMed] [Google Scholar]
- 11.Ranøyen I, Lydersen S, Larose TL, Weidle B, Skokauskas N, Thomsen PH, et al. Developmental course of anxiety and depression from adolescence to young adulthood in a prospective Norwegian clinical cohort. Eur Child Adolesc Psychiatry. 2018;27:1413–23. [DOI] [PubMed] [Google Scholar]
- 12.Choi KW, Kim YK, Jeon HJ. Comorbid anxiety and depression: clinical and conceptual consideration and transdiagnostic treatment. Adv Exp Med Biol. 2020;1191:219–35. [DOI] [PubMed] [Google Scholar]
- 13.Anker JJ, Kushner MG. Co-Occurring alcohol use disorder and anxiety: bridging psychiatric, psychological, and neurobiological perspectives. Alcohol Res Curr Rev. 2019;40:arcr.v40.1.03. [DOI] [PMC free article] [PubMed]
- 14.Fluharty M, Taylor AE, Grabski M, Munafò MR. The association of cigarette smoking with depression and anxiety: A systematic review. Nicotine Tob Res Off J Soc Res Nicotine Tob. 2017;19:3–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Pearce M, Garcia L, Abbas A, Strain T, Schuch FB, Golubic R, et al. Association between physical activity and risk of depression: A systematic review and Meta-analysis. JAMA Psychiatry. 2022;79:550–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Wootton RE, Richmond RC, Stuijfzand BG, Lawn RB, Sallis HM, Taylor GMJ, et al. Evidence for causal effects of lifetime smoking on risk for depression and schizophrenia: a Mendelian randomisation study. Psychol Med. 2020;50:2435–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Wang M, Li T, Xie Y, Zhang D, Qu Y, Zhai S, et al. Clustered health risk behaviors with comorbid symptoms of anxiety and depression in young adults: moderating role of inflammatory cytokines. J Affect Disord. 2024;345:335–41. [DOI] [PubMed] [Google Scholar]
- 18.Wang K, Li Y, Zhang T, Luo J. The relationship among college students’ physical Exercise, Self-Efficacy, emotional Intelligence, and subjective Well-Being. Int J Environ Res Public Health. 2022;19:11596. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Ghotbi N, Pilz LK, Winnebeck EC, Vetter C, Zerbini G, Lenssen D, et al. The µMCTQ: an Ultra-Short version of the Munich chronotype questionnaire. J Biol Rhythms. 2020;35:98–110. [DOI] [PubMed] [Google Scholar]
- 20.Paruthi S, Brooks LJ, D’Ambrosio C, Hall WA, Kotagal S, Lloyd RM, et al. Recommended amount of sleep for pediatric populations: A consensus statement of the American academy of sleep medicine. J Clin Sleep Med JCSM Off Publ Am Acad Sleep Med. 2016;12:785–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Spitzer RL, Kroenke K, Williams JBW, Löwe B. A brief measure for assessing generalized anxiety disorder: the GAD-7. Arch Intern Med. 2006;166:1092–7. [DOI] [PubMed] [Google Scholar]
- 22.Accuracy of Patient Health. Questionnaire-9 (PHQ-9) for screening to detect major depression: individual participant data meta-analysis. BMJ. 2019;365:l1781. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Ssewamala FM, Brathwaite R, Neilands TB, Economic Empowerment HIVR. Behavior, and mental health among School-Going adolescent girls in uganda: longitudinal Cluster-Randomized controlled Trial, 2017–2022. Am J Public Health. 2023;113:306–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.van Dijk MT, Murphy E, Posner JE, Talati A, Weissman MM. Association of multigenerational family history of depression with lifetime depressive and other psychiatric disorders in children: results from the adolescent brain cognitive development (ABCD) study. JAMA Psychiatry. 2021;78:778–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Yu S, Zheng J, Xu Z, Zhang T. The transformation of parents’ perception of education Involution under the background of ‘Double reduction’ policy: the mediating role of education anxiety and perception of education equity. Front Psychol. 2022;13:800039. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Sinha P, Calfee CS, Delucchi KL. Practitioner’s guide to latent class analysis: methodological considerations and common pitfalls. Crit Care Med. 2021;49:e63–79. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Gardner LA, Champion KE, Parmenter B, Grummitt L, Chapman C, Sunderland M, et al. Clustering of six key risk behaviors for chronic disease among adolescent females. Int J Environ Res Public Health. 2020;17:7211. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Assanangkornchai S, Li J, McNeil E, Saingam D. Clusters of alcohol and drug use and other health-risk behaviors among Thai secondary school students: a latent class analysis. BMC Public Health. 2018;18:1272. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Nauha L, Jurvelin H, Ala-Mursula L, Niemelä M, Jämsä T, Kangas M, et al. Chronotypes and objectively measured physical activity and sedentary time at midlife. Scand J Med Sci Sports. 2020;30:1930–8. [DOI] [PubMed] [Google Scholar]
- 30.Qu Y, Wu X, Tao S, Yang Y, Zou L, Xie Y, et al. Association between cell phone dependence and anxiety-depression co-morbid symptoms among college students. Chin J SCHOOL HEALTH. 2021;42(12):1842–6. [Google Scholar]
- 31.Small DM, Simons AD, Yovanoff P, Silva SG, Lewis CC, Murakami JL, et al. Depressed adolescents and comorbid psychiatric disorders: are there differences in the presentation of depression? J Abnorm Child Psychol. 2008;36:1015–28. [DOI] [PubMed] [Google Scholar]
- 32.Bitsika V, Sharpley CF. Comorbidity of anxiety-depression among Australian university students: implications for student counsellors. Br J Guid Couns. 2012;40:385–94. [Google Scholar]
- 33.Orji CC, Ghosh S, Nwaobia OI, Ibrahim KR, Ibiloye EA, Brown CM. Health behaviors and Health-Related quality of life among U.S. Adults aged 18–64 years. Am J Prev Med. 2021;60:529–36. [DOI] [PubMed] [Google Scholar]
- 34.Baldini V, Gnazzo M, Varallo G, Atti AR, De Ronchi D, Fiorillo A, et al. Inflammatory markers and suicidal behavior: A comprehensive review of emerging evidence. Ann Gen Psychiatry. 2025;24(1):36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Jiang L, Cao Y, Ni S, Chen X, Shen M, Lv H, et al. Association of sedentary behavior with Anxiety, Depression, and suicide ideation in college students. Front Psychiatry. 2020;11:566098. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Xiaoyong SQL, Qinghai G. Pan W, Huang Y. Association between latent class of health-risk behaviors and depressive symptoms among middle school students. Preventive Medicine. 2024;36(6):474–478.
- 37.Ramón-Arbués E, Sagarra-Romero L, Echániz-Serrano E, Granada-López JM, Cobos-Rincón A, Juárez-Vela R, et al. Health-related behaviors and symptoms of anxiety and depression in Spanish nursing students: an observational study. Front Public Health. 2023;11:1265775. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
No datasets were generated or analysed during the current study.