Association between somatic symptoms and depression and anxiety in adolescents: a cross-sectional school-based study

Abstract

Objectives

To explore the associations between somatic symptoms, depression and anxiety among Chinese adolescents in non-clinical settings.

Design

A cross-sectional study.

Setting

The questionnaire was distributed among Chinese students in Grades 10–11 from four vocational high schools (two in Anhui Province and two in Sichuan Province).

Participants

Across the four schools, 85 out of 4500 eligible students were excluded because of parental refusal, 296 students were excluded because they were absent from school during the survey time, and 4119 completed the questionnaire (38.29% females).

Primary outcome measures

Students’ somatic symptoms, depression and anxiety were measured using the somatisation subscale of the Brief Symptom Inventory 18, the Patient Health Questionnaire-9 and the Generalised Anxiety Disorder-7, respectively.

Results

After controlling for gender, age, family economic status, and paternal and maternal education level, compared with students without somatic symptoms, students with mild (OR=8.15; 95% CI=6.46 to 10.29), moderate (OR=18.78; 95% CI=13.08 to 26.96) and severe (OR=23.07; 95% CI=8.86 to 60.07) somatic symptoms reported significantly higher prevalence of depression; students with mild (OR=6.70; 95% CI=5.03 to 8.93), moderate (OR=16.41; 95% CI=11.38 to 23.67) and severe (OR=20.03; 95% CI=8.52 to 47.11) somatic symptoms reported significantly higher prevalence of anxiety.

Conclusion

The associations between somatic symptoms, depression and anxiety urge attention from caregivers and educators on young individuals experiencing somatic symptoms. Our findings highlight the potential predictive effects of somatic symptoms on depression and anxiety among Chinese students.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• This study used a large sample size of students from four schools in two geographically distant provinces (Anhui Province and Sichuan Province), so sampling bias issues are expected to be minimal.

• The time lag existing in the tools we used to measure students’ somatic symptoms, depression and anxiety symptoms may influence our results.

• This is a self-administered survey, and our results exclusively relied on students’ self-reporting.

Background

Somatic symptoms, which are physical manifestations not fully explained by a medical condition, are a complex and multifaceted phenomenon. They can range from mild complaints such as faintness and nausea to more severe issues like difficulty breathing.¹ These symptoms are not only prevalent in the general population but also constitute a significant portion of paediatric outpatient visits, underscoring their ubiquity and clinical relevance. Many studies have examined somatic symptoms using varied methods, including self-reported health states,² medical record analyses³ and direct clinical observations.⁴

In the paediatric domain, somatic complaints are alarmingly common, with nearly half of all children and adolescents reporting at least one such symptom within a 2-week period.⁵ While many of these complaints are transient and resolve on their own,⁶ a concerning subset of children experience chronic somatic symptoms that significantly disrupt their daily life, leading to psychosocial challenges such as school absenteeism and social isolation.⁷

The misinterpretation of these symptoms by caregivers, particularly when they are not corroborated by medical professionals, can lead to a dismissal of the child’s distress as a means to avoid school or other obligations. This oversight is problematic, as current aetiological frameworks recognise the intricate interplay between physiological and psychological factors in the manifestation of somatic symptoms.⁸ It is crucial to enhance awareness of the potential psychiatric underpinnings in youth presenting with somatic symptoms to ensure comprehensive clinical management.

Globally, approximately 20% of children and adolescents face psychiatric challenges, a figure that is expected to rise with increasing environmental stressors.⁹ Despite the significant impact of these conditions, many young individuals remain untreated due to the subtle nature of their symptoms, which can be easily overlooked by caregivers and educators. Societal reticence to discuss mental health, particularly among certain cultural groups that favour discussing physical complaints, further compounds this issue.^{10 11} While there are numerous validated tools to assess psychiatric issues in children and adolescents, concerns about privacy and the fear of discrimination can hinder honest responses to such assessments. Building on previous research that has linked an increased prevalence of unintentional injuries among youth with self-harm behaviours,¹² this study aims to explore the association between somatic symptoms and psychiatric conditions.

Adolescents with these disorders often present with a wide array of somatic complaints, from respiratory distress to energy depletion.^{13 14} In children and adolescents, somatic symptoms may delay the diagnosis of underlying psychiatric conditions such as depression and anxiety, as caregivers often seek medical attention for these physical manifestations.¹⁵ Children and adolescents may also lack the ability to advocate for further medical support when their caregivers dismiss these symptoms as trivial, particularly when medical professionals fail to validate them. Most previous studies have been conducted in clinical settings, which, while informative, may not capture the full spectrum of these issues in non-clinical populations. Data from non-clinical settings could enhance population-based mental health initiatives for young people.

Contextualising these findings within China’s 2021 Ministry of Education mandate for nationwide adolescent mental health screening,¹⁶ like depression, reveals implementation challenges. Concerns over privacy disclosure and fear of discrimination can lead some children and adolescents to hesitate from responding honestly to such surveys. This disclosure disparity highlights the strategic value of investigating somatic-mental-health associations among Chinese youth in non-clinical settings. Evidence in non-clinical settings about the associations between somatic symptoms, depression and anxiety among Chinese adolescents can greatly help optimise this screening initiative since the disclosure of mental health problems is always more stigmatised than that of somatic symptoms.

Vocational education is far less attractive in China than in other education sectors, as in other countries.¹⁷ Vocational education is often perceived as a ‘last resort’ for academically underperforming students, reinforced by the ‘vocational-academic tracking’ policy that diverts 50% of middle school graduates to vocational pathways.¹⁸ This systemic labelling cultivates pervasive social stigma, leading to internalised shame and low self-efficacy.¹⁹ These issues are exacerbated by their overrepresentation from marginalised backgrounds (eg, rural, left-behind or low-income families). To our knowledge, there is a dearth of research on the associations between somatic symptoms, depression and anxiety among Chinese vocational high schools, where mental health resources are presumably scarce. I believe the background should clearly articulate the unique characteristics specifically. The current study, a school-based large-scale survey, aims to address this gap with two primary hypotheses:

Hypothesis 1: there is a positive association between the prevalence of somatic symptoms and the occurrence of depression and anxiety in students.

Hypothesis 2: the severity of somatic symptoms increases in conjunction with a higher prevalence of depression and anxiety among students.

Methods

Data collection

This was a cross-sectional study focusing on the physical and mental health of public vocational high school students conducted in June 2024. In China, the senior high school entrance exam serves as a pivotal gateway. Only the top half of candidates secure admission to public academic high schools, while the remainder navigate towards vocational education pathways. This strategic stratification, aimed at optimising the allocation of public educational resources, has been further emphasised by the Chinese government in recent times.

This study was conducted across two provinces: Anhui, an eastern region with a population of 61 million inhabitants and a per capita Gross Domestic Product (GDP) of 10 893 USD, and Sichuan, a western province boasting 82 million residents and a per capita GDP of 10 185 USD. Two vocational high schools in Wuhu City, Anhui Province, alongside one each from Chengdu City and Liangshan Yi Autonomous Prefecture, both within Sichuan Province, were selected using purposive sampling.

Thus, data were collected from four vocational high schools in China using a passive informed consent procedure. Students in Grades 10–11 and without mild to severe cognitive impairment were identified as eligible and were invited to participate. Students were excluded if their parents objected or if they were absent from school during the survey time. At each vocational high school, four investigators organised students to complete a self-administered electronic questionnaire in the school computer room using uniform instruction under the supervision of a computer teacher and a school psychology counsellor. No personal information of students, including those who did not participate in the survey, was recorded. Across the four schools, 85 out of 4500 eligible students were excluded because of parental refusal, 296 students were excluded because they were absent from school during the survey time, and 4119 completed the questionnaire, representing a response rate of 93.30%. There were no missing data.

Patient and public involvement

None.

Outcome variables

Students’ depression was assessed with the Chinese version of the Patient Health Questionnaire-9 (PHQ-9).²⁰ The answer categories were based on a 4-point response scale, with the categories ‘not at all’ (0), ‘several days’ (1), ‘more than half of the days’ (2) and ‘nearly every day’ (3). Therefore, PHQ-9 scores of 10 and above represented moderate-severe depression.²¹ The Cronbach’s alpha in the present study was 0.897.

Students’ anxiety was assessed with the Chinese version of General Anxiety Disorder-7 (GAD-7).²² The answer categories were also based on a 4-point response scale, with the categories ‘not at all’ (0), ‘several days’ (1), ‘more than half of the days’ (2) and ‘nearly every day’ (3). Therefore, GAD-7 scores of 10 and above represented moderate-severe anxiety.²³ The Cronbach’s alpha in the present study was 0.927.

Exposure variables

Somatic symptoms related to faintness, heart pain, nausea/upset stomach, trouble breathing, numbness and feeling weak in the past week were measured by the somatisation subscale of the brief symptom inventory 18.²⁴ One previous study has confirmed that the Chinese version of the brief symptom inventory 18 has acceptable psychometric quality and can be applied to comparative studies of high school students.²⁵ Each item can be answered on a 5-point scale, ranging from 0 = ‘not at all’ to 4 = ‘extremely’. Scores of 7, 12 and 19 represented mild, moderate and severe somatic symptoms.

Covariates

Socio-demographic characteristics collected for the present study included gender, age, family economic status (poor/fair/wealthy), parental marital status and the highest education level of the parent (primary school or below/middle school/high school or above). Based on our previous studies, we found that most students can’t report their family income correctly. So we usually ask them to report a perceived family economic status as poor, fair or wealthy.

Data analysis

Descriptive statistics, χ² tests, t-tests and Fisher’s exact tests were used to explore differences among students with and without depression and anxiety, including somatic symptoms and socio-demographic characteristics. Then, we used logistic regression to examine the associations among depression, anxiety and somatic symptoms, adjusting for socio-demographic characteristics. We employed category collapsing to avoid wide CIs because of the low severe somatic symptoms (n=23). We analysed all data using SPSS (IBM Released 2011. IBM SPSS Statistics for Windows, V.20.0. Armonk, NY: IBM) and assumed a statistical significance level of p<0.05.

Results

The descriptive statistics of the 4119 students included in this study are summarised in table 1. Females comprised approximately 40% of the sample, with a mean age of 16.54 years across all participants. 32.48% of students reported being only children, and 31.27% of students experienced parental divorce or other circumstances. Over 60% of participants indicated that at least one parent had attained education beyond primary school, and the majority perceived their family’s economic status as fair or wealthy. Mild, moderate and severe somatic symptoms were prevalent in 9.57%, 3.84% and 0.56% of students, respectively. Depression and anxiety were identified in 16.29% and 8.04% of the students.

Table 1. Descriptive statistics.

Variables N(%)	Total
Gender
Male	2542 (61.71)
Female	1577 (38.29)
Age mean(SD)	16.54 (0.95)
Singleton
Yes	1338 (32.48)
No	2781 (67.52)
Parental marital status
Married	2831 (68.73)
Divorced or other	1288 (31.27)
Paternal education
Primary school or below	926 (22.48)
Middle school	1688 (40.98)
High school or above	1125 (27.31)
Don't know	380 (9.23)
Maternal education
Primary school or below	1113 (27.02)
Middle school	1520 (36.90)
High school or above	1056 (25.64)
Don't know	430 (10.44)
Family economic status
Poor	950 (23.05)
Fair	2475 (60.09)
Wealthy	694 (16.85)
Somatic symptoms
No	3544 (86.04)
Mild	394 (9.57)
Moderate	158 (3.84)
Severe	23 (0.56)

Table 2 presents regression coefficients for depression on somatic symptoms with adjustment for socio-demographic characteristics. Compared with students without somatic symptoms, students with mild (OR=8.15; 95% CI=6.46 to 10.29) and moderate or severe (OR=19.24; 95% CI=13.66 to 27.10) somatic symptoms reported significantly higher prevalence of depression.

Table 2. Regression coefficients for depression on somatic symptoms with adjustment for social-demographic characteristics.

Variables N(%)	OR (95% CI)	P value
Somatic symptoms
No	1.00
Mild	8.15 (6.46 to 10.29)	<0.001
Moderate or severe	19.24 (13.66 to 27.10)	<0.001
Gender
Male	1.00
Female	1.34 (1.10 to 1.62)	0.003
Age Mean(SD)	0.97 (0.87 to 1.06)	0.472
Singleton
Yes	1.00
No	1.04 (0.84 to 1.27)	0.747
Parental marital status
Married	1.00
Divorced or other	1.30 (1.06 to 1.59)	0.010
Paternal education
Primary school or below	1.00
Middle school	0.79 (0.61 to 1.04)	0.088
High school or above	1.08 (0.79 to 1.49)	0.627
Don't know	1.11 (0.69 to 1.79)	0.658
Maternal education
Primary school or below	1.00
Middle school	0.95 (0.73 to 1.24)	0.708
High school or above	1.17 (0.85 to 1.61)	0.344
Don't know	1.05 (0.66 to 1.66)	0.833
Family economic status
Poor	1.00
Fair	0.53 (0.42 to 0.66)	<0.001
Wealthy	0.37 (0.27 to 0.51)	<0.001

Table 3 presents regression coefficients for anxiety on somatic symptoms with adjustment for socio-demographic characteristics. Compared with students without somatic symptoms, students with mild (OR=6.70; 95% CI=5.03 to 8.93) and moderate or severe (OR=16.84; 95% CI=11.89 to 23.83) somatic symptoms reported significantly higher prevalence of anxiety.

Table 3. Regression coefficients for anxiety on somatic symptoms with adjustment for social-demographic characteristics.

Variables N(%)	OR (95% CI)	P value
Somatic symptoms
No	1.00
Mild	6.70 (5.03 to 8.93)	<0.001
Moderate or severe	16.84 (11.89 to 23.83)	<0.001
Gender
Male	1.00
Female	1.27 (0.99 to 1.63)	0.060
Age mean(SD)	0.93 (0.82 to 1.06)	0.298
Singleton
Yes	1.00
No	0.95 (0.73 to 1.24)	0.701
Parental marital status
Married	1.00
Divorced or other	1.36 (1.05 to 1.76)	0.022
Paternal education
Primary school or below	1.00
Middle school	0.84 (0.60 to 1.18)	0.315
High school or above	0.74 (0.49 to 1.12)	0.154
Don't know	0.72 (0.38 to 1.37)	0.321
Maternal education
Primary school or below	1.00
Middle school	0.79 (0.56 to 1.11)	0.174
High school or above	1.31 (0.87 to 1.95)	0.194
Don't know	0.77 (0.41 to 1.42)	0.401
Family economic status
Poor	1.00
Fair	0.57 (0.43 to 0.75)	<0.001
Wealthy	0.58 (0.39 to 0.86)	0.007

Discussion

To the best of our knowledge, this is the first study to use a large-scale school-based survey to examine the associations between somatic symptoms, depression and anxiety among Chinese adolescents. Our findings reveal significant positive association: as the prevalence of somatic symptoms increases, so too does the prevalence of depression and anxiety among students, aligning with our initial hypotheses. Additionally, the severity of somatic symptoms appears to rise concurrently with the prevalence of these psychiatric conditions, which is also supported by other studies.²⁶

In light of our findings, we strongly advocate for an expansion of the nationwide screening initiative for mental health among children and adolescents put forward by the Ministry of Education of China.¹⁶ This enhancement should encompass an evaluation of somatic symptoms, which are often overlooked yet crucial indicators of mental health challenges. A common issue with existing depression and anxiety scales is a lack of enough items for accurately assessing and monitoring individual physical symptoms.²⁷ By incorporating this dimension, we can create a more holistic and sensitive diagnostic framework that is better equipped to identify the subtle and varied presentations of depression among the youth in China. And this expansion is also a significant step towards dismantling the pervasive stigma that surrounds mental health discussions.²⁸ The inclusion of somatic symptom assessment in the screening process can help normalise conversations about mental health by highlighting the physical manifestations of emotional distress. This, in turn, can encourage more young people to seek help without fear of judgement or misunderstanding.

The Ministry of Education of China also demands that by 2025, over 95% of primary and middle schools should be equipped with at least one psychological teacher, aimed at addressing the mental health needs of students. While a fortunate few schools can afford to engage dedicated psychological teachers, the majority face constraints, often resorting to appointing teachers with limited specialised training to concurrently fulfil this crucial role. These non-specialist mental health providers often find themselves insufficiently equipped to tackle complex issues such as depression and suicidal ideation among students. Given that somatic symptoms are more overt and easier to discuss,²² we recommend that schools prioritise attention to these symptoms, especially in settings where psychological teachers may not yet be fully qualified as counsellors. This strategy can facilitate initial conversations, making the process less intimidating for both the provider and the student. Additionally, it becomes more straightforward for these psychological teachers to advise caregivers to seek further medical consultations for their children.

By recognising the physical signs of depression, healthcare providers can offer more targeted interventions and support, leading to improved treatment outcomes.²¹ It also empowers educators, parents and peers to be more proactive in identifying and addressing the mental health needs of young individuals. This approach not only benefits those directly affected by depression but also contributes to a broader societal shift towards greater empathy and understanding of mental health issues.

This study has several strengths. First, this is the first study to examine the associations between somatic symptoms, depression and anxiety among Chinese adolescents in non-clinical settings. Our findings can help optimise large-scale school-based students’ mental health screening. Second, by conducting a population-based survey that included students from four schools in two geographically distant provinces (Anhui Province and Sichuan Province), sampling bias issues are expected to be minimal.

The present study has several limitations that require consideration. First, although we tried our best to enrich the sample diversity, we only included students from vocational high schools. Therefore, caution should be taken when generalising our findings to other age groups or academic high school students. Second, the tools we selected measured students’ somatic symptoms in the past week but depression and anxiety symptoms in the past 2 weeks. Although we don’t know to what direction and extent this time lag can influence our results, we believe this won’t twist our findings. What’s more, we were unable to address potential confounding from organic causes such as gastrointestinal conditions, sleep disorders or neurodevelopmental traits. Thirdly, this is a self-administered survey, and our results exclusively relied on students’ self-reporting. Thus, problems like shared method variance and self-presentation should be noticed. Fourthly, students who were absent during the survey period (n=296) were excluded from the analysis. Based on our ongoing engagement with local schools and stakeholders, we understand that the majority of absences on the survey day were attributable to routine sick leave (eg, mild illnesses like colds or fever) or personal/family matters. Still, this operation may introduce systematic bias and underestimate the prevalence and severity of depression and anxiety.

Conclusions

In this cross-sectional study, we found that students who experience severe somatic symptoms have a higher prevalence of depression and anxiety. Thus, the systematic screening for somatic symptoms in adolescents should be a requisite evaluation procedure in school-based students’ mental health screening. It will enhance the early detection and treatment of depression and anxiety, foster open dialogue and ultimately pave the way for a more supportive and inclusive society for Chinese youth. By integrating somatic symptom assessment into mental health initiatives, we can bridge the gap between physical and mental health, promoting a more comprehensive understanding of adolescent well-being and fostering a culture of empathy and support.

Data availability statement

Data are available upon reasonable request.