Introduction

Global trends indicate a distressing rise in adolescent obesity, positioning it as a significant global health challenge. This alarming increase not only jeopardizes the well-being of young populations but also elevates the risk of early mortality in adulthood. A projection by the World Obesity Federation underscores a concerning future, particularly for China. By 2030, the nation is expected to lead the world in childhood obesity, with over one million affected children, surpassing other high-prevalence countries such as India, the United States, and Indonesia. This trend necessitates urgent and tailored public health strategies to mitigate the impending health crisis. Based on whether metabolic abnormalities accompanied it, the population with obesity is further divided into metabolically healthy obesity (MHO) and metabolically unhealthy obesity (MUO). A comprehensive analysis involving 12,346 children and adolescents across seven provinces revealed that metabolically healthy obesity (MHO) and metabolically unhealthy obesity (MUO) were present in 1.96% and 3.03% of the participants, respectively1. Concurrently, another study highlighted that 13.4% of Chinese children and adolescents suffer from conditions like metabolically unhealthy overweight (MUOW) and MUO, underscoring the public health challenge these metabolic disorders pose in this age group2. In contrast, in the adolescent population with obesity, the prevalence of MHO ranges from 19.7 to 41.0%3,4, and the prevalence of MUO ranges from about 29.1–71.7%3,5,6. Studies have pointed out that younger age, insufficient sleep, excessive screen use7, poor dietary patterns8, and poorer mental health9 are important risk factors for the development of MHO and MUO in children and adolescents.

The comorbidity problem of depression and weight gain in children and adolescents is an unavoidable issue in clinical diagnosis and treatment. Research indicates that within this demographic, 15.5% struggle with overweight and 10.4% with obesity10. These conditions are compounded by high fasting glucose levels, elevated triglycerides, and intensified depressive symptoms, all of which stand as independent predictors of obesity in these children and adolescents. Furthermore, a study emphasized that among young people with depression, the prevalence of overweight is as high as 52.9%, while the prevalence of obesity is only 4.1%, and obesity was closely related to the patient’s age, age of onset and gender11. Besides, the implications of overweight and obesity extend beyond physical health, influencing mental health risks as well. Notably, there is a significant correlation between increased body weight and the risk of suicidal ideation and attempts12. In addition to adolescents with overweight or obesity, adolescents who self-perceived as overweight or having abnormal body weight also had significantly higher rates of suicide13. The increased suicide risk among adolescents with obesity and depressive disorders may be related to multiple factors. First, chronic inflammation and metabolic disorders exacerbate the imbalance of neurotransmitters and weaken the patients’ ability to regulate emotions14. Secondly, obesity and depression cause social discrimination, leading to low self-esteem and physical shame in patients, which in turn intensifies feelings of despair15. Thirdly, compensatory behaviors such as overeating and lack of exercise form a vicious circle16.

There is a significant bidirectional association between depression and adolescent obesity, forming a mutually reinforcing vicious cycle. On the one hand, depression increases the risk of obesity through multiple pathways, such as hormonal secretion disorders and neurotransmitter level imbalances17. Furthermore, the decline in exercise motivation and the disorder of circadian rhythm further exacerbates the imbalance of energy metabolism18. On the other hand, obesity induces depressive symptoms through both physiological and psychological mechanisms18. For instance, adipose tissue can release pro-inflammatory factors, affecting mood regulation14. Obesity-related insulin resistance and other factors interfere with the synthesis of monoamine neurotransmitters19. Moreover, social stigmatization and body image distress significantly increase psychological stress15. This two-way effect is particularly prominent during adolescence. Therefore, it is of great significance to study obesity and metabolic problems in children and adolescents with depressive disorders.

Despite the growing interest in metabolic health outcomes among young populations, research specifically investigating MHO and MUO within adolescent groups facing depressive disorders remains scant. This study aims to bridge this knowledge gap by classifying children and adolescents into four distinct categories based on their body mass index (BMI) levels and metabolic health markers: Metabolically Healthy Non-Obesity (MHNO), Metabolically Unhealthy Non-Obesity (MUNO), MHO, and MUO. The objective is to delineate the specific factors independently associated with MHO and MUO in this demographic.

Methods

Design, population and sample

This is a cross-sectional survey, the patients were children and adolescents hospitalized (from January-December 2021) in the Third People’s Hospital of Fuyang. Inclusion criteria: (1) age: 8–18 years old; (2) meeting the International Classification of Diseases (10th edition) (ICD-10) diagnostic criteria for depressive disorders (ICD-10 codes were diagnoses for F32 and F33 and their affiliates), and diagnosed independently by two senior psychiatrists. Exclusion criteria: (1) co-morbid ICD-10 other psychiatric disorders (e.g., attention deficit hyperactivity disorder, autism spectrum disorder, obsessive compulsive disorder, conduct disorders, eating disorders, etc.); (2) in pregnancy or breastfeeding; and (3) combined with severe physical illnesses (e.g., neurological, respiratory, and immune system diseases, etc.). The Ethics Committee of the Third People’s Hospital of Fuyang sanctioned this research (approval number: 2018-340-10). The researchers informed the patients and their guardians of the research purpose, main research process, and data utilization of this study. Both the patients and their guardians agreed to participate in the project and signed a paper version of the informed consent form for retention. The estimated prevalence of MUO among children and adolescents with depressive disorders was set at 5%, the allowable error was selected as 5%, and the confidence level was set at 0.95. The minimum sample size to be investigated was calculated using the PASS software, which was 334 people. A total of 515 research subjects were finally included in the study.

General information and biochemical indicators

General information

Collected demographic data of the enrolled subjects, such as gender, age, education level. Clinical data were obtained through the electronic medical record system and both the patients and their guardians, and were cross-checked with each other, such as age of onset, age of first hospitalization, disease duration, medication category, medication dosage. The dose of antidepressants taken by the patients was converted to fluoxetine equivalents20. In line with the protocols established by China’s Screening for Malnutrition in School-Aged Children and Adolescents alongside the Screening Form for Overweight and Obesity in School-Aged Children and Adolescents, the classification of obesity among the study’s participants was conducted. This categorization was based on comparisons of BMI across various age brackets10.

Biochemical indicators

Blood samples were collected from fasting patients for analysis. After the patients joined the project and completed the collection of general and clinical data, the following morning, between 6:00 and 7:00 a.m., hematological parameters were assessed, including fasting blood glucose (FBG), total cholesterol (TC), triglycerides (TG), high-density lipoprotein (HDL), low-density lipoprotein (LDL), glutamic pyruvic transaminase (ALT), glutamic oxaloacetic transaminase (AST), uric acid, and creatinine. Additionally, the triglyceride-glucose (TyG) index, an indicator of insulin resistance, was computed using the equation: ln [(fasting triglyceride (mg/dL) × fasting blood glucose (mg/dL))/2]21.

Hamilton depression scale (HAMD)

The 24-item HAMD evaluated the severity of depressive symptoms in patients. Suicidal thoughts were specifically measured through item 3, labeled “Suicide”22. If the patient has “feeling that life is meaningless”, “wishing that they have died, or often thinking about things related to death” or “negative ideas (suicidal thoughts)”, it is determined that they have suicidal ideation. If a patient has “serious suicidal behavior”, it is determined that they are accompanied by suicidal behavior.

Determination of MUO, MHO, MUNO and MHNO

The patient’s BMI level was in the obesity category and was not accompanied by any of the following conditions23: (1) HDL-C > 40 mg/dL (or > 1.03 mmol/L); (2) TG ≤ 150 mg/dL (or ≤ 1.7 mmol/L); (3) systolic blood pressure (SBP) and diastolic blood pressure (DBP) ≤ 90th percentile; and (4) FBG ≤ 100 mg /dL (or ≤ 5.6 mmol/L). The MUO group was determined if any of the metabolic abnormality indicators were present. MHO refers to a condition where a patient meets the diagnostic criteria for obesity but shows no abnormalities in the above-mentioned metabolic indicators. MUNO refers to a situation where a patient does not meet the diagnostic criteria for obesity, but any of the above metabolic indicators is abnormal. MHNO refers to the situation where the patient does not meet the diagnostic criteria for obesity and there are no abnormalities in the above metabolic indicators.

Statistical methods

Firstly, to assess the distribution of continuous variables, normality tests were conducted. Variables that followed a normal distribution were analyzed using One-Way ANOVA to evaluate differences across groups. For variables not meeting this criterion, the Kruskal Wallis Test was applied. Secondly, the χ2 test was employed for the examination and comparison of categorical variables. In addition, after controlling the confounding variables, the covariance analysis was performed to compare the variable which was statistically different by one-way analysis of variance between MHO and MUO group or not. Thirdly, regression analysis was used to explore the independent correlates of MHO or MUO in adolescent patients with depressive disorders, using whether the patient was MHO or MUO as the dependent variable (assignment: yes = 1, no = 0) and variables (assigned variables were age, gender, age of onset, age of first hospitalization, BMI, SBP, DBP, blood indicators and medication use) included in this study that were statistically different by one-way analysis of variance as the independent variables. The test level was two-sided α < 0.05.

Results

Demographic data and biochemical indicators

The detection rates of MHNO, MUNO, MHO, and MUO were 63.3% (326/515), 25.0% (129/515), 3.7% (19/515), and 8.0% (41/515), respectively. Statistical analysis comparing various groups revealed marked differences across numerous parameters including age, gender, onset age, initial hospitalization age, BMI, SBP, DBP, FBG, TC, TG, HDL, LDL, TyG index, ratios of TC to HDL, TG to HDL, alongside levels of ALT, AST, and uric acid. Further subgroup evaluations indicated significant distinctions specifically in onset age, age at first hospitalization, BMI, LDL, TyG index, and uric acid between the MHO and MHNO subjects. In the comparison between MUO and MHNO groups, significant disparities were also observed in BMI, SBP, DBP, FBG, TC, TG, HDL, LDL, TyG index, TC/HDL ratio, TG/HDL ratio, ALT, AST, and uric acid, all demonstrating statistical significance (P < 0.05). Secondly, compared to the MHO group, patients in the MUO group showed greater age, age of onset and age of first hospitalization, higher SBP and DBP, higher levels of TG, TC/HDL, TG/HDL, TyG index and AST, and low levels of HDL. See Table 1.

Table 1 Comparison of general information and biologic indices of different groups.
Full size table

Binary logistic regression analysis of MHO and MUO in adolescent depressive disorders patients

Using whether the patient was MHO or MUO as the dependent variable (assignment: yes = 1, no = 0) and variables that were statistically different by univariate analysis as the independent variables, binary regression analysis showed that a high level of LDL (OR = 2.76, P = 0.007) was an independent risk factor for MHO; whereas, older age at onset of the disorders (OR = 0.69, P = 0.002) was a protective factor for MHO. In addition, high levels of TC/HDL (OR = 2.66, P = 0.003), TG/HDL (OR = 1.81, P = 0.034), AST (OR = 1.03, P < 0.001), and uric acid (OR = 1.004, P = 0.018) were independent risk factors for MUO. See Table 2.

Table 2 Binary logistic regression analysis of MHO and MUO in children and adolescents with depressive disorders.
Full size table

Discussion

Our research focused on exploring the occurrence and determinants of MHO and MUO among children and adolescents suffering from depressive disorders. The findings revealed that 3.7% of the subjects were categorized under MHO, while 8.0% fell into the MUO category. In another study within a similar demographic, MHO was observed at 3.5% and MUO at 5.6%7. Interestingly, data from an Iranian adolescent cohort showed a higher prevalence of MHO at 8.9% and a lower rate of MUO at 1.8%, respectively24, . While another study in Iran indicated that MHO and MUO were 3.4% and 8.2%25.This variation underscores the influence of geographical and cultural factors on the prevalence rates of these conditions among children and adolescents with depressive disorders. Former studies demonstrate that there is a difference between children and adolescents with depressive disorders with higher MHO and MUO than the general adolescent population20. In this research, we observed that among children and adolescents with obesity diagnosed with depressive disorders, 31.7% (19 out of 60 participants) exhibited MHO. Conversely, 68.3% (41 out of 60 participants) were classified as having MUO. This distribution aligns closely with results from prior investigations into the subject. Variability in MHO and MUO rates across different studies can likely be attributed to ethnic variations, regional dietary habits, the size of the study sample, and the specific benchmarks used for defining research indicators.

In the conducted research, compared to the MHO group, patients in the MUO group exhibited greater age, age of onset and age of first hospitalization, higher SBP and DBP, high levels of TG, TC/HDL, TG/HDL, TyG index and AST, and low levels of HDL. Previous studies have shown that compared to the normal-weight population, MHO is associated with younger age, lower levels of physical activity, and puberty-onset obesity or overweight factors26. In logistic regression, older onset age is a protective factor for MHO. Previous studies have pointed out that early-onset depressive disorders are often characterized by a long history of depressive episodes, duration of symptoms, and higher levels of neuroticism, and are associated with the patient’s higher BMI. This makes an earlier age of onset a related clinical factor for obesity in young patients11. On the contrary, older patients have stronger self-control ability, are better at regulating emotions, and maintain the balance before eating and exercising. Previous literature found that the level of insulin resistance can be an essential predictor of MHO27. Clinical trials noted that the MHO group exhibited younger age, waist circumference, and insulin resistance levels than the MUO group4. Our research results provide certain references for the study of related mechanisms.

Moreover, in the current analysis, the TyG index emerged as notably elevated in the MUO children and adolescents compared to their MHO counterparts, a finding consistent with existing literature28. Additionally, the TC/HDL ratio not only showed a strong correlation with the TG/HDL ratio but also proved to be an effective indirect marker for assessing insulin resistance29,30. This relationship underscores its utility in reliably predicting metabolic health status. Furdela et al. showed that the TyG index, TG/HDL, is one of the valuable combinations of predictors of MUOW and MUO in children and adolescents3. Insulin-sensitive MHO patients are at less risk of abnormal metabolic indices than MUO patients with higher levels of insulin resistance31. It can be seen that in addition to individual biological indicators, the combination of TyG index, TC/HDL, TG/HDL, and other indicators are equally essential references for the diagnosis of MHO or MUO.

Elevated uric acid levels were consistently observed in MUO children and adolescents compared to their MHO counterparts, as reported in studies32,33. In contrast, children and adolescents maintaining normal weight exhibited markedly lower uric acid concentrations32. Additionally, reduced uric acid and ALT levels were identified as significant correlates with the onset of metabolic health in individuals with obesity34. Furthermore, serum uric acid alongside insulin resistance emerged as critical biochemical markers for predicting MUO5. The present study similarly pointed out that high uric acid levels were one of the independent risk factors for MUO. In addition, blood uric acid levels in children and adolescents showed a significant positive correlation with high levels of BMI or obesity, among others35, and uric acid levels were significantly reduced after weight loss in children and adolescents with obesity36. In addition, obesity and abnormal glucose and lipid metabolism, as essential risk factors for hyperuricemia, positively contribute to their higher prevalence37. The reduction of MUO risk is accompanied by a corresponding decrease in the risk of hyperuricemia38. Thus, it is evident that there is a strong association between uric acid levels and MUO.

Metabolically healthy obesity is not entirely harmless, and over time, it may transition to a metabolically unhealthy phenotype, while increased MUO increases future cardiovascular morbidity and premature mortality in early adulthood39. Significant differences in BMI, SBP, DBP, TG and HDL levels were found between the two groups of patients due to MHO and MUO limiting conditions40. In addition, both lipocalin and ApoA-1 levels were significantly higher in MUO patients compared to MHO patients33. Secondly, it was noted that poor dietary composition (e.g., fried foods, fast foods, less plant-based foods) and uncontrolled eating behaviors were more pronounced in MUO children and adolescents compared to MHO children and adolescents8,41. In contrast, after diet combined with exercise intervention for MHO or MUO children and adolescents, their BMI, waist circumference, SBP, DBP, TG, and TC were effectively improved, and the benefits were more pronounced in the MUO group42. On the contrary, children and adolescents with good dietary patterns, such as implementing the Mediterranean dietary pattern and consuming of fewer snacks or convenience-type foods, are less likely to develop MUO24,43. Research indicates that individuals characterized as MUO exhibit notably higher oxidative stress levels than their normal-weight counterparts44. Additionally, levels of high-sensitivity C-reactive protein (HsCRP), an inflammation marker, are significantly elevated in MUO children and adolescents relative to those of normal weight32. Furthermore, it has been documented that poor dietary choices contribute to heightened inflammation within the body45. Strong evidence suggests that there may be a continuum of action between unhealthy dietary patterns, disordered inflammatory mechanisms, and the development of MUO.

Research indicates that various lifestyle and psychological factors play crucial roles in the health outcomes of children and adolescents, particularly in relation to MHO and MUO. Elevated socioeconomic status, diminished physical activity, increased screen exposure, and a diet low in fruits correlate with heightened risks of both MHO and MUO26,46. Conversely, children and adolescents who experience lower levels of psychological stress, demonstrate higher resilience, and exhibit better school adjustment are less likely to develop MUO47. Moreover, the contemporary dietary patterns among children and adolescents, characterized by a high intake of junk food, are linked to escalated symptoms of depression, anxiety, and stress, alongside deteriorating sleep quality48,49. Importantly, the relationship between MUO and depressive symptoms appears to be bidirectional, suggesting that each condition may exacerbate the other. In the context of obesity, both depression and anxiety are more prevalent among children and adolescents. Notably, the intensity of depressive symptoms is considerably more severe in children and adolescents with unhealthy metabolic profiles9,50. This interplay of dietary habits, psychological well-being, and metabolic health underscores the complex nature of adolescent obesity and highlights the need for integrated approaches in treatment and prevention strategies. Moreover, depressive symptoms may exacerbate the risk of obesity through multiple mechanisms. First, chronic stress causes a continuous increase in cortisol, promoting visceral fat deposition51. Secondly, disorders of monoamine neurotransmitters (such as 5-HT) can cause hyperappetite and increase the intake of high-calorie foods52. Thirdly, emotional eating, as a psychological compensatory mechanism, leads to excessive energy intake53. Fourth, the decline in motor motivation and circadian rhythm disorders reduce energy expenditure16. Fifth, insulin resistance and leptin resistance accompanied by long-term depression disrupt the regulation of energy homeostasis54. These factors jointly contribute to a positive energy balance and ultimately promote the occurrence and development of obesity. Clinical intervention requires an integrated strategy. Cognitive behavioral therapy should be implemented simultaneously to improve emotion regulation55. Exercise-nutrition intervention should be combined to break metabolic disorders56. Attention should also be paid to the construction of medical, rehabilitation and social support systems to achieve the coordinated management of depressive symptoms and obesity.

Study limitations

There are several limitations to this paper. First, this study was cross-sectional, and it was impossible to clarify the causal relationship between the included study variables and MHO or MUO. Second, this study highlights the study population by targeting children and adolescents with depressive disorders. Still, our research only included some Han children and adolescents, with a small sample size and a single-center survey, the representativeness is more limited, and a multicenter, large-sample survey is needed to explore uric acid levels and related factors in this population. Thirdly, although comorbidities are common among children and adolescents with emotional problems, this study only included patients with a single diagnosis of depressive disorder. Nevertheless, the possibility of comorbidity of other mental illnesses in the study population cannot be completely ruled out, and using a standardized screening questionnaire might be more helpful. Finally, for the metabolic or obesity issues of the population, there are many other confounding factors that may have an impact, such as physical activity, sleep, dietary structure, socioeconomic status, etc., which were not included in this study. It awaits further discussion in specific research projects.

Applications to clinical practice

In this study, the research variables included not only BMI, FBG, SDP, DBP, and blood lipid indicators, but also liver function indicators (ALT and AST), TyG index, TC/HDL ratio, TG/HDL ratio, etc. The included variables are more complete than those in previous studies and have certain reference value. Obesity has emerged as a significant global public health challenge. It is increasingly recognized that not all individuals with obesity experience the same metabolic consequences. Among these, some exhibit MHO while others suffer from MUO, with the latter group facing heightened risks of cardiovascular diseases. In light of these findings, it becomes imperative for medical practitioners to closely monitor obesity trends, including distinctions between MHO and MUO. Early and targeted interventions are crucial to address the underlying risk factors and to avert the progression from MHO to MUO. This proactive approach in clinical settings can significantly contribute to mitigating the broader impact of obesity on public health.

Conclusion

Our research highlights a concerning correlation between depressive disorders in children and adolescents and elevated incidences of both MHO and MUO. Analysis reveals that children and adolescents categorized under MUO typically present with advanced age, earlier age of onset, and younger age at first hospitalization. Additionally, they show elevated systolic and diastolic blood pressures, higher levels of TG, TC/HDL ratio, TG/HDL ratio, TyG index, and AST, along with reduced levels of HDL compared to their MHO counterparts. Moreover, our findings indicate that high LDL levels independently predict MHO. Conversely, an older age at onset appears to act as a protective factor against MHO. For MUO, elevated TC/HDL, TG/HDL, AST, and uric acid levels emerge as independent risk factors. In medical practice, the diagnosis and treatment of patients’ diseases and the monitoring of abnormal indicators can effectively improve the condition of metabolic disorders.