Association Between Metabolic Syndrome and Psychiatric Morbidity in a Nationwide Taiwanese Population Study

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Thank you for submitting your manuscript to Nutrients.

Abstract
Please specify the study design (cross-sectional or longitudinal).
The authors calculated the odds ratio (OR). There are a large number of participants. Could a longitudinal study be conducted with these participants, given that measurements were taken more than once? Could a relative risk (RR) analysis be performed?
The choice of terminology for "psychiatric morbidity" is crucial, especially in the field of psychiatry. Individuals diagnosed by a physician are indeed considered to have a mental illness. However, when using questionnaires for assessment, it's important to be more cautious, as these individuals are not formally "diagnosed." Please consider using a more flexible and comprehensive term such as "mental disorder."

Conclusion
In a cross-sectional study, definitive statements like the one in the first sentence can cause considerable controversy. Did you confirm the prevalence of psychiatric morbidity in individuals with metabolic syndrome?
Similarly, if hypertension, increased waist circumference, and hypertriglyceridemia increase psychiatric morbidity, there should be an explanation for this. Therefore, the wording should be softened and made less assertive.

-------------------------

Introduction
The introduction is presented as a general overview. To emphasize the necessity of this study, please include specific results from previous studies when citing them. The introduction should explain that this study was designed to address the limitations of previous research. Also, many of the cited references are somewhat outdated. Please use references from within the last 5 years whenever possible. Please also ensure that the reference style conforms to MDPI guidelines.

Methods
I assume that the initial data pool was larger to obtain the final 121,575 participants. Therefore, please add a diagram showing the inclusion and exclusion criteria.

Results
Table 1: Please express continuous variables, including age and BMI, to one decimal place. Does the age range start from adulthood (20 years old)? Is there no information on monthly income? Is information available on employment status and religious affiliation? Is there any information on physical activity?
Metabolic syndrome, as well as psychiatric morbidity, is associated with various physical, social, and economic issues. Please provide as much diverse information as possible that has been collected. While having a large amount of data is an advantage, it is better to have more specific information about the participants, even if the amount of data analyzed is reduced depending on the direction and purpose of the study.
Table 2: What is BMI compared to?
Tables 4 and 5 could be better presented using bar graphs.
Consider using "—" instead of "to" for the 95% CI.

Discussion
The relationship between metabolic syndrome and psychiatric problems is very complex. Psychological distress can be caused by metabolic syndrome, and mental health problems can lead to difficulties in controlling diet. Considering these complex issues, the authors' discussion is very brief and lacks depth. Fortunately, numerous studies have already been conducted on these two issues. For a more in-depth discussion, please refer to more studies and interpret the results of this study accordingly. Also, please add possible interpretations regarding the biomechanisms based on the existing knowledge from previous research.

Author Response

Abstract

Comments: Please specify the study design (cross-sectional or longitudinal)

• Response: We agree and now explicitly state this is a cross-sectional study.
• We further revised the abstract as follows:” This study aimed to examine the association between MetS and psychiatric morbidity in a nationwide Taiwanese adult cohort using a cross-sectional design.” (Page 1, abstract)

 

Comments: The authors calculated the odds ratio (OR). There are a large number of participants. Could a longitudinal study be conducted with these participants, given that measurements were taken more than once? Could a relative risk (RR) analysis be performed?

• Response: We agree this is important. However, the present analysis used a cross-sectional snapshot from the community screening dataset; we therefore used logistic regression and report ORs. We now explicitly note that prospective/longitudinal designs (and RR/incident analyses) are needed and are a priority for future work.
• We further revised the abstract as follows:” In this large Taiwanese community cohort, MetS was associated with a modest increase in psychiatric morbidity. Given the cross-sectional design and symptom-based screening measures, these findings should be interpreted as associations rather than evidence of causality; prospective studies are warranted to clarify temporal relationships and mechanisms.” (Page 2, abstract)

 

Comments: The choice of terminology for "psychiatric morbidity" is crucial, especially in the field of psychiatry. Individuals diagnosed by a physician are indeed considered to have a mental illness. However, when using questionnaires for assessment, it's important to be more cautious, as these individuals are not formally "diagnosed." Please consider using a more flexible and comprehensive term such as "mental disorder."

• Response: We appreciate your careful consideration of terminology. We respectfully retained the term psychiatric morbidity because, in epidemiological and public health research, it refers to the presence of clinically relevant psychological symptom burden identified through validated screening instruments rather than exclusively physician-confirmed psychiatric diagnoses. In fact, the term mental disorder typically implies a formal diagnostic classification (e.g., DSM or ICD criteria), whereas psychiatric morbidity is commonly used in population-based studies to capture non-diagnostic but clinically meaningful mental health burden detected using standardized questionnaires. Therefore, replacing the term with mental disorder would risk overstating the diagnostic certainty of the present study. To ensure clarity and avoid overinterpretation, we have explicitly defined psychiatric morbidity in the manuscript as screening-based symptom burden rather than confirmed psychiatric disease.
• We further revised the Abstract as follows: “Psychiatric morbidity was defined as depressive and/or anxiety burden identified by validated screening instruments (Patient Health Questionnaire-2 score ≥ 3 or Generalized Anxiety Disorder-2 score ≥ 3) or self-reported physician-diagnosed depression.” (Page 1, abstract)
• We further revised the Methods section as follows: “In this research, psychiatric morbidity was defined as depressive and/or anxiety burden identified by validated screening instruments (PHQ-2 ≥ 3 or GAD-2 ≥ 3) or self-reported physician-diagnosed depression.” (Page 4, methods)
• We further revised the Limitation section as follows: “Third, our outcome definition combined screening-positive symptoms with self-reported physician-diagnosed depression; this approach may increase sensitivity but also introduces heterogeneity, and the self-reported diagnosis could not be verified in this de-identified dataset. Accordingly, psychiatric morbidity should be interpreted as a population-level mental health burden, reflecting screening-detected symptoms and self-reported prior diagnosis rather than clinically confirmed disorders.”(Page 9, limitations)

Conclusion
Comments: In a cross-sectional study, definitive statements like the one in the first sentence can cause considerable controversy. Did you confirm the prevalence of psychiatric morbidity in individuals with metabolic syndrome?

• Response: We agree. We now (i) report the symptom prevalence by MetS status in the Abstract/Results, and (ii) soften causal language to “associated with” and emphasize cross-sectional limitations.
• We further revised the Results section of abstract as follows: “The prevalence of psychiatric morbidity was higher among participants with MetS than those without MetS (5.0% vs 4.3%).” (Page 1, abstract)
• We further revised the conclusion section of abstract as follows: “In this large Taiwanese community cohort, MetS was associated with a modest increase in psychiatric morbidity.” (Page 2, abstract)

Comments: Similarly, if hypertension, increased waist circumference, and hypertriglyceridemia increase psychiatric morbidity, there should be an explanation for this. Therefore, the wording should be softened and made less assertive.

• Response: We agree. We expanded the Discussion to provide plausible biological/behavioral explanations and clearly label them as hypothesis-generating (inflammation, HPA axis, lifestyle/diet-related pathways), while avoiding causal language.
• We further revised the Discussion section as follows: “Prior research suggests that these associations may operate through interconnected bi-ological and behavioral pathways. At the biological level, chronic low-grade inflam-mation, oxidative stress, autonomic imbalance, and hypothalamic–pituitary–adrenal axis dysregulation have been proposed as shared mechanisms contributing to both cardiometabolic dysfunction and mood-related symptoms [31–34]. At the behavioral level, depressive and anxiety symptoms may adversely affect diet quality, physical ac-tivity, sleep, and medication adherence, whereas cardiometabolic symptoms and weight-related stigma may increase psychological distress [35–37]. Within a cross-sectional framework, our findings are therefore more consistent with shared-pathway hypotheses than with a specific causal direction. From a nutritional perspective, diet quality represents a plausible upstream factor: dietary patterns high in ultra-processed foods and added sugars are associated with cardiometabolic risk, whereas healthier dietary patterns (e.g., higher intake of fiber, fruits/vegetables, and unsaturated fats) have been linked to a lower depressive symptom burden [38–41]. Although detailed dietary data were not available in this dataset, future Taiwanese cohort studies incorporating nutritional assessment may clarify whether diet mediates or modifies the MetS–mental health relationship.” (Page 9, Discussion)

 

Introduction
Comments: The introduction is presented as a general overview. To emphasize the necessity of this study, please include specific results from previous studies when citing them. The introduction should explain that this study was designed to address the limitations of previous research. Also, many of the cited references are somewhat outdated. Please use references from within the last 5 years whenever possible. Please also ensure that the reference style conforms to MDPI guidelines.

• Response: We revised the Introduction to better articulate (i) what prior work has shown, (ii) why Taiwanese large-scale evidence is still valuable, and (iii) what a cross-sectional design can and cannot address. We also added nutrition-relevant framing for Nutrients.
• We further revised the Introduction as follows: “Recent meta-analyses have reported that individuals with depression have ap-proximately 1.3–1.5-fold higher odds of developing MetS compared with those without depression, while MetS has also been associated with a higher prevalence of depressive and anxiety symptoms [3,5,7]. These associations may be explained by shared mecha-nisms, including chronic systemic inflammation [7], dysregulation of the hypothalam-ic–pituitary–adrenal (HPA) axis [8,9], and unhealthy lifestyle behaviors such as poor diet quality and reduced physical activity [10,11]. However, most prior studies have relied on clinical samples, relatively small cohorts, or single disease endpoints, which may limit generalizability to the general population.

Although consistent associations have been observed, temporal direction and population-level burden remain unclear because many studies were conducted in Western populations or lacked large community-based samples. Cultural norms, die-tary habits, and social environments differ across populations and may modify both risk and clinical expression of these conditions. Moreover, large-scale community screening data can estimate real-world symptom burden but cannot establish causality; therefore, clarifying the magnitude and pattern of association in such populations re-mains clinically informative.

Therefore, we aimed to quantify the prevalence of psychiatric morbidity according to MetS status and evaluate the associations of individual MetS components and cu-mulative metabolic burden with psychiatric morbidity in a population-based cohort.” (Page 2, introduction)

 

Methods
Comments: I assume that the initial data pool was larger to obtain the final 121,575 participants. Therefore, please add a diagram showing the inclusion and exclusion criteria.

• Response: Added Figure 1 flowchart describing the analytic cohort selection (data availability for MetS and symptom measures).
• We revised the Methods section as follows: “Participants enrolled in the Taiwan Biobank between 2008 and 2019 (n = 122,068) were screened for eligibility. Individuals with missing data on metabolic syndrome components (n = 409) or psychiatric morbidity assessment (n = 84) were excluded. A total of 121,575 participants were included in the final analysis (Figure 1).” (Page 3, Methods)

 

Results
Comments: Table 1: Please express continuous variables, including age and BMI, to one decimal place. Does the age range start from adulthood (20 years old)? Is there no information on monthly income? Is information available on employment status and religious affiliation? Is there any information on physical activity? Metabolic syndrome, as well as psychiatric morbidity, is associated with various physical, social, and economic issues. Please provide as much diverse information as possible that has been collected. While having a large amount of data is an advantage, it is better to have more specific information about the participants, even if the amount of data analyzed is reduced depending on the direction and purpose of the study.

• Response: We reformatted key continuous variables (e.g., age and BMI) to one decimal place and clarified that the analytic age range was 30–70 years based on the cohort description. Monthly income, employment status, and religion were not available in the dataset and therefore could not be included; this limitation is now acknowledged in the manuscript. Physical activity was represented by regular exercise, which was already included and retained in the analysis.
• We further revised the Table 1.
• We revised the Methods section as follows: “The dataset for this study was derived from a nationwide, population-based cohort of volunteers aged 30–70 years who were enrolled between 2008 and 2019 through 29 community recruitment centers across Taiwan, as detailed in previous publications.” (Page 2, Methods)
• We further added the Limitation as follows: “Fourth, residual confounding is possible because important factors such as medication use (psychotropic and metabolic agents), dietary patterns, psychosocial stress, monthly income, employment status, religion, and sleep were not available for adjustment.” (Page 10, limitations)

Comments: Table 2: What is BMI compared to?

• Response: We clarified the unit of BMI in the table label as per 1 kg/m².

Comments: Tables 4 and 5 could be better presented using bar graphs.

• Response: We agree and have added graphical representations: Supplementary Figure 1, a bar plot of adjusted odds ratios by number of MetS traits, and Supplementary Figure 2, a bar plot showing the fully adjusted associations of individual MetS components.

Comments: Consider using "—" instead of "to" for the 95% CI.

• Response: Implemented across tables (e.g., “1.152–1.325” instead of “1.152 to 1.325,” which appeared in the original tables)

 

Discussion
Comments: The relationship between metabolic syndrome and psychiatric problems is very complex. Psychological distress can be caused by metabolic syndrome, and mental health problems can lead to difficulties in controlling diet. Considering these complex issues, the authors' discussion is very brief and lacks depth. Fortunately, numerous studies have already been conducted on these two issues. For a more in-depth discussion, please refer to more studies and interpret the results of this study accordingly. Also, please add possible interpretations regarding the biomechanisms based on the existing knowledge from previous research.

• Response: We clarified the interpretation to emphasize a non-causal relationship consistent with the cross-sectional design, expanded the discussion of biological and behavioral pathways (including inflammation, HPA-axis dysregulation, and lifestyle factors), and incorporated a nutrition-specific perspective by highlighting diet quality as a potential upstream factor and suggesting future studies incorporating dietary assessment, in alignment with the editor’s request.
• We further revised the Limitations section as follows: “First, the cross-sectional design precludes causal inference and does not allow determination of whether MetS precedes depressive/anxiety symptoms or vice versa.” (Page 9, limitations)
• We revised the mechanism section as follows: “Prior research suggests that these associations may operate through interconnected bi-ological and behavioral pathways. At the biological level, chronic low-grade inflam-mation, oxidative stress, autonomic imbalance, and hypothalamic–pituitary–adrenal axis dysregulation have been proposed as shared mechanisms contributing to both cardiometabolic dysfunction and mood-related symptoms [31–34]. At the behavioral level, depressive and anxiety symptoms may adversely affect diet quality, physical ac-tivity, sleep, and medication adherence, whereas cardiometabolic symptoms and weight-related stigma may increase psychological distress [35–37]. Within a cross-sectional framework, our findings are therefore more consistent with shared-pathway hypotheses than with a specific causal direction. From a nutritional perspective, diet quality represents a plausible upstream factor: dietary patterns high in ultra-processed foods and added sugars are associated with cardiometabolic risk, whereas healthier dietary patterns (e.g., higher intake of fiber, fruits/vegetables, and unsaturated fats) have been linked to a lower depressive symptom burden [38–41]. Although detailed dietary data were not available in this dataset, future Taiwanese cohort studies incorporating nutritional assessment may clarify whether diet mediates or modifies the MetS–mental health relationship.” (Page 9, Discussion)

Reviewer 2 Report

Comments and Suggestions for Authors

The authors aimed to investigate the association between metabolic syndrome (MetS) and depressive and anxiety symptoms in a nationwide Taiwanese adult cohort. Mental health status was assessed using the Generalized Anxiety Disorder‑2 (GAD‑2) and Patient Health Questionnaire‑2 (PHQ‑2) instruments. Depressive and anxiety symptoms were identified in 1,366 of 27,349 participants with MetS (5.0%) and in 4,047 of 94,226 participants without MetS (4.3%). After multivariable adjustment, MetS was significantly associated with increased odds of depressive and anxiety symptoms (adjusted odds ratio [aOR] 1.235; 95% confidence interval [CI] 1.152–1.325). Among individual MetS components, hypertension, increased waist circumference, and hypertriglyceridemia were independently associated with higher odds of depressive and anxiety symptoms. The authors concluded that MetS was associated with a modest but significant increase in mental health symptom burden among adults, and that prospective studies are warranted to clarify the temporal relationship and underlying biological mechanisms.

First, the reviewer has concerns regarding the use of GAD‑2 and PHQ‑2 to assess psychiatric morbidity. These instruments are widely used as screening tools for common depressive and anxiety symptoms rather than for diagnosing psychiatric disorders or psychiatric morbidity per se. Therefore, the reviewer recommends revising the terminology throughout the manuscript from psychiatric morbidity to depressive and anxiety symptoms or common mental health symptoms to better reflect the measurement tools used.

Second, in the fourth paragraph of the Introduction, the authors state: “Although numerous studies have identified significant associations between MetS and mood or anxiety disorders, the direction of causality remains uncertain. Furthermore, most investigations have focused on Western populations, leaving a paucity of data from Asian regions such as Taiwan. Cultural norms, dietary habits, and social environments differ across populations and may modify both the risk and clinical expression of these disorders.” This framing may lead readers to expect either a longitudinal or interventional design, or analyses explicitly examining population-specific or cultural factors influencing the expression of depressive or anxiety symptoms. However, such data are not presented in the current study. Although the participants were Taiwanese residents, the findings appear largely comparable to those reported in Western populations, and no in-depth comparison or discussion of potential population-specific differences is provided.

Furthermore, in the Discussion section (study limitations), the authors acknowledge that a cross-sectional design precludes causal inference and limits the ability to determine temporal relationships. When the Introduction and Discussion are considered together, the reviewer found that the rationale for choosing a cross-sectional design was not sufficiently articulated.

On the other hand, this study has several notable strengths. It is based on a large nationwide population sample of Taiwanese adults, providing substantial statistical power to detect modest associations and enhancing the generalizability of the findings. The use of standardized and widely validated screening instruments (PHQ‑2 and GAD‑2) allows for efficient and consistent assessment of depressive and anxiety symptoms in a large epidemiological setting. Additionally, the analysis of individual MetS components provides more granular insight into which specific metabolic abnormalities are associated with mental health symptoms. Finally, the inclusion of a non-Western population contributes valuable evidence supporting the reproducibility of previously reported associations between metabolic syndrome and depressive and anxiety symptoms in an Asian context. From a public health perspective, the study highlights the potential relevance of obesity and metabolic health interventions for both physical and mental health promotion in the community.

Taken together, there appears to be a mismatch between the issues emphasized in the Introduction, particularly causality and population-specific mechanisms, and what a cross-sectional design can reasonably address. In addition, given that PHQ-2 and GAD-2 are screening instruments for depressive and anxiety symptoms rather than diagnostic tools, terminology such as “psychiatric morbidity” may overstate what can be inferred from the data. Therefore, the Introduction should be revised to clearly articulate the study’s descriptive aims, use symptom-level language consistent with the measures employed, and ensure that both the framing and terminology are methodologically aligned with the study design throughout the manuscript (not only in the Introduction).

The reviewer believes these revisions would help improve clarity and ensure that the conclusions are interpreted appropriately, without diminishing the value of the current findings.

Author Response

Comments:  First, the reviewer has concerns regarding the use of GAD‑2 and PHQ‑2 to assess psychiatric morbidity. These instruments are widely used as screening tools for common depressive and anxiety symptoms rather than for diagnosing psychiatric disorders or psychiatric morbidity per se. Therefore, the reviewer recommends revising the terminology throughout the manuscript from psychiatric morbidity to depressive and anxiety symptoms or common mental health symptoms to better reflect the measurement tools used.

• Response: We appreciate the reviewer’s important observation. We agree that PHQ-2 and GAD-2 are screening instruments rather than diagnostic tools. Our intention, however, was not to indicate clinically confirmed psychiatric disorders but to describe population-level mental health burden. In epidemiological research, the term psychiatric morbidity is commonly used to represent clinically relevant psychological symptom burden detected using standardized questionnaires, rather than formal DSM/ICD diagnoses. To avoid overinterpretation, we have explicitly clarified this definition throughout the manuscript. Psychiatric morbidity in the present study refers to screen-positive depressive and/or anxiety symptoms (PHQ-2 ≥ 3 or GAD-2 ≥ 3) or self-reported physician-diagnosed depression, and should therefore be interpreted as a population-level indicator of mental health burden rather than confirmed psychiatric disease.
• We further revised the Methods section as follows: “In this research, psychiatric morbidity was defined as depressive and/or anxiety burden identified by validated screening instruments (PHQ-2 ≥ 3 or GAD-2 ≥ 3) or self-reported physician-diagnosed depression.” (Page 4, Methods)
• We further revised the Limitation section as follows: “Third, our outcome definition combined screening-positive symptoms with self-reported physician-diagnosed depression; this approach may increase sensitivity but also introduces heterogeneity, and the self-reported diagnosis could not be verified in this de-identified dataset. Accordingly, psychiatric morbidity should be interpreted as a population-level mental health burden, reflecting screening-detected symptoms and self-reported prior diagnosis rather than clinically confirmed disorders.”(Page 9, Limitations)

 

Comments: Second, in the fourth paragraph of the Introduction, the authors state: “Although numerous studies have identified significant associations between MetS and mood or anxiety disorders, the direction of causality remains uncertain. Furthermore, most investigations have focused on Western populations, leaving a paucity of data from Asian regions such as Taiwan. Cultural norms, dietary habits, and social environments differ across populations and may modify both the risk and clinical expression of these disorders.” This framing may lead readers to expect either a longitudinal or interventional design, or analyses explicitly examining population-specific or cultural factors influencing the expression of depressive or anxiety symptoms. However, such data are not presented in the current study. Although the participants were Taiwanese residents, the findings appear largely comparable to those reported in Western populations, and no in-depth comparison or discussion of potential population-specific differences is provided.

• Response: We thank the reviewer for this important observation. We agree that the previous wording in the Introduction could imply that the present study was designed to evaluate causal relationships or population-specific mechanisms. Our intention, however, was to emphasize the value of large-scale population-based evidence from an Asian cohort rather than to test cultural determinants directly. Accordingly, we revised the Introduction to clarify that the aim of this study is descriptive and epidemiological: to quantify the prevalence and magnitude of the association between metabolic syndrome and psychiatric morbidity in a large community-based Taiwanese population. We removed wording suggesting mechanistic or causal inference and now explicitly state that cross-sectional community screening data provide estimates of real-world symptom burden but cannot determine causality or cultural mechanisms. We also clarified in the Discussion that our findings are generally consistent with prior studies and should be interpreted as supporting the reproducibility of previously reported associations in a non-Western population rather than identifying population-specific effects.
• We further revised the introduction section as follows: “Although consistent associations have been observed, temporal direction and population-level burden remain unclear because many studies were conducted in Western populations or lacked large community-based samples. Cultural norms, dietary habits, and social environments differ across populations and may modify both risk and clinical expression of these conditions. Moreover, large-scale community screening data can estimate real-world symptom burden but cannot establish causality; therefore, clarifying the magnitude and pattern of association in such populations remains clinically informative. Therefore, we aimed to quantify the prevalence of psychiatric morbidity according to MetS status and evaluate the associations of individual MetS components and cumulative metabolic burden with psychiatric morbidity in a population-based cohort.” (Page 2, introduction)

• We further revised the Discussion section as follows: “Within a cross-sectional framework, our findings are therefore more consistent with shared-pathway hypotheses than with a specific causal direction.” (Page 9, line 31 to 33)
• We revised the Limitations section as follows: “First, the cross-sectional design precludes causal inference and does not allow determination of whether MetS precedes depressive/anxiety symptoms or vice versa.”(Page 9, Limitations)

 

Comments: Furthermore, in the Discussion section (study limitations), the authors acknowledge that a cross-sectional design precludes causal inference and limits the ability to determine temporal relationships. When the Introduction and Discussion are considered together, the reviewer found that the rationale for choosing a cross-sectional design was not sufficiently articulated.

• Response: We thank the reviewer for this helpful suggestion. We agree that the purpose of using a cross-sectional design should be clarified. The present study used a nationwide community screening cohort, which provides a single-time-point assessment of metabolic and psychological status. Our primary objective was therefore not to evaluate temporal sequence or causality, but to estimate population-level symptom burden and quantify the strength and pattern of associations between metabolic syndrome and psychiatric morbidity in a real-world community population. Such large-scale screening data are particularly suitable for identifying epidemiological relationships and generating hypotheses for future longitudinal or interventional studies. To clarify this rationale, we revised the Introduction and Discussion to explicitly state that the cross-sectional design was chosen to characterize prevalence and association patterns rather than causal pathways.
• We revised the introduction section as follows: “Moreover, large-scale community screening data can estimate real-world symptom burden but cannot establish causality; therefore, clarifying the magnitude and pattern of association in such populations remains clinically informative.” (Page 2, line 10 from bottom)
• We revised the Discussion section as follows: “Within a cross-sectional framework, our findings are therefore more consistent with shared-pathway hypotheses than with a specific causal direction.” (Page 9, line 31 to 33)

 

Comments: On the other hand, this study has several notable strengths. It is based on a large nationwide population sample of Taiwanese adults, providing substantial statistical power to detect modest associations and enhancing the generalizability of the findings. The use of standardized and widely validated screening instruments (PHQ‑2 and GAD‑2) allows for efficient and consistent assessment of depressive and anxiety symptoms in a large epidemiological setting. Additionally, the analysis of individual MetS components provides more granular insight into which specific metabolic abnormalities are associated with mental health symptoms.

• Response: We thank the reviewer for the positive evaluation of our study and for recognizing the strengths of the large nationwide cohort, the use of validated screening instruments, and the component-level analysis of metabolic syndrome.

Comments: Finally, the inclusion of a non-Western population contributes valuable evidence supporting the reproducibility of previously reported associations between metabolic syndrome and depressive and anxiety symptoms in an Asian context. From a public health perspective, the study highlights the potential relevance of obesity and metabolic health interventions for both physical and mental health promotion in the community.

• Response: We thank the reviewer for recognizing the contribution of this study in providing evidence from a large Asian population and its potential public health relevance. In the revised manuscript, we have further clarified in the Discussion that our findings support the reproducibility of previously reported associations in a non-Western population and highlight the potential value of integrated metabolic and mental health prevention strategies at the community level.

Reviewer 3 Report

Comments and Suggestions for Authors

The background clearly outlines the rationale for examining the association between metabolic syndrome and psychiatric morbidity and appropriately highlights the gap in large-scale population-based evidence. The objective is well defined and aligned with the study design. The results are clearly presented and support the study’s conclusions. 

Minor suggestions

1. The definition of metabolic syndrome would benefit from briefly listing the key diagnostic criteria and corresponding cutoff values used in the analysis.
2. The inclusion of a flowchart detailing the eligible participant selection process is appropriate and improves the clarity and transparency of the study design.
3. The manuscript primarily presents results in tabular form. Inclusion of graphical representations—such as bar charts or forest plots illustrating univariable associations with psychiatric morbidity—could enhance clarity and facilitate interpretation of the findings.

Author Response

Comments: The definition of metabolic syndrome would benefit from briefly listing the key diagnostic criteria and corresponding cutoff values used in the analysis.

• Response: We clarified that MetS was defined using standard criteria and ensured the component cutoffs are explicitly listed in Methods.
• We revised the Methods as follows: “This study applied the diagnostic criteria for MetS commonly adopted in Taiwan, which are based on the recommendations of the International Diabetes Federation (IDF) and the National Cholesterol Education Program Adult Treatment Panel III (NCEP-ATP III), with adjustments to better reflect Asian population characteristics. Participants were identified as having MetS if they met at least three of the following five criteria:

(1)    Abdominal obesity: Waist circumference ≥90 cm in men or ≥80 cm in women.

(2)    Triglycerides: Fasting triglyceride levels ≥150 mg/dL (1.7 mmol/L) or current use of lipid-lowering medication.

(3)    High-density lipoprotein cholesterol (HDL-C): HDL-C <40 mg/dL (1.03 mmol/L) for men or <50 mg/dL (1.29 mmol/L) for women, or undergoing treatment aimed at raising HDL-C.

(4)    Blood pressure: Systolic blood pressure ≥130 mmHg or diastolic blood pressure ≥85 mmHg, or use of antihypertensive medication.

(5)    Fasting blood glucose: Fasting plasma glucose ≥100 mg/dL (5.6 mmol/L) or a documented diagnosis of diabetes under current treatment.”

 

Comments: The inclusion of a flowchart detailing the eligible participant selection process is appropriate and improves the clarity and transparency of the study design.

• Response: Added Figure 1 flowchart.

 

Comments: The manuscript primarily presents results in tabular form. Inclusion of graphical representations—such as bar charts or forest plots illustrating univariable associations with psychiatric morbidity—could enhance clarity and facilitate interpretation of the findings.

• Response: We agree and have added graphical representations: Supplementary Figure 1, a bar plot of adjusted odds ratios by number of MetS traits, and Supplementary Figure 2, a bar plot showing the fully adjusted associations of individual MetS components.

Reviewer 4 Report

Comments and Suggestions for Authors

The study explores the intersection between metabolic and mental health, a topic of growing clinical relevance. Its main contribution lies in validating this association within a large Asian cohort, demonstrating a progressive increase in risk as additional components of metabolic syndrome (MetS) accumulate.

However, the following points for improvement are suggested:

1. Although the PHQ-2 and GAD-2 are validated screening tools, their use to define “psychiatric morbidity” may lead to an overestimation of true clinical prevalence. A more in-depth discussion is needed regarding the sensitivity and specificity of these ultra-short instruments compared with formal clinical diagnoses.

2. The study design is cross-sectional, which precludes establishing whether MetS precedes depression/anxiety or vice versa. This limitation should be more explicitly emphasized in the Abstract.

3. The Discussion states that waist circumference did not retain statistical significance in the multivariable model; however, Table 5 and the Conclusion report the opposite (aOR 1.087, p < 0.001). This inconsistency must be corrected. Section 4 (Discussion) should be revised to align strictly with the data presented in Table 5 regarding which MetS components are truly independent predictors.

4. It should be clarified whether the “self-reported medical diagnosis of depression” was verified through medical records or relied solely on participant recall.

5. A sensitivity analysis excluding participants who met the criteria solely based on self-report should be added to assess whether the association with PHQ/GAD scales remains equally robust.

Author Response

Comments: Although the PHQ-2 and GAD-2 are validated screening tools, their use to define “psychiatric morbidity” may lead to an overestimation of true clinical prevalence. A more in-depth discussion is needed regarding the sensitivity and specificity of these ultra-short instruments compared with formal clinical diagnoses.

• Response: We agree that PHQ-2 and GAD-2 are screening instruments and may not reflect clinically confirmed psychiatric disorders. We expanded the Discussion to clarify that these tools identify symptom burden rather than diagnostic prevalence and added description regarding their screening performance compared with structured clinical interviews.
• We revised the discussion section as follows: “This approach increases sensitivity for capturing population-level mental health bur-den but introduces heterogeneity, as screening tools identify current symptoms whereas self-reported diagnosis may reflect prior clinical recognition. The self-reported diagnosis could not be verified in this de-identified dataset. Therefore, the estimates should be interpreted as reflecting overall psychological symptom burden rather than a single uniform clinical entity, rather than clinically confirmed psychiatric disorders.”(Page 10, line 2 to 7)

 

Comments: The study design is cross-sectional, which precludes establishing whether MetS precedes depression/anxiety or vice versa. This limitation should be more explicitly emphasized in the Abstract.

• Response: We agree and have now explicitly stated this limitation in the Abstract.
• We revised the Abstract section as follows: “Given the cross-sectional design and symptom-based screening measures, the findings should be interpreted as associations rather than evidence of causality.”(Page 2, abstract)

 

Comments: The Discussion states that waist circumference did not retain statistical significance in the multivariable model; however, Table 5 and the Conclusion report the opposite (aOR 1.087, p < 0.001). This inconsistency must be corrected. Section 4 (Discussion) should be revised to align strictly with the data presented in Table 5 regarding which MetS components are truly independent predictors.

• Response: We thank the reviewer for identifying this error. Waist circumference remained statistically significant in the fully adjusted model (aOR 1.087, p < 0.001). The Discussion text has been corrected to align with Table 5 and the Conclusion.
• We revised the Discussion as follows: “Among individual MetS components, hypertension, increased waist circumference and hypertriglyceridemia remained significantly associated with psychiatric morbidity after full adjustment, corroborating findings from earlier research [26,27]. In contrast, impaired glucose tolerance, and low HDL-C did not retain statistical significance in the multivariable model, suggesting that their associations may be mediated indirectly through other metabolic or behavioral factors.”(Page 9, line 12 to 17)

 

Comments: It should be clarified whether the “self-reported medical diagnosis of depression” was verified through medical records or relied solely on participant recall.

• Response: The Taiwan Biobank dataset contains self-reported physician-diagnosed depression and does not include medical record verification. We clarified this explicitly in the Methods and Limitations sections.
• We revised the Limitations section as follows: “The self-reported diagnosis could not be verified in this de-identified dataset.”(Page 10, limitations)

 

Comments: A sensitivity analysis excluding participants who met the criteria solely based on self-report should be added to assess whether the association with PHQ/GAD scales remains equally robust.

• Response: We agree and performed an additional sensitivity analysis excluding participants classified solely by self-reported physician-diagnosed depression. The association between metabolic syndrome and psychiatric morbidity remained statistically significant, indicating that the main findings were not driven by recall-based diagnosis. Specifically, metabolic syndrome remained significantly associated with psychiatric morbidity defined by screening instruments (adjusted OR 1.208, 95% CI 1.024–1.424).
• We further revised the Results section as follows: “Moreover, a sensitivity analysis excluding participants classified solely by self-reported physician-diagnosed depression yielded similar results, with MetS remaining signifi-cantly associated with psychiatric morbidity defined by screening instruments (aOR 1.208, 95% CI 1.024–1.424, p = 0.025)(Supplementary Table 1).”(Page 6, results)

Reviewer 5 Report

Comments and Suggestions for Authors

This paper explores the correlation between metabolic syndrome (MetS) and psychiatric morbidity (PM) in a large nationwide Taiwanese cohort. The study investigates a highly relevant issue at the crossroads of metabolic and mental health, which is a rapidly expanding area of interest in public health and clinical research. A key feature of this paper is the extraordinarily large sample along with using standardized health screening data collected at various community locations. The authors explain the analysis well and demonstrate a strong effort in adjusting for a wide range of confounding variables. Thus, this paper effectively provides solid epidemiological evidence from an Asian population that has been largely overlooked and contributes significantly to the existing literature.

Below, I present my critical assessment and detailed remarks on the paper.

• The cross-sectional design limits the ability to make inferences about the direction and causality of the relationships. Although the limitation is recognized in the manuscript, some sections of the discussion allude to or imply bidirectional/causal pathways that are not directly supported by the data, and the authors should use more cautious language when presenting these hypotheses.
• Mental health status is determined using self-reports of physician diagnosed depression in combination with screening tools (PHQ-2 and GAD-2). On one hand, such an approach increases sensitivity; on the other hand, it also creates a different case definition in terms of heterogeneity. A short discussion on the consequences of mixing screening-based and self-reported diagnostic indicators would help unveil the methodology from the inside.
• Very brief screening tools like PHQ-2 and GAD-2 were employed to keep large-scale studies feasible, but these are not diagnostic instruments. The manuscript lacks a proper definition between psychiatric morbidity as a screening-based concept and clinically diagnosed disorders and could gain from this clarification.
• The authors' regression analysis accounts for several covariates, but nevertheless, they allow for the possibility of residual confounding. There are some influential factors, e.g., medication usage (psychotropic and metabolic drugs), dietary patterns, psychosocial stress, which have not been controlled for and thus may impact both metabolic and mental health outcomes.
• While the authors have statistically significant findings, the effect sizes reported are small. Placing more weight on effect sizes and how the clinical or public health relevance of the observed associations could be accounted for when interpreting the results would be helpful.
• Due to the large size of the cohort studied, the paper could use some sensitivity or stratified analyses (e.g., by sex or age groups) to identify any effect modifications. This aspect should be especially worth exploring considering the pronounced sex differences in psychiatric morbidity demonstrated in the results.
• Lastly, the paper closely matches the thematic of Nutrients, yet the nutritional angle is almost entirely missing. A few lines describing the connection between metabolic syndrome, dietary habits, and mental health would go a long way towards accounting for the journal's focus in this submission.

In conclusion, I find this paper to be a well-conducted and methodologically rigorous epidemiological research study. My points above serve more as a suite of suggestions for improved clarity and refinement rather than as major flaws, so the paper's contribution would be still considered strong even without them.

Author Response

Comments: The cross-sectional design limits the ability to make inferences about the direction and causality of the relationships. Although the limitation is recognized in the manuscript, some sections of the discussion allude to or imply bidirectional/causal pathways that are not directly supported by the data, and the authors should use more cautious language when presenting these hypotheses.

• Response: We thank the reviewer for this important suggestion. We have carefully revised the Discussion to avoid causal or bidirectional interpretations and to ensure that all mechanistic explanations are presented as hypothesis-generating rather than inferential. Specifically, wording implying directionality (e.g., “leads to”, “results in”, or “increases risk”) has been replaced with neutral terminology such as “associated with” or “may be related to.” We also clarified that biological and behavioral mechanisms represent potential shared pathways rather than demonstrated causal processes.
• We further revised the Discussion section as follows: “Within a cross-sectional framework, our findings are therefore more consistent with shared-pathway hypotheses than with a specific causal direction.”
• We revised the Limitations section as follows: “First, the cross-sectional design precludes causal inference and does not allow determination of whether MetS precedes depressive/anxiety symptoms or vice versa.”
• We revised the mechanism section as follows: “Prior research suggests that these associations may operate through interconnected bi-ological and behavioral pathways. At the biological level, chronic low-grade inflam-mation, oxidative stress, autonomic imbalance, and hypothalamic–pituitary–adrenal axis dysregulation have been proposed as shared mechanisms contributing to both cardiometabolic dysfunction and mood-related symptoms [31–34]. At the behavioral level, depressive and anxiety symptoms may adversely affect diet quality, physical ac-tivity, sleep, and medication adherence, whereas cardiometabolic symptoms and weight-related stigma may increase psychological distress [35–37]. Within a cross-sectional framework, our findings are therefore more consistent with shared-pathway hypotheses than with a specific causal direction. From a nutritional perspective, diet quality represents a plausible upstream factor: dietary patterns high in ultra-processed foods and added sugars are associated with cardiometabolic risk, whereas healthier dietary patterns (e.g., higher intake of fiber, fruits/vegetables, and unsaturated fats) have been linked to a lower depressive symptom burden [38–41]. Although detailed dietary data were not available in this dataset, future Taiwanese cohort studies incorporating nutritional assessment may clarify whether diet mediates or modifies the MetS–mental health relationship.”

Comments: Mental health status is determined using self-reports of physician diagnosed depression in combination with screening tools (PHQ-2 and GAD-2). On one hand, such an approach increases sensitivity; on the other hand, it also creates a different case definition in terms of heterogeneity. A short discussion on the consequences of mixing screening-based and self-reported diagnostic indicators would help unveil the methodology from the inside.

• Response: We thank the reviewer for this insightful comment. We agree that combining screening-based symptom measures (PHQ-2/GAD-2) with self-reported physician-diagnosed depression introduces heterogeneity in outcome definition. The combined definition was chosen to capture population-level mental health burden in an epidemiological setting, increasing sensitivity but potentially incorporating individuals at different stages of symptom recognition or healthcare utilization. To evaluate whether recall-based diagnosis influenced the observed association, we performed an additional sensitivity analysis excluding participants classified solely by self-reported physician-diagnosed depression. The association between metabolic syndrome and psychiatric morbidity remained statistically significant, indicating that the main findings were not driven by self-reported diagnosis. Specifically, metabolic syndrome remained significantly associated with psychiatric morbidity defined by screening instruments (adjusted OR 1.208, 95% CI 1.024–1.424).
• We further revised the Results section as follows: “Moreover, a sensitivity analysis excluding participants classified solely by self-reported physician-diagnosed depression yielded similar results, with MetS remaining significantly associated with psychiatric morbidity defined by screening instruments (aOR 1.208, 95% CI 1.024–1.424, p = 0.025)(Supplementary Table 1).”
• We further revised the Limitations section as follows: “Third, our outcome definition combined screening-positive symptoms with self-reported physician-diagnosed depression. This approach increases sensitivity for capturing population-level mental health burden but introduces heterogeneity, as screening tools identify current symptoms whereas self-reported diagnosis may reflect prior clinical recognition. The self-reported diagnosis could not be verified in this de-identified dataset. Therefore, the estimates should be interpreted as reflecting over-all psychological symptom burden rather than a single uniform clinical entity, rather than clinically confirmed psychiatric disorders.”

Comments: Very brief screening tools like PHQ-2 and GAD-2 were employed to keep large-scale studies feasible, but these are not diagnostic instruments. The manuscript lacks a proper definition between psychiatric morbidity as a screening-based concept and clinically diagnosed disorders and could gain from this clarification.

• Response: We thank the reviewer for this helpful comment. We agree that PHQ-2 and GAD-2 are screening instruments rather than diagnostic tools, and we have clarified the distinction between screening-based psychiatric morbidity and clinically diagnosed psychiatric disorders throughout the manuscript. In this study, psychiatric morbidity is intended to represent population-level psychological symptom burden detected by standardized questionnaires rather than DSM/ICD-confirmed disease. Specifically, psychiatric morbidity was defined as screen-positive depressive and/or anxiety symptoms (PHQ-2 ≥ 3 or GAD-2 ≥ 3) or self-reported physician-diagnosed depression. We now explicitly state that this construct reflects symptom burden in epidemiologic research and should not be interpreted as the prevalence of clinically confirmed psychiatric disorders.
• We further revised the Methods section as follows: “In this research, psychiatric morbidity was defined as depressive and/or anxiety burden identified by validated screening instruments (PHQ-2 ≥ 3 or GAD-2 ≥ 3) or self-reported physician-diagnosed depression.”
• We revised the Limitations section as follows: “Our outcome definition combined screening-positive symptoms with self-reported physician-diagnosed depression… Therefore, psychiatric morbidity should be interpreted as a population-level mental health burden rather than clinically confirmed psychiatric disorders.”

 

Comments: The authors' regression analysis accounts for several covariates, but nevertheless, they allow for the possibility of residual confounding. There are some influential factors, e.g., medication usage (psychotropic and metabolic drugs), dietary patterns, psychosocial stress, which have not been controlled for and thus may impact both metabolic and mental health outcomes.

• Response: Added to limitations (psychotropic/metabolic medications, dietary patterns, psychosocial stress, sleep).
• We revised the Limitations section as follows: “Fourth, residual confounding is possible because important factors such as medication use (psychotropic and metabolic agents), dietary patterns, psychosocial stress, monthly income, employment status, religion, and sleep were not available for adjustment.”

 

Comments: While the authors have statistically significant findings, the effect sizes reported are small. Placing more weight on effect sizes and how the clinical or public health relevance of the observed associations could be accounted for when interpreting the results would be helpful.

• Response: We thank the reviewer for this important suggestion. We agree that the observed effect sizes are modest at the individual level. We have therefore revised the Discussion to emphasize interpretation from a population-health perspective. Because metabolic syndrome is highly prevalent in the general population, even small relative increases in psychological symptom burden may translate into a meaningful number of affected individuals at the community level. Accordingly, our findings should be interpreted as indicating potential public health relevance rather than strong individual-level prediction. We have added clarifying statements to avoid overstating clinical impact.
• We revised the Discussion as follows: “Although the observed effect sizes were modest, the high population prevalence of metabolic syndrome implies that even small increases in relative risk may translate into a considerable mental health burden at the community level. Therefore, the findings are more relevant to population-level awareness and prevention strategies than to individual-level clinical prediction.”

 

Comments: Due to the large size of the cohort studied, the paper could use some sensitivity or stratified analyses (e.g., by sex or age groups) to identify any effect modifications. This aspect should be especially worth exploring considering the pronounced sex differences in psychiatric morbidity demonstrated in the results.

• Response: We thank the reviewer for this helpful suggestion. We performed sex-stratified multivariable logistic regression analyses to evaluate potential effect modification. Metabolic syndrome remained significantly associated with psychiatric morbidity in both men and women (men: aOR 1.340, 95% CI 1.183–1.518; women: aOR 1.199, 95% CI 1.101–1.305).
• We have added these results as Supplementary Table S2 and incorporated a corresponding description in the Results and Discussion sections.

 

Comments: Lastly, the paper closely matches the thematic of Nutrients, yet the nutritional angle is almost entirely missing. A few lines describing the connection between metabolic syndrome, dietary habits, and mental health would go a long way towards accounting for the journal's focus in this submission.

• Response: We agree that the nutritional relevance should be more clearly articulated given the scope of Nutrients. The Discussion has been expanded to incorporate a nutrition-focused perspective. Specifically, diet quality is now described as a potential upstream factor linking metabolic and mental health outcomes. We added text noting that dietary patterns high in ultra-processed foods and added sugars are associated with cardiometabolic risk, whereas healthier dietary patterns (e.g., higher intake of fiber, fruits/vegetables, and unsaturated fats) are associated with lower depressive symptom burden. These pathways are explicitly presented as hypothesis-generating within the constraints of a cross-sectional design.
• We further revised the Discussion section as follows: “From a nutritional perspective, diet quality represents a plausible upstream factor: dietary patterns high in ultra-processed foods and added sugars are associated with cardiometabolic risk, whereas healthier dietary patterns (e.g., higher intake of fiber, fruits/vegetables, and unsaturated fats) have been linked to a lower depressive symptom burden [38–41]. Although detailed dietary data were not available in this dataset, future Taiwanese cohort studies incorporating nutritional assessment may clarify whether diet mediates or modifies the MetS–mental health relationship.”

 

In conclusion, I find this paper to be a well-conducted and methodologically rigorous epidemiological research study. My points above serve more as a suite of suggestions for improved clarity and refinement rather than as major flaws, so the paper's contribution would be still considered strong even without them.

• Response: We appreciate the reviewer’s overall positive assessment of the study. We have carefully considered the suggested refinements and revised the manuscript accordingly to improve clarity, precision of terminology, and interpretation of findings. We believe these revisions have strengthened the presentation while preserving the original conclusions.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

I do not have comment.

Author Response

Comments: I do not have comment.

Response: Thank you for reviewing our revision and for confirming that there are no further comments.
We sincerely appreciate your helpful suggestions, which have improved the manuscript.

Reviewer 2 Report

Comments and Suggestions for Authors

First, the reviewer would like to thank the authors for revising their manuscript in response to his/her comments.

The authors' revisions appear to have significantly improved the logic of the paper. However, some parts could benefit from improved coherence. For example, the sentence in the introduction, "Moreover, large-scale community screening data can estimate real-world symptom burden but cannot establish causality; therefore, clarifying the magnitude and pattern of association in such populations remains clinically informative," seems somewhat disconnected from the surrounding context, and the rationale for conducting a cross-sectional study of local residents in Taiwan could be articulated more clearly. One possible way to improve clarity might be to revise the final two paragraphs of the introduction as follows: "Although community screening data do not establish causality, they do demonstrate the magnitude and pattern of associations, which are clinically useful. Previous studies have focused on Western countries, and data from Taiwan are scarce. Therefore, this study, a baseline cross-sectional analysis of a population-based cohort in Taiwan, aimed to (1) quantify the prevalence of psychiatric morbidity by MetS status and (2) evaluate the association between each component of MetS, cumulative metabolic burden, and psychiatric morbidity."

 

 

 

Furthermore, the abstract conclusions, "These findings should be interpreted as associations rather than evidence of causality," may inadvertently place more emphasis on the limitations rather than the key findings. For example, how about revising it to rewrite, "Consistent with previous findings, MetS was associated with a higher prevalence and greater symptom burden of psychiatric morbidity in this large Taiwanese community cohort. These cross-sectional associations underscore the need for future longitudinal studies in Taiwan to clarify the temporal sequence and underlying mechanisms linking metabolic and mental health."

 

He/she believes that the value of this paper will be enhanced by clearly discussing the purpose, conclusions, and significance of the research throughout the paper. 

I would appreciate the authors’ thoughts on these suggestions.

 

Author Response

Comment 1: The authors' revisions appear to have significantly improved the logic of the paper. However, some parts could benefit from improved coherence. For example, the sentence in the introduction, "Moreover, large-scale community screening data can estimate real-world symptom burden but cannot establish causality; therefore, clarifying the magnitude and pattern of association in such populations remains clinically informative," seems somewhat disconnected from the surrounding context, and the rationale for conducting a cross-sectional study of local residents in Taiwan could be articulated more clearly. One possible way to improve clarity might be to revise the final two paragraphs of the introduction as follows: "Although community screening data do not establish causality, they do demonstrate the magnitude and pattern of associations, which are clinically useful. Previous studies have focused on Western countries, and data from Taiwan are scarce. Therefore, this study, a baseline cross-sectional analysis of a population-based cohort in Taiwan, aimed to (1) quantify the prevalence of psychiatric morbidity by MetS status and (2) evaluate the association between each component of MetS, cumulative metabolic burden, and psychiatric morbidity."

• Response: Thank you for your suggestions, we further revised the Introduction as follows: “Although consistent associations have been observed, the population-level burden and pattern of association remain uncertain. Community-based screening data, while not suitable for establishing causality, can provide clinically meaningful estimates of the magnitude of association in real-world populations. Furthermore, most existing studies have been conducted in Western populations, and large population-based evi-dence from Asian communities, particularly Taiwan, remains limited. Differences in cultural context, diet, and lifestyle may influence both metabolic risk and psychiatric expression, highlighting the need for region-specific data. Therefore, this study used a population-based cohort in Taiwan to (1) quantify the prevalence of psychiatric morbidity according to MetS status and (2) evaluate the as-sociations between individual MetS components, cumulative metabolic burden, and psychiatric morbidity.” (Page 2, line 14 from bottom to line 3 from bottom)

 

Comment 2: Furthermore, the abstract conclusions, "These findings should be interpreted as associations rather than evidence of causality," may inadvertently place more emphasis on the limitations rather than the key findings. For example, how about revising it to rewrite, "Consistent with previous findings, MetS was associated with a higher prevalence and greater symptom burden of psychiatric morbidity in this large Taiwanese community cohort. These cross-sectional associations underscore the need for future longitudinal studies in Taiwan to clarify the temporal sequence and underlying mechanisms linking metabolic and mental health."

• Response: Thank you for your suggestion, we revised the Abstract as follows: “MetS was associated with a modest increase in psychiatric morbidity in this large Taiwanese community cohort. Because of the cross-sectional design, causal inference is limited. Future longitudinal studies are needed to clarify the direction of association and underlying mechanisms linking metabolic and mental health conditions.”(Page 2, line 3 to 7)