Pith. sign in

REVIEW 5 major objections 6 minor 45 references

Trustworthy Chronic Disease Risk Prediction For Self-Directed Preventive Care via Medical Literature Validation

T0 review · 5 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

Pith's one-line read For all 13 chronic diseases, the model's most influential features match established medical risk factors.

desk verdict Post-hoc literature matching makes the trustworthiness claim unfalsifiable; the data work is decent but the central conclusion needs a real validation protocol. read the letter →

arxiv 2506.17620 v1 pith:UOK4L7HT submitted 2025-06-21 cs.LG cs.CY

classification cs.LGcs.CY
keywords chronicdiseaseriskpredictionSHAPexplainabilitymedicalliteraturevalidationBehavioralFactorSurveillanceSystemdeeplearningself-directedpreventivecaremodeltrustworthinessnon-clinicalfactors
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to establish that a deep learning model can be trusted to predict chronic-disease risk from ordinary personal and lifestyle questions, because the model's most influential features agree with established medical knowledge. The authors train a ResNet on the 2023 Behavioral Risk Factor Surveillance System survey to score risk for 13 chronic diseases using 38 non-clinical inputs, then use SHAP to rank feature influence for each disease. Across all 13 diseases, the three most influential features match documented risk factors, most often age, weight, sex, smoking, alcohol use, and physical activity. If that alignment is accepted, the models provide a low-cost, transparent screening signal for self-directed prevention and a reusable recipe for auditing health-risk tools.

What carries the argument

The load-bearing mechanism is an explanatory audit: SHAP (Shapley Additive exPlanations) with the Kernel Explainer is run on each trained ResNet, using 500 randomly sampled survey responses and a 100-point k-means background set, and the per-feature SHAP values are summed to produce a global ranking. For each disease, the top three ranked features are then checked against peer-reviewed medical studies. Two categorical inputs, employment and marital status, are recast as age proxies on the basis of their answer categories (student, retired, unable to work; divorced, widowed), which is what lets the age-risk literature cover most diseases.

What would settle it

Retrain the model with an explicit age feature among the inputs: if age becomes a top-three feature and employment and marital status fall out of the top three, the paper's age-proxy interpretation is supported, whereas if employment and marital status stay on top even with age controlled, the model is not simply encoding age and the age-based medical literature used for 11 of the 13 diseases does not explain the model's behavior.

Watch

Extended reading notes

Core claim

The central claim is that for every one of the 13 chronic diseases, the three inputs SHAP ranks highest for the deep model's risk score correspond to risk factors the medical literature identifies as significant, and this cross-disease agreement is evidence that the models can be broadly trusted for chronic disease prediction. The authors deliberately exclude general-health and disability-like fields because those are more likely consequences than causes, and they interpret employment and marital status as indicators of age based on survey answer categories. The result is framed as the first self-directed-care model whose explanations are validated against medical literature rather than merely reported.

Load-bearing premise

The argument depends on treating employment status (and, for two diseases, marital status) as a reliable proxy for age in the survey data; if those variables do not track age, the age-related medical studies cited for 11 of the 13 diseases no longer validate the models' most influential features.

Editorial extensions

If this is right

  • Anyone can get a first-pass chronic disease risk score from a short lifestyle questionnaire, because none of the 38 inputs requires a lab test or exam.
  • The same explainability-plus-literature audit can be applied to other health-risk models before public release, flagging models whose top features are medically implausible.
  • Because the pattern repeats across 13 diseases, the approach is presented as a general trustworthiness check rather than a one-off result for a single condition.
  • Classification accuracy and recall between 65% and 75%, comparable to earlier non-clinical models, mean the trustworthiness argument applies to models with ordinary predictive performance.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The authors do not report how stable the top-three SHAP rankings are across retraining runs; that variance would matter for anyone relying on the trustworthiness claim.
  • Because employment and marital status are validated mainly as age stand-ins, adding age as an explicit feature would probably absorb their influence and could yield a simpler model with the same behavior.
  • A stricter validation would check not only whether the top features appear in the literature but also whether the direction of each SHAP effect matches the known risk direction (for example, smoking raising COPD risk), which the paper does not test.
  • The age-proxy reading implies the models are in part social-surrogate age predictors, so deploying them in populations where employment and marital patterns differ from the U.S. survey sample would require re-validation.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

5 major / 6 minor

Summary. The paper trains deep ResNet models on the 2023 BRFSS dataset to predict 13 chronic diseases from 38 non-clinical personal and lifestyle features, using weighted cross-entropy loss to handle class imbalance. For each disease, SHAP is used to identify the three most influential features, and these are then paired with citations from the medical literature that associate the feature with the disease. The authors report that this alignment holds across all 13 diseases and conclude that the machine learning approach 'can be broadly trusted for chronic disease prediction.' The manuscript includes model training metrics (loss, accuracy, recall), SHAP-based feature rankings, disease-specific literature citations, and a short limitations section.

Significance. If the central claim were rigorously established, the paper would make a useful contribution by proposing that model explanations be validated against external medical knowledge as a trustworthiness check for self-directed risk tools. The use of a large, public dataset and the construction of 13 separate disease models are practical strengths, and the framing around self-directed preventive care addresses a real gap. However, the validation is presented as a post-hoc narrative rather than a falsifiable test: features are selected first, citations are chosen afterward, and no quantitative alignment metric, pre-registered criteria, or negative controls are reported. Because of this, the paper's main conclusion is currently unsupported, and the contribution is more of a proposal for a validation approach than a demonstration that the models are trustworthy. The manuscript also contains an unaddressed temporal mismatch between the cross-sectional BRFSS outcomes and the stated goal of predicting risk of 'developing' disease, which further limits the significance of the results as they stand.

major comments (5)
  1. [Abstract, §I, §III.A] The paper consistently describes the task as predicting the risk of 'developing' a chronic disease, but the BRFSS outcome variables (e.g., BPHIGH6, DIABETE4) are self-reported lifetime diagnoses, i.e., prevalent disease status, not incident disease. The models are therefore trained and evaluated on current or past disease status, and the results cannot be interpreted as prospective risk prediction without additional evidence. This distinction is central to the paper's stated purpose of self-directed preventive care, so the mismatch must be addressed by either reframing the claims or re-analyzing with an appropriate longitudinal or incident-disease design.
  2. [§IV-B, §IV-C] The central validation is post hoc and unfalsifiable as presented. The authors first obtain the SHAP top-three features, then search the literature for citations linking each feature to the disease, without a pre-specified alignment criterion, a quantitative overlap metric, or any negative control. With only three features per disease and flexible narrative interpretation, finding supporting citations for common demographic and lifestyle variables is close to guaranteed; the reported 'strong alignment' is therefore consistent with a model that uses spurious or confounded associations. To support the claim, the authors should pre-register the feature-disease hypotheses, define a measurable alignment score with a threshold, compare the observed alignment to a null distribution (e.g., random features or models trained on shuffled labels), and report cases where the literature did not support a top feature.
  3. [§IV-B1, §IV-B3, Table V, Table VII] The reinterpretation of employment and marital status as proxies for age is not demonstrated. For employment, Table V compares only students against retirees, but the employment variable includes 'Unable to work,' which the authors themselves describe as a likely consequence of chronic disease, and the 'Others' category conflates several heterogeneous statuses. The claim that the model uses these variables 'as an indicator of age' requires direct evidence (e.g., showing that the SHAP dependence on employment is mediated by age, or that the model's predictions track age-like patterns after controlling for employment status). Without such evidence, the cited age-risk literature does not validate the model's most influential features for the 8 of 13 diseases where employment is the top predictor.
  4. [§IV-B] The exclusion of general-health features is applied inconsistently. The authors remove general health, physical health, and poor-health days because they are 'more likely to be consequences, rather than causes,' yet they retain mental health as the top predictor for depressive disorder and retain 'Unable to work' within the employment feature without applying the same consequence-based criterion. This selective application weakens the cause/consequence filter as a principled design choice and makes the resulting top-feature lists appear constructed to align with the medical literature. A consistent and pre-specified feature-inclusion rule is needed.
  5. [§V] The limitations section states that 'a model's alignment with medical literature doesn't guarantee trustworthiness' and that alignment might reinforce existing biases, yet the Abstract and Conclusion assert that the approach 'can be broadly trusted.' This tension is not resolved in the manuscript. Either the conclusion should be weakened to match the acknowledged limitations, or the authors should provide evidence that the validation, despite its limitations, supports the specific trustworthiness claim being made.
minor comments (6)
  1. [§IV-A] There is a typo: 'the performance of the models isare comparable' should read 'the performance of the models is comparable' or 'the performances are comparable.'
  2. [§IV-B] The sentence 'We use SHAP to explain of our machine learning models' contains an extra 'of'; it should read 'to explain our machine learning models.'
  3. [§IV-B2] The phrase 'females are more at risk forto asthma' should be corrected to 'females are more at risk for asthma.'
  4. [Table I, Table IV] The feature name 'Difficulty walking' in Table IV (stroke row) does not appear in the categories listed in Table I, where the disability items are described as 'difficulty making decisions, climbing stairs, dressing, doing errands.' The naming should be harmonized with the dataset's actual BRFSS variable.
  5. [§IV-C7] References [28] and [29] are cited out of order in the text; the Sasco et al. citation should appear before the Coughlin citation, or the numbering should be adjusted.
  6. [§IV-C6] The abbreviation 'NSMC' is introduced as 'Non-melanoma skin cancer (NSMC),' but the standard abbreviation used elsewhere is NMSC; please correct the typo.

Circularity Check

1 steps flagged · score 3.0 of 10

One self-definitional feature for depressive disorder weakens the 'all 13 diseases' validation; the rest of the literature matching is external and not circular.

  1. self definitional [Section IV-B (feature exclusion) and Table IV (top predictors for depressive disorder)]
    "Notably, general health, physical health, and days with poor health are excluded, because they are more likely to be consequences, rather than causes, of chronic diseases. ... Depressive disorder: Mental health, Sex, Weight [Table IV]."

    The paper discards general-health and physical-health features as likely consequences of chronic disease, yet retains 'Mental health' as the leading SHAP feature for depressive disorder. In BRFSS, the mental-health input is essentially the number of days the respondent's mental health was not good—a direct symptom of depression, not an independent causal risk factor. The literature cited for this feature (Liu et al., bidirectional asthma-mental-health link) does not establish an external risk factor; it validates a feature that is, by construction, the same construct as the target label. Thus the claimed 'strong alignment' for depressive disorder is guaranteed by input-label overlap, not by independent medical confirmation.

full rationale

The paper's central validation approach compares SHAP-identified features with external medical literature, which is generally independent and not definitionally circular. There are no load-bearing self-citations, no fitted parameters renamed as predictions, and no uniqueness theorem imported from the authors' prior work. The employment-as-age and marital-status-as-age reinterpretations are post hoc and flexible, but they are data-driven rather than constructed to equal the target. The main genuine circularity is the depressive disorder row of Table IV: mental health, the top predictor, is the same construct as the outcome, while the paper explicitly excluded analogous general-health and physical-health features as consequences. That single self-definitional feature means the 'validation holds across all 13 diseases' claim is not fully independent. The paper's own limitation section honestly notes that literature alignment does not guarantee trustworthiness, which counts against a charge of deliberate circularity. On balance, the derivation is mostly self-contained and externally anchored, but the one definitional overlap and the post hoc flexibility justify a low-to-moderate circularity score rather than zero.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The central claim rests on domain assumptions about survey data quality, proxy variables, and the representativeness of the chosen literature. Several analysis choices (k=3, SHAP sample sizes, post-hoc feature exclusions) are hand-selected and directly affect whether the alignment appears. No invented entities are introduced.

free parameters (3)
  • Neural network hyperparameters and architecture = Partially reported: learning rate 0.001, batch size 32, 50 epochs; layer sizes omitted
    These choices affect which features SHAP ranks highest; without the full architecture, the SHAP ranking is not reproducible and could change with different hyperparameters.
  • SHAP sample sizes = 500 explained points; 100 background points
    SHAP values are averaged over 500 points and 100 k-means centers; no sensitivity analysis shows that the top-3 features are stable across different samples.
  • Number of top features validated (k) = 3 per disease
    The success of the validation depends on k; no justification is given for k=3, and a larger or smaller k could change the alignment result.
assumptions (5)
  • domain assumption BRFSS self-reported responses accurately measure lifestyle factors and diagnosed disease status
    The entire model and validation treat survey answers as ground truth; errors in self-reporting, memory, stigma, or misunderstanding propagate into both predictions and explanations. (Section II-A)
  • ad hoc to paper Employment and marital status can be used as proxies for age
    The authors reinterpret the top-ranked employment and marital status features as age indicators after seeing the SHAP results; this interpretation is load-bearing for validating 8 of 13 diseases. (Sections IV-B1, IV-B3)
  • ad hoc to paper General health, physical health, and poor-health days are consequences, not causes, of chronic disease
    This assumption justifies excluding potentially dominant features from the validation and is asserted without independent evidence. (Section IV-B)
  • domain assumption The cited medical literature is representative and correctly interpreted as supporting the features
    The alignment is only as strong as the authors' selection and reading of the references; no systematic review or independent adjudication is used. (Section IV-C)
  • domain assumption SHAP values computed on 500 points with 100 background centers represent the global model behavior
    The authors acknowledge computational limits and do not show that the top-3 ranking is stable across different SHAP samples. (Section III-B)

how reviews work

0 comments
Cite this review

Pith. "Pith review of Trustworthy Chronic Disease Risk Prediction For Self-Directed Preventive Care via Medical Literature Validation." pith.science (2026). https://pith.science/paper/UOK4L7HT

@misc{pith2026250617620,
  author       = {Pith},
  title        = {Pith review of: Trustworthy Chronic Disease Risk Prediction For Self-Directed Preventive Care via Medical Literature Validation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UOK4L7HT}},
  note         = {Machine review of arXiv:2506.17620}
}
read the original abstract

Chronic diseases are long-term, manageable, yet typically incurable conditions, highlighting the need for effective preventive strategies. Machine learning has been widely used to assess individual risk for chronic diseases. However, many models rely on medical test data (e.g. blood results, glucose levels), which limits their utility for proactive self-assessment. Additionally, to gain public trust, machine learning models should be explainable and transparent. Although some research on self-assessment machine learning models includes explainability, their explanations are not validated against established medical literature, reducing confidence in their reliability. To address these issues, we develop deep learning models that predict the risk of developing 13 chronic diseases using only personal and lifestyle factors, enabling accessible, self-directed preventive care. Importantly, we use SHAP-based explainability to identify the most influential model features and validate them against established medical literature. Our results show a strong alignment between the models' most influential features and established medical literature, reinforcing the models' trustworthiness. Critically, we find that this observation holds across 13 distinct diseases, indicating that this machine learning approach can be broadly trusted for chronic disease prediction. This work lays the foundation for developing trustworthy machine learning tools for self-directed preventive care. Future research can explore other approaches for models' trustworthiness and discuss how the models can be used ethically and responsibly.

Figures

Figures reproduced from arXiv: 2506.17620 by the authors.

Figure 1
Figure 1. (a) Architecture of the residual block, forming part of the machine learning model; (b) Architecture of the entire machine learning model. IV. RESULTS A. Machine Learning Model Training We train machine learning models using data batches of 32, with a weighted cross-entropy loss inversely proportional to class frequency to address the imbalance. We use the Adam optimizer, with an initial learning rate of 0.001 (halv… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

45 extracted references · 39 canonical work pages

  1. [1]

    About chronic diseases,

    “About chronic diseases,” Centers for Disease Control and Prevention, https://www.cdc.gov/chronic-disease/about/ (accessed Jun. 18, 2025)

  2. [2]

    Trends in Multiple Chronic Conditions Among US Adults, By Life Stage, Behavioral Risk Factor Surveillance System, 2013-2023,

    K. B. Watsonet al., “Trends in Multiple Chronic Conditions Among US Adults, By Life Stage, Behavioral Risk Factor Surveillance System, 2013-2023,” Preventing Chronic Disease, vol. 22, Apr. 2025, doi: https://doi.org/10.5888/pcd22.240539

  3. [3]

    Chronic kidney disease prediction using machine learning techniques,

    D. A. Debal and T. M. Sitote, “Chronic kidney disease prediction using machine learning techniques,” Journal of Big Data, vol. 9, no. 1, Nov. 2022, doi: https://doi.org/10.1186/s40537-022-00657-5

  4. [4]

    Construction of machine learning diagnostic models for cardiovascular pan-disease based on blood routine and biochemical detection data,

    Z. Wanget al., “Construction of machine learning diagnostic models for cardiovascular pan-disease based on blood routine and biochemical detection data,” Cardiovascular Diabetology, vol. 23, no. 1, Sep. 2024, doi: https://doi.org/10.1186/s12933-024-02439-0

  5. [5]

    Efficient Automated Disease Diagnosis Using Machine Learning Models,

    N. Kumaret al., “Efficient Automated Disease Diagnosis Using Machine Learning Models,” Journal of Healthcare Engineering, vol. 2021, pp. 1–13, May 2021, doi: https://doi.org/10.1155/2021/9983652

  6. [6]

    Trust and medical ai: The challenges we face and the expertise needed to overcome them,

    T. P. Quinn, M. Senadeera, S. Jacobs, S. Coghlan, and V . Le, “Trust and medical ai: The challenges we face and the expertise needed to overcome them,”Journal of the American Medical Informatics Association, vol. 28, no. 4, pp. 890–894, Dec. 2020. doi:10.1093/jamia/ocaa268

  7. [7]

    Towards a rigorous science of interpretable machine learning,

    F. Doshi-Velez and B. Kim, “Towards a rigorous science of interpretable machine learning,” arXiv.org, https://arxiv.org/abs/1702.08608 (accessed Jun. 18, 2025)

  8. [8]

    Interactive Diabetes Risk Prediction Using Explainable Machine Learning: A Dash-Based Approach with SHAP, LIME, and Comorbidity Insights

    U. Allani, “Interactive diabetes risk prediction using explainable ma- chine learning: A dash-based approach with shap, lime, and comorbidity insights,” arXiv.org, https://arxiv.org/abs/2505.05683 (accessed Jun. 17, 2025)

Show all 45 references
  1. [9]

    S. B. Akteret al.,Improving heart disease probability prediction sensitivity with a grow network model, Mar. 2024. doi:10.1101/2024.02.28.24303495

  2. [10]

    CDC - 2023 BRFSS survey data and Documenta- tion,

    “CDC - 2023 BRFSS survey data and Documenta- tion,” Centers for Disease Control and Prevention, https://www.cdc.gov/brfss/annual data/annual 2023.html (accessed Jun. 18, 2025)

  3. [11]

    Predicting asthma using imbalanced data mod- eling techniques: Evidence from 2019 Michigan BRFSS data,

    N. Budhathokiet al., “Predicting asthma using imbalanced data mod- eling techniques: Evidence from 2019 Michigan BRFSS data,” PLoS ONE, vol. 18, no. 12, pp. e0295427–e0295427, Dec. 2023, doi: https://doi.org/10.1371/journal.pone.0295427

  4. [12]

    Obesity and Hypertension,

    G. Seravalle and G. Grassi, “Obesity and Hypertension,” Springer eBooks, pp. 65–79, Jan. 2024, doi: https://doi.org/10.1007/978-3-031- 62491-9 5

  5. [13]

    Age-related changes in the risk of high blood pressure,

    W. Chenget al., “Age-related changes in the risk of high blood pressure,” Frontiers in Cardiovascular Medicine, vol. 9, Sep. 2022, doi: https://doi.org/10.3389/fcvm.2022.939103

  6. [14]

    Management of high cholesterol levels in older people,

    M. Bertolottiet al., “Management of high cholesterol levels in older people,” Geriatrics & Gerontology International, vol. 19, no. 5, pp. 375–383, Mar. 2019, doi: https://doi.org/10.1111/ggi.13647

  7. [15]

    Cholesterol Production in Obesity,

    T. A. Miettinen, “Cholesterol Production in Obesity,” Circulation, vol. 44, no. 5, pp. 842–850, Nov. 1971, doi: https://doi.org/10.1161/01.cir.44.5.842

  8. [16]

    General Health Checks in Adult Primary Care,

    D. T. Lisset al., “General Health Checks in Adult Primary Care,” JAMA, vol. 325, no. 22, p. 2294, Jun. 2021, doi: https://doi.org/10.1001/jama.2021.6524

  9. [17]

    Aging and Diseases of the Heart,

    S. Sternet al., “Aging and Diseases of the Heart,” Circulation, vol. 108, no. 14, Oct. 2003, doi: https://doi.org/10.1161/01.cir.0000086898.96021.b9

  10. [18]

    W. B. Kannel,Cardiovascular Drugs and Therapy, vol. 11, no. 1+, pp. 199–212, 1997. doi:10.1023/a:1007792820944

  11. [19]

    Influence of age and health behaviors on stroke risk: Lessons from longitudinal studies,

    M. Kelly-Hayes, “Influence of age and health behaviors on stroke risk: Lessons from longitudinal studies,”Journal of the American Geriatrics Society, vol. 58, no. s2, Oct. 2010. doi:10.1111/j.1532- 5415.2010.02915.x

  12. [20]

    Obesity and the lung: 4{middle dot} obesity and asthma,

    D. D. Sin and E. R. Sutherland, “Obesity and the lung: 4{middle dot} obesity and asthma,”Thorax, vol. 63, no. 11, pp. 1018–1023, Nov. 2008. doi:10.1136/thx.2007.086819

  13. [21]

    Bidirectional associations between asthma and types of mental disorders,

    X. Liuet al., “Bidirectional associations between asthma and types of mental disorders,”The Journal of Allergy and Clinical Immunology: In Practice, vol. 11, no. 3, Mar. 2023. doi:10.1016/j.jaip.2022.11.027

  14. [22]

    Hormones, sex, and asthma,

    J. A. Yunget al., “Hormones, sex, and asthma,”Annals of Allergy, Asthma & Immunology, vol. 120, no. 5, pp. 488–494, May 2018. doi:10.1016/j.anai.2018.01.016

  15. [23]

    Age distribution of patients with ad- vanced non-melanoma skin cancer in the United States,

    S. Dacosta Byfieldet al., “Age distribution of patients with ad- vanced non-melanoma skin cancer in the United States,”Archives of Dermatological Research, vol. 305, no. 9, pp. 845–850, Apr. 2013. doi:10.1007/s00403-013-1357-2

  16. [24]

    Alcohol consumption decreases the protection efficiency of the antioxidant network and increases the risk of sunburn in human skin,

    M. E. Darvinet al., “Alcohol consumption decreases the protection efficiency of the antioxidant network and increases the risk of sunburn in human skin,”Skin Pharmacology and Physiology, vol. 26, no. 1, pp. 45–51, Nov. 2012. doi:10.1159/000343908

  17. [25]

    Dietary components and the risk of non-melanoma skin cancer: A systematic review of Epidemiological Studies,

    E. Hezavehet al., “Dietary components and the risk of non-melanoma skin cancer: A systematic review of Epidemiological Studies,”Critical Reviews in Food Science and Nutrition, vol. 63, no. 21, pp. 5290–5305, Dec. 2021. doi:10.1080/10408398.2021.2016600

  18. [26]

    Changing trends in the disease burden of non-melanoma skin cancer globally from 1990 to 2019 and its predicted level in 25 years,

    W. Huet al., “Changing trends in the disease burden of non-melanoma skin cancer globally from 1990 to 2019 and its predicted level in 25 years,”BMC Cancer, vol. 22, no. 1, Jul. 2022. doi:10.1186/s12885- 022-09940-3

  19. [27]

    The age of cancer,

    R. A. DePinho, “The age of cancer,”Nature, vol. 408, no. 6809, pp. 248–254, Nov. 2000. doi:10.1038/35041694

  20. [28]

    Tobacco smoking and cancer: A brief review of recent epidemiological evidence,

    A. J. Sascoet al., “Tobacco smoking and cancer: A brief review of recent epidemiological evidence,”Lung Cancer, vol. 45, Aug. 2004. doi:10.1016/j.lungcan.2004.07.998

  21. [29]

    Social determinants of colorectal cancer risk, stage, and survival: A systematic review,

    S. S. Coughlin, “Social determinants of colorectal cancer risk, stage, and survival: A systematic review,”International Journal of Colorectal Disease, vol. 35, no. 6, pp. 985–995, Apr. 2020. doi:10.1007/s00384- 020-03585-z

  22. [30]

    Systematic review with meta-analysis of the epidemiological evidence relating smoking to COPD, chronic bronchitis and emphysema,

    B. A. Foreyet al., “Systematic review with meta-analysis of the epidemiological evidence relating smoking to COPD, chronic bronchitis and emphysema,”BMC Pulmonary Medicine, vol. 11, no. 1, Jun. 2011. doi:10.1186/1471-2466-11-36

  23. [31]

    Chronic bronchitis in the general population: Influence of age, gender and socio-economic conditions,

    A. Ferr ´eet al., “Chronic bronchitis in the general population: Influence of age, gender and socio-economic conditions,”Respiratory Medicine, vol. 106, no. 3, pp. 467–471, Mar. 2012. doi:10.1016/j.rmed.2011.12.002

  24. [32]

    Association of Exercise Tolerance with Respiratory Health Outcomes in Mild-to-Moderate COPD,

    Q. Wanet al., “Association of Exercise Tolerance with Respiratory Health Outcomes in Mild-to-Moderate COPD,” Annals of the American Thoracic Society, Nov. 2024, doi: https://doi.org/10.1513/annalsats.202404-408oc

  25. [33]

    Gender differences in depression: Critical review,

    M. Piccinelli and G. Wilkinson, “Gender differences in depression: Critical review,”British Journal of Psychiatry, vol. 177, no. 6, pp. 486–492, 2000. doi:10.1192/bjp.177.6.486

  26. [34]

    Major depressive disorder,

    W. Marxet al., “Major depressive disorder,”Nature Reviews Disease Primers, vol. 9, no. 1, Aug. 2023. doi:10.1038/s41572-023-00454-1

  27. [35]

    Chronic kidney disease in older people,

    A. S. Leveyet al., “Chronic kidney disease in older people,”JAMA, vol. 314, no. 6, p. 557, Aug. 2015. doi:10.1001/jama.2015.6753

  28. [36]

    Perfluoroalkyl Chemicals and Chronic Kidney Disease in US Adults,

    A. Shankaret al., “Perfluoroalkyl Chemicals and Chronic Kidney Disease in US Adults,” American Journal of Epidemiology, vol. 174, no. 8, pp. 893–900, Aug. 2011, doi: https://doi.org/10.1093/aje/kwr171

  29. [37]

    A community-based study of explanatory factors for the excess risk for early renal function decline in blacks vs whites with diabetes,

    J. S. Kropet al., “A community-based study of explanatory factors for the excess risk for early renal function decline in blacks vs whites with diabetes,”Archives of Internal Medicine, vol. 159, no. 15, p. 1777, Aug

  30. [38]

    The influence of ageing on the development and management of rheumatoid arthritis,

    A. M. Bootset al., “The influence of ageing on the development and management of rheumatoid arthritis,”Nature Reviews Rheumatology, vol. 9, no. 10, pp. 604–613, Jun. 2013. doi:10.1038/nrrheum.2013.92

  31. [39]

    Sex and management of rheumatoid arthritis,

    E. G. Favalliet al., “Sex and management of rheumatoid arthritis,” Clinical Reviews in Allergy & Immunology, vol. 56, no. 3, pp. 333–345, Jan. 2018. doi:10.1007/s12016-018-8672-5

  32. [40]

    Diabetes educators,

    D. Sherr and R. D. Lipman, “Diabetes educators,”American Journal of Preventive Medicine, vol. 44, no. 4, Apr. 2013. doi:10.1016/j.amepre.2012.12.013

  33. [41]

    Dynamics of diabetes and obesity: Epidemiolog- ical perspective,

    A. Boleset al., “Dynamics of diabetes and obesity: Epidemiolog- ical perspective,”Biochimica et Biophysica Acta (BBA) - Molecu- lar Basis of Disease, vol. 1863, no. 5, pp. 1026–1036, May 2017. doi:10.1016/j.bbadis.2017.01.016

  34. [42]

    Diabetes mellitus and alcohol,

    A. van de Wiel, “Diabetes mellitus and alcohol,”Diabetes/Metabolism Research and Reviews, vol. 20, no. 4, pp. 263–267, Jul. 2004. doi:10.1002/dmrr.492

  35. [43]

    Explainable, trustworthy, and ethical ma- chine learning for healthcare: A survey,

    K. Rasheedet al., “Explainable, trustworthy, and ethical ma- chine learning for healthcare: A survey,” Computers in Biology and Medicine, vol. 149, no. 106043, p. 106043, Oct. 2022, doi: https://doi.org/10.1016/j.compbiomed.2022.106043

  36. [44]

    Gender differences in coronary heart disease,

    A. H. E. M. Maas and Y . E. A. Appelman, “Gender differences in coronary heart disease,” Netherlands Heart Journal, vol. 18, no. 12, pp. 598–603, Nov. 2010, doi: https://doi.org/10.1007/s12471-010-0841-y

  37. [1999]

    doi:10.1001/archinte.159.15.1777

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.