Logistic Regression Analysis on the Dietary Behavior and the Risk of Nutritional Deficiency Dermatosis: The Case of Bicol Region, Philippines

John Ben S Temones

REVIEW 3 major objections 1 minor 4 references

Logistic Regression Analysis on the Dietary Behavior and the Risk of Nutritional Deficiency Dermatosis: The Case of Bicol Region, Philippines

T0 review · 3 major / 1 minor · reviewed 2026-05-23 · grok-4.3

Challenge this review Download PDF Read on arXiv ↗

Pith's one-line read Logistic regression on Bicol food purchase data links riboflavin-rich foods to lower nutritional deficiency dermatosis risk.

desk verdict Standard logistic regression on one region's FNRI survey data with concrete metrics but thin methods and clear confounding risks. read the letter →

arxiv 2411.00063 v2 submitted 2024-10-31 q-bio.QM

John Ben S Temones This is my paper

classification q-bio.QM

keywords nutritionaldeficiencydermatosislogisticregressiondietarybehaviorriboflavinintakeBicolregionmalnutritionPhilippinesfoodpurchasesurvey

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 applies logistic regression to survey data on daily food purchases in the Bicol region of the Philippines to test associations between dietary patterns and the risk of nutritional deficiency dermatosis. It reports that each additional gram of riboflavin-rich items such as meat, eggs, and dairy reduces the odds by up to 3 percent, while rice and plant-based foods show a smaller protective effect of 0.3 percent per gram despite low riboflavin content. The model fits the observed 15.75 percent prevalence rate with Nagelkerke R² of 0.765, 94.1 percent accuracy, and 84.5 percent precision. A sympathetic reader would care because the results identify measurable dietary levers that could inform targeted nutrition efforts in regions with ongoing malnutrition concerns.

What carries the argument

Logistic regression model that uses daily quantities of purchased food items as predictors to estimate individual NDD risk probability.

What would settle it

A follow-up study that measures actual riboflavin blood levels and confirms NDD diagnoses in the same population to test whether the reported per-gram risk reductions persist.

Watch

Extended reading notes

Core claim

The study establishes through logistic regression that riboflavin-rich foods lower NDD risks by up to 3% per gram while rice and plant-based diets reduce odds by 0.3% per gram; the model achieves Nagelkerke R² = 0.765, accuracy = 94.1%, and precision = 84.5% on FNRI food purchase data from the Bicol region where NDD prevalence is 15.75% and concentrated in Masbate and Camarines Sur.

Load-bearing premise

Recorded food purchase quantities from the survey accurately reflect actual individual dietary intake and riboflavin consumption without major unmeasured confounding.

Editorial extensions

If this is right

Higher daily intake of meat, eggs, and dairy is associated with substantially greater NDD risk reduction than reliance on rice and plant staples.
The fitted model can classify new cases from food purchase patterns with over 94 percent accuracy.
Public health measures such as rice enrichment and improved access to animal-source foods could lower regional NDD rates.
Interventions should prioritize provinces with the highest case concentration such as Masbate and Camarines Sur.

Reading between the lines

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

If the purchase data serve as a reliable proxy, fortification programs targeting staples could produce measurable population-level effects on skin deficiency conditions.
Repeating the analysis with direct intake diaries or biomarkers would clarify how much of the observed effect is due to riboflavin versus correlated factors like overall diet quality.
The same regression framework could be applied to other micronutrient deficiency outcomes in comparable low-resource settings.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

The paper runs logistic regression on FNRI household purchase data for the Bicol region and links per-gram food amounts to NDD odds, reporting 0.3% reduction for rice and up to 3% for meat/eggs/dairy, plus model stats of Nagelkerke R² 0.765, 94.1% accuracy, and 84.5% precision. Prevalence sits at 15.75% with concentration in two provinces. That is the core output. It is a direct application of an off-the-shelf model to an existing regional slice of national survey data, aimed at local nutrition planning rather than any methodological advance. The concrete performance numbers and per-gram figures are the parts that stand out as usable for someone tracking regional patterns. The work does not introduce new variables, new estimators, or first-principles derivations; it fits the model and reports coefficients. The soft spots are the missing pieces that matter for interpretation. No sample size appears, no variable coding or missing-data rules, and no validation steps. Household purchase records are treated as proxies for individual riboflavin intake without discussion of waste, intra-household sharing, cooking losses, or total energy. Socioeconomic status, market access, and correlated behaviors are not mentioned as controls. The stress-test concern about purchases failing to equal actual intake therefore lands directly on the reported effects; if that assumption does not hold, the odds reductions cannot be read as dietary impacts. The paper is aimed at public-health workers focused on Bicol or similar Philippine settings who might use the numbers for program targeting. It is not set up for readers seeking generalizable claims about nutrition and dermatosis. I would not bring it to a reading group. It does not look ready for peer review in its current form because the central quantitative claims rest on unexamined data assumptions and lack the basic methodological transparency needed for assessment.

Referee Report

3 major / 1 minor

Summary. The manuscript applies logistic regression to FNRI survey data from the Bicol region to examine associations between household food purchase quantities (emphasizing riboflavin sources) and nutritional deficiency dermatosis (NDD) risk. It reports 15.75% NDD prevalence, per-gram odds reductions of 0.3% for rice/plant-based foods and up to 3% for meat/eggs/dairy, and model metrics of Nagelkerke R² = 0.765, accuracy = 94.1%, precision = 84.5%. The study concludes with recommendations for enriched rice, market access, and dietary education.

Significance. If the associations prove robust after addressing data limitations and confounding, the per-gram effect sizes and regional prevalence estimates could inform targeted nutrition policy in the Philippines. The reported model performance metrics provide a concrete basis for predictive utility, and the use of existing FNRI survey data is a strength for reproducibility. However, the observational design and unverified assumptions about intake measurement substantially weaken causal interpretation of the dietary effects.

major comments (3)

[Abstract] Abstract: The reported odds reductions (0.3–3% per gram) and performance metrics (Nagelkerke R² = 0.765, accuracy = 94.1%) are presented without sample size, number of observations, variable definitions, missing-data handling, or any validation procedure (e.g., cross-validation or hold-out set). These omissions are load-bearing because the central claims rest on the fitted logistic coefficients being reliable and generalizable.
[Abstract] Abstract / Results: The interpretation of household-level purchase quantities (e.g., rice 1590.93 g/day) as individual riboflavin intake and direct predictors of NDD odds is not justified. FNRI data are typically aggregate purchases; the manuscript provides no evidence of adjustment for intra-household allocation, food waste, preparation losses, total energy intake, or socioeconomic status, all of which are plausible confounders correlated with both purchases and NDD risk.
[Abstract] Abstract: Observational purchase data are used to claim specific risk reductions, yet no causal inference framework, instrumental variables, or sensitivity analysis for unmeasured confounding is mentioned. This undermines the policy recommendations (enriched rice, market access) that treat the per-gram effects as actionable dietary impacts.

minor comments (1)

[Abstract] The abstract would be clearer if it explicitly listed the food categories included as predictors and stated whether the model was adjusted for age, sex, or household size.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the detailed and constructive feedback. We address each major comment below, indicating where revisions will be made to improve clarity, acknowledge limitations, and strengthen the presentation of our observational analysis.

read point-by-point responses

Referee: [Abstract] Abstract: The reported odds reductions (0.3–3% per gram) and performance metrics (Nagelkerke R² = 0.765, accuracy = 94.1%) are presented without sample size, number of observations, variable definitions, missing-data handling, or any validation procedure (e.g., cross-validation or hold-out set). These omissions are load-bearing because the central claims rest on the fitted logistic coefficients being reliable and generalizable.

Authors: We agree that the abstract should include these key details for proper evaluation of the results. In the revised manuscript, we will expand the abstract to report the sample size and number of observations from the FNRI survey, provide brief definitions of the main variables (e.g., household purchase quantities in grams per day), note the handling of missing data, and mention the validation approach (e.g., any cross-validation or goodness-of-fit checks) used for the logistic regression model. revision: yes
Referee: [Abstract] Abstract / Results: The interpretation of household-level purchase quantities (e.g., rice 1590.93 g/day) as individual riboflavin intake and direct predictors of NDD odds is not justified. FNRI data are typically aggregate purchases; the manuscript provides no evidence of adjustment for intra-household allocation, food waste, preparation losses, total energy intake, or socioeconomic status, all of which are plausible confounders correlated with both purchases and NDD risk.

Authors: This is a substantive limitation of the data. The FNRI survey provides household-level food purchase quantities, which we analyzed as proxies for dietary patterns associated with NDD risk at the household level. No adjustments for intra-household allocation, waste, losses, energy intake, or SES were performed, as these variables were not available or derivable in the dataset. In revision, we will add an explicit limitations paragraph clarifying that results reflect household-level associations and discuss potential confounding. We will also state that individual-level inferences require caution. revision: partial
Referee: [Abstract] Abstract: Observational purchase data are used to claim specific risk reductions, yet no causal inference framework, instrumental variables, or sensitivity analysis for unmeasured confounding is mentioned. This undermines the policy recommendations (enriched rice, market access) that treat the per-gram effects as actionable dietary impacts.

Authors: We concur that the study is strictly observational and does not employ causal methods such as instrumental variables or sensitivity analyses for unmeasured confounding. The per-gram odds ratios represent adjusted associations from the logistic model, not causal effects. We will revise the discussion and conclusion sections to emphasize the associative nature of the findings, remove or qualify language implying direct risk reduction, and reframe policy suggestions (e.g., enriched rice and market access) as hypotheses that could be tested in future studies rather than direct recommendations based on causal impacts. revision: yes

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; logistic regression on external survey data

full rationale

The paper applies standard logistic regression to FNRI survey data to estimate associations between per-gram food purchases and NDD odds. Reported quantities (odds reductions of 0.3–3% per gram, Nagelkerke R²=0.765, accuracy=94.1%) are direct outputs of the fitted model on the input data. No equations reduce results to fitted parameters by construction, no self-citations form a load-bearing chain, and no ansatz or uniqueness claim is imported from prior author work. The derivation is self-contained empirical modeling against an external dataset.

Assumptions & free parameters 1 free parameters · 2 assumptions · 0 invented entities

The central claim rests on fitted logistic regression coefficients and standard statistical assumptions about the survey data. No new physical entities are postulated.

free parameters (1)

logistic regression coefficients (beta values for each food category)
These are estimated from the FNRI survey data to produce the reported per-gram odds reductions and model performance metrics.

assumptions (2)

domain assumption Logistic regression assumptions hold (linearity in the logit, independence of observations, no perfect multicollinearity).
Invoked implicitly by fitting and interpreting the model on survey data.
domain assumption Food purchase quantities recorded in the FNRI survey accurately proxy actual dietary intake and riboflavin consumption.
Required to translate purchase patterns into nutritional risk estimates.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Logistic Regression Analysis on the Dietary Behavior and the Risk of Nutritional Deficiency Dermatosis: The Case of Bicol Region, Philippines." pith.science (2026). https://pith.science/paper/2411.00063

@misc{pith2026241100063,
  author       = {Pith},
  title        = {Pith review of: Logistic Regression Analysis on the Dietary Behavior and the Risk of Nutritional Deficiency Dermatosis: The Case of Bicol Region, Philippines},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2411.00063}},
  note         = {Machine review of arXiv:2411.00063}
}

read the original abstract

This study explores the link between dietary behavior and the risk of nutritional deficiency dermatoses (NDD) in the Bicol region, where malnutrition remains a concern. Using regression analysis on FNRI data, it examines food purchase patterns, particularly riboflavin intake. Findings show an NDD risk prevalence of 15.75%, with Masbate and Camarines Sur contributing over half of cases. While rice (1590.93 g/day) and plant-based diets (523.30 g/day) are not rich in riboflavin, they still reduce NDD odds by 0.3% per gram. Riboflavin-rich foods like meat, eggs, and dairy lower risks by up to 3% per gram. The logistic regression model demonstrated strong performance (Nagelkerke = 0.765, accuracy = 94.1%, precision = 84.5%). Findings highlight the need for nutrition interventions, including enriched rice, better market access, and food diversity education to improve riboflavin intake and mitigate NDD risks.

Figures

Figures reproduced from arXiv: 2411.00063 by the authors.

**Figure 3.** Performance Metrics of the Model From table 3, it can be gleaned that cereals and cereal products is a significant predictor (OR=0.997, p<.001) of NDD, which means that for every additional gram of purchase of these products, the odds of developing NDD decrease by 0.3%. The case is similar to that of vegetables, even when they are not good sources of riboflavin. Notable sources of riboflavin, namely meat and meat pr… view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

4 extracted references · 4 canonical work pages

[1]

Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes , 50(2), 179 –211. https://doi.org/10.1016/0749- 5978(91)90020-T Amugsi, D.A., Mittelmark, M.B. & Lartey, A. An analysis of socio-demographic patterns in child malnutrition trends using Ghana demographic and health survey data in the period 1993 –2008. B...

work page doi:10.1016/0749- 1991
[2]

R., Chen, Y

Isaura, E. R., Chen, Y. C., & Yang, S. H. (2022). Childhood socioeconomic status and adulthood dietary diversity among Indonesian adults. Frontiers in Nutrition , 9, 948208. https://doi.org/10.3389/FNUT.2022.948208/BIBTEX Ivers LC, Cullen KA. Food insecurity: special considerations for women. Am J Clin Nutr. 2011 Dec;94(6):1740S-1744S. doi: 10.3945/ajcn.1...

work page doi:10.3389/fnut.2022.948208/bibtex 2022
[3]

Johnson, L.E

PMID: 22089447; PMCID: PMC3226027. Johnson, L.E. (2024). Riboflavin Deficiency - Nutritional Disorders - MSD Manual Professional Edition . Retrieved March 26, 2025, from https://www.msdmanuals.com/professional/nutritional-disorders/vitamin- deficiency-dependency-and-toxicity/riboflavin-deficiency Kennedy, G., Ballard, T., & Dop, M. C. (2007). Guidelines f...

work page arXiv 2024
[4]

(2024, August 15)

https://doi.org/10.3389/FNUT.2022.866400 Mapa, C.D.S. (2024, August 15). Family Income and Expenditure Survey. Philippine Statistics Authority: Philippines. Retrieved from https://psa.gov.ph/statistics/income-expenditure/fies Marashly ET, Bohlega SA. Riboflavin Has Neuroprotective Potential: Focus on Parkinson's Disease and Migraine. Front Neurol. 2017 Ju...

work page doi:10.3389/fnut.2022.866400 2022

Pith tools

Reviewed May 23, 2026 · model on record in the stance chip above.

[1] [1]

Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes , 50(2), 179 –211. https://doi.org/10.1016/0749- 5978(91)90020-T Amugsi, D.A., Mittelmark, M.B. & Lartey, A. An analysis of socio-demographic patterns in child malnutrition trends using Ghana demographic and health survey data in the period 1993 –2008. B...

work page doi:10.1016/0749- 1991

[2] [2]

R., Chen, Y

Isaura, E. R., Chen, Y. C., & Yang, S. H. (2022). Childhood socioeconomic status and adulthood dietary diversity among Indonesian adults. Frontiers in Nutrition , 9, 948208. https://doi.org/10.3389/FNUT.2022.948208/BIBTEX Ivers LC, Cullen KA. Food insecurity: special considerations for women. Am J Clin Nutr. 2011 Dec;94(6):1740S-1744S. doi: 10.3945/ajcn.1...

work page doi:10.3389/fnut.2022.948208/bibtex 2022

[3] [3]

Johnson, L.E

PMID: 22089447; PMCID: PMC3226027. Johnson, L.E. (2024). Riboflavin Deficiency - Nutritional Disorders - MSD Manual Professional Edition . Retrieved March 26, 2025, from https://www.msdmanuals.com/professional/nutritional-disorders/vitamin- deficiency-dependency-and-toxicity/riboflavin-deficiency Kennedy, G., Ballard, T., & Dop, M. C. (2007). Guidelines f...

work page arXiv 2024

[4] [4]

(2024, August 15)

https://doi.org/10.3389/FNUT.2022.866400 Mapa, C.D.S. (2024, August 15). Family Income and Expenditure Survey. Philippine Statistics Authority: Philippines. Retrieved from https://psa.gov.ph/statistics/income-expenditure/fies Marashly ET, Bohlega SA. Riboflavin Has Neuroprotective Potential: Focus on Parkinson's Disease and Migraine. Front Neurol. 2017 Ju...

work page doi:10.3389/fnut.2022.866400 2022