Dietary Mineral Intake and Vascular Health in Patients with Long COVID-19: The BioICOPER Study

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

1.Mineral intakes correlate with total energy intake. Without energy adjustment (nutrients per 1000 kcal or energy as a covariate), associations may reflect overall diet quantity rather than specific mineral effects. Re-run models including total energy intake.

2.Hypertension, dyslipidemia, diabetes, and obesity lie on the causal pathway between mineral intake and arterial stiffness. Adjusting for them attenuates true effects. Present a reduced model (age, sex, smoking, physical activity, alcohol, energy) alongside the full model and discuss overadjustment.

3. Mg, Se, K, P show significant inverse associations with cfPWV in regression (Figs 3–5), but Table 3 shows only Mg reaching significance. Dichotomization and non‑linearity may explain this. Use restricted cubic splines and discuss the discrepancy.

4. Nine minerals × three outcomes = 27 primary associations, plus sex‑stratified analyses, with no correction. Risk of false positives is high. Apply FDR (q<0.10) or Bonferroni, and temper interpretation of marginal findings ( P with VAI, p=0.045).

5. Supplement use (common for Mg, Ca, Zn, Se) not assessed; food sources not considered. Collinearity among minerals (Mg, K from plant foods) suggests a healthy dietary pattern may drive associations. Report supplement prevalence or add pattern analysis.

6. Stratified analyses show significant associations in women only, but formal sex×mineral interactions are all non‑significant (p>0.05). This does not support effect modification. Present interaction p‑values explicitly and interpret stratified results as descriptive, not as evidence of sex differences.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript addresses a relevant topic, examining the association between dietary mineral intake and vascular health in adults with long COVID. The study has several strengths, including a relatively large sample, the use of a 7-day dietary record, and a detailed vascular assessment based on cIMT, cfPWV, baPWV and the vascular ageing index. However, several methodological and interpretative issues require clarification before the manuscript can be considered for publication. The comments and suggestions are outlined below:

1. In Section 2.3, Dietary Intake Assessment, it is unclear whether the dietary assessment included the consumption of mineral-fortified products and whether minerals added through fortification were included in the estimation of total mineral intake.

2. In Section 2.3, Dietary Intake Assessment, the authors should provide the full name of the dietary reference values or standards used, together with the appropriate reference. The manuscript currently cites reference no. 58 for carbohydrate and fibre intake and reference no. 59 for fat intake; however, the reference for mineral intake appears to be missing in lines 213–216. In addition, the description of how groups with adequate and inadequate mineral intake were defined requires clarification.

3. The Discussion section should include information on the prevalence of dietary supplement use and should also consider the contribution of fortified foods as a source of minerals in the diet. The authors should discuss in greater detail whether, and how, these sources may have influenced total mineral intake.

4. The authors present interesting results regarding phosphorus. However, the Discussion should address in greater detail the potential contribution of phosphorus derived from food additives and indicate the main dietary sources that may substantially increase total phosphorus intake.

5. The information provided in the Strengths and Limitations section, particularly in lines 556–561, is too general and insufficiently detailed.

6. At the end of the manuscript, it is recommended that the authors introduce a separate section entitled: Evidence gaps and future research.

 

Author Response

Reviewer 2

The manuscript addresses a relevant topic, examining the association between dietary mineral intake and vascular health in adults with long COVID. The study has several strengths, including a relatively large sample size, the use of a 7-day dietary record, and a detailed vascular assessment based on cIMT, cfPWV, baPWV, and the vascular aging index.

General response:

We thank Reviewer 2 for the careful and constructive evaluation of our manuscript. In the revised version, we have clarified the dietary assessment methodology, the handling of fortified foods and supplements, the dietary reference standards used to define mineral adequacy, and the interpretation of phosphorus intake. We have also expanded the Strengths and Limitations section and added a new section entitled “Evidence gaps and future research”. Importantly, as also described in our response to Reviewer 1, the main regression analyses were recalculated using energy-adjusted mineral density variables expressed per 1000 kcal/day, hierarchical models, FDR correction, and additional restricted cubic spline analyses. Therefore, some interpretations from the initial version have been revised and are now presented more cautiously as exploratory findings.

However, several methodological and interpretative aspects require clarification before the manuscript can be considered for publication. The comments and suggestions are detailed below:

1. Comment 1

Reviewer comment: In Section 2.3, Assessment of Dietary Intake, it is unclear whether the dietary assessment included consumption of fortified mineral products and whether minerals added by fortification were included in the estimate of total mineral intake.

Response:

We have clarified in Section 2.3 that dietary intake was estimated from the 7-day food record using the EVIDENT tool and the Spanish food composition tables. In this approach, when a fortified food was recorded by the participant and was represented as such in the compositional basis used, its mineral content was incorporated into the estimate of the food consumed. However, the tool did not allow to quantify separately the fraction of minerals from fortification or to distinguish between naturally occurring minerals and minerals added during processing. Therefore, we have added a methodological clarification indicating that the estimated mineral intake represents the intake from the registered foods according to their available composition, but that it was not possible to carry out a specific analysis of fortified products or minerals added by fortification. This point has also been included as a limitation, as it could have produced some misclassification of dietary exposure, especially in minerals that can be added to processed or fortified foods.

Changes made in the manuscript:

Section 2.3. Dietary Intake Assessment: When fortified foods were recorded by participants and were available as such in the food composition database used by the EVIDENT tool, their mineral content was incorporated as part of the corresponding food item. However, the dietary assessment tool did not allow the separate quantification of minerals derived specifically from food fortification, nor did it distinguish naturally occurring minerals from minerals added during food processing or enrichment. Therefore, mineral intake estimates should be interpreted as total dietary intake from recorded foods according to the available food composition data, without separate attribution to fortification.”

2. Comment 2

Reviewer comment:

In Section 2.3, Assessment of Dietary Intake, authors should provide the full name of the dietary reference values or standards used, along with the corresponding reference. Currently, the manuscript cites reference no. 58 for carbohydrate and fiber intake and reference no. 59 for fat intake; however, the reference for mineral intake appears to be missing in lines 213–216. In addition, the description of how groups with adequate and insufficient mineral intake were defined requires clarification.

Response:

We have revised Section 2.3 and Supplementary Table S2 to further specify the standards used. In the revised version, it is expressly stated that the adequacy of mineral intake was defined using the Dietary Reference Values (DRV) of the European Food Safety Authority (EFSA) [1] and the nutritional objectives/recommendations of the Spanish Society of Community Nutrition (SENC) [2] selecting sex-specific cut-off points when available. We have also corrected the bibliographic reference, since in the previous version the references cited corresponded to other macronutrients and did not adequately identify the mineral recommendations. In addition, we have clarified that participants were classified as having adequate intake when they reached the corresponding cut-off point for each mineral and as having inadequate intake or risk when they did not reach it. In the updated analyses, mineral intake variables expressed per 1000 kcal were used in the regression models, while adequacy categories were retained for descriptive marginal mean analyses.

Changes made in the manuscript:

Section 2.3. Dietary Intake Assessment: “Adequate and inadequate/risk intake groups were defined using mineral-specific reference values derived from the European Food Safety Authority (EFSA) Dietary Reference Values [1] for nutrients and the Spanish Society of Community Nutrition (SENC) nutritional objectives/recommended intakes for the Spanish population [2]. Sex-specific cut-off points were applied when available. Participants whose intake met or exceeded the corresponding reference value were classified as having adequate intake, whereas those below the cut-off point were classified as having inadequate or at-risk intake. The cut-off points used for each mineral are presented in Supplementary Table S2.”Supplementary Table S2 was revised to include the full name of the dietary reference standards used and the corresponding references.

 

3. Comment 3

Reviewer comment: The Discussion section should include information on the prevalence of dietary supplement use and should also consider the contribution of fortified foods as a source of minerals in the diet. The authors should discuss in greater detail whether and how these sources may have influenced total mineral intake.

Response:

We have expanded the Discussion to include available information on dietary supplements and fortified foods. In the sample with available data, only 8 of 286 participants reported dietary supplement use, which represents a low prevalence of approximately 2.8%. However, we did not have information on the type of supplement, dosage, duration of consumption or mineral composition. For this reason, it was not possible to quantify the mineral intake from supplements or determine whether they contained magnesium, calcium, zinc, selenium or other minerals. Given the low number of supplement users, no separate models were performed in this subgroup as they would have provided shaky estimates. Instead, and in line with the response to Reviewer 1, a sensitivity analysis was performed excluding participants who reported supplement use. A specific discussion of fortified foods has also been added, indicating that they could be included in the dietary estimate if they were recorded as consumed foods and appeared in the composition base, but that it was not possible to quantify their contribution separately. This limitation could influence the estimate of total mineral intake and should be considered when interpreting the results.

Changes made in the manuscript:

Discussion: “Information on dietary supplement use was limited. Only 8 of the 286 participants with available information reported using dietary supplements, corresponding to a prevalence of approximately 2.8%. However, data on supplement type, dose, duration, and mineral composition were not available. Therefore, mineral intake from supplements could not be quantified separately, and we could not determine whether these supplements contained magnesium, calcium, zinc, selenium, or other minerals. Fortified foods may also contribute to mineral intake. In the present study, fortified products were captured only insofar as they were recorded as consumed foods and represented in the food composition database used by the EVIDENT tool. However, minerals derived from fortification could not be separated from naturally occurring minerals, which may have introduced some exposure misclassification”

Discussion/Limitations: “Fortified foods may also contribute to mineral intake. In the present study, fortified products were captured only insofar as they were recorded as consumed foods and represented in the food composition database used by the EVIDENT tool. However, minerals derived from fortification could not be separated from naturally occurring minerals. This may have introduced some exposure misclassification and limits our ability to identify the specific contribution of fortified foods to total mineral intake.”

4. Comment 4

Reviewer comment: The authors present interesting results on phosphorus. However, the Discussion should address in greater detail the potential contribution of phosphorus derived from food additives and indicate the main dietary sources that can substantially increase total phosphorus intake.

Response:

We agree with this comment. However, after the updated analyses performed in response to Reviewer 1, phosphorus associations should be interpreted more cautiously, as 1000Kcal-adjusted models and correction for multiple comparisons attenuate the interpretation of nominal findings. For this reason, we have revised the Discussion to avoid a causal or excessively specific interpretation of phosphorus. However, we consider it relevant to discuss dietary sources of phosphorus. We have added that phosphorus comes from both organic sources naturally present in foods such as dairy, meat, fish, eggs, legumes, nuts and cereals, and from inorganic sources added by phosphate additives. These additives can be present in processed meats, deli products, processed cheeses, industrial bakery products, ready meals, fast food, colas, and other ultra-processed foods. We have also indicated that inorganic phosphorus from additives is usually more bioavailable than organic phosphorus from natural foods, so it could substantially increase the total phosphorus load without being fully reflected in overall dietary patterns. However, our dietary record did not allow us to differentiate natural phosphorus from phosphorus from additives, which has been incorporated as a limitation.

Changes made in the manuscript:

Discussion: “Phosphorus intake deserves specific consideration because total dietary phosphorus may originate from both naturally occurring organic phosphorus and inorganic phosphate additives [3,4]. Natural sources include dairy products, meat, fish, eggs, legumes, nuts, and cereals [4,5]. In contrast, phosphate additives are frequently used in processed meats, processed cheese, industrial bakery products, ready-to-eat meals, fast foods, cola-type soft drinks, and other ultra-processed foods [3,4,6]. Inorganic phosphate additives are generally more readily absorbable than naturally occurring organic phosphorus and may therefore increase the total phosphorus load disproportionately [3,4,7]. However, the present dietary assessment did not allow us to distinguish phosphorus naturally present in foods from phosphorus derived from additives. Consequently, the interpretation of phosphorus-related findings should be cautious, particularly after energy adjustment and correction for multiple comparisons.”

5. Comment 5

Reviewer comment: The information provided in the Strengths and Limitations section, especially lines 556–561, is too general and not sufficiently detailed.

Response:

We've rewritten and expanded the Strengths and Limitations section to make it more specific and transparent. In the revised version, the relatively large sample size for a long COVID cohort, the use of a 7-day dietary registry using a previously validated tool, the objective vascular assessment of different arterial territories using cIMT, cfPWV, baPWV and VAI, and the performance of complementary analyses that include adjustment for energy density,  hierarchical models, multicollinearity assessment, FDR correction, constrained cubic splines, and sensitivity analysis excluding supplement users. The section has also been expanded to detail specific limitations: cross-sectional design, inability to infer causality, possible reverse causality, dietary information bias by self-recording, possible underreporting or error in portion sizes, absence of serum or urinary mineral biomarkers, absence of detailed information on supplements, inability to quantify fortified foods and phosphate additives separately,  lack of identification of specific food sources, risk of multiple comparisons and possibility of residual confounding due to clinical, pharmacological, socioeconomic or severity-related variables of long COVID. It is also clarified that, after applying the FDR correction, nominal findings should be interpreted as exploratory and hypothesis-generating.

Changes made in the manuscript:

4.3. Limitations and Strengths

The main strengths of this study include the relatively large and well-characterized sample of adults with LC, the use of a 7-day dietary record collected with the validated EVIDENT tool, and a comprehensive vascular assessment including cIMT, cfPWV, baPWV, and VAI. In addition, the revised analytical strategy incorporated mineral density variables expressed per 1000 kcal/day, hierarchical regression models, multicollinearity diagnostics, FDR correction for multiple testing, restricted cubic spline analyses, formal sex × mineral interaction testing, and a sensitivity analysis excluding supplement users. Together, these elements strengthen the transparency of the findings and support a cautious interpretation of the results.

Several limitations should be acknowledged. First, the cross-sectional design precludes causal inference and does not exclude reverse causality. Second, dietary intake was self-reported through a 7-day dietary record; although this method provides detailed information, underreporting, portion-size error, recall difficulties, and fatigue-related recording errors in individuals with LC may have affected the accuracy of mineral intake estimates. Third, mineral intake was estimated from dietary records and food composition data, but serum, urinary, or other objective biomarkers of mineral status were not available. Fourth, supplement use was recorded only in a limited way, without information on supplement type, dose, duration, or mineral composition. Fifth, fortified foods and phosphate additives could not be quantified separately, and naturally occurring minerals could not be distinguished from minerals added during processing. Sixth, mineral intakes may partly reflect broader dietary patterns or shared food sources rather than isolated mineral-specific effects. Seventh, residual confounding cannot be excluded, particularly due to unmeasured socioeconomic factors, renal function within the non-excluded range, inflammatory biomarkers, hormonal status, medication use, or detailed food-source information. Finally, although the total sample size was adequate for the overall analyses, the study may have been underpowered to detect small effects, particularly in sex-stratified analyses and interaction testing.

6. Comment 6

Reviewer comment: At the end of the manuscript, it is recommended that the authors introduce a separate section entitled: Evidence gaps and future research.

Response:

Following their recommendation we have added a new independent section at the end of the Discussion entitled "Evidence gaps and future research". This section synthesizes the main knowledge gaps identified: the scarcity of studies on mineral intake and vascular health specifically in long COVID, the lack of longitudinal studies that allow establishing temporality, the absence of nutritional intervention trials, the need for objective biomarkers of mineral status, the more accurate evaluation of supplements, fortified foods and additives,   and the need to study food sources and dietary patterns beyond isolated minerals. It has also been suggested that future studies should evaluate whether there are optimal ranges of mineral intake per 1000 Kcal associated with lower arterial stiffness or vascular aging, especially for magnesium, potassium, selenium, phosphorus, and zinc. It has also been pointed out that the analyses by sex should be designed with sufficient power and supported by formal interaction tests, since in the updated analyses the sex-× mineral interactions did not confirm a robust modification of the effect after considering the correction for multiple comparisons.

Changes made in the manuscript:

4.4. Evidence Gaps and Future Research

Evidence on the relationship between dietary mineral intake and vascular health in LC remains scarce. Future longitudinal studies are needed to determine the temporal direction of the observed associations and to clarify whether mineral intake precedes changes in arterial stiffness or vascular aging. Intervention trials would also be useful to assess whether improving overall diet quality or correcting inadequate mineral intake can modify vascular outcomes in this population. Future research should combine detailed dietary assessment with objective biomarkers of mineral status, including serum, urinary, or functional markers when appropriate. Studies should also quantify the contribution of dietary supplements, fortified foods, and phosphate additives, and should distinguish between naturally occurring minerals and minerals added during food processing. Finally, adequately powered studies are required to evaluate sex-specific patterns using formal interaction testing rather than relying only on stratified analyses.

References

1. Authority, E.F.S. Dietary reference values for nutrients summary report; 2397-8325; Wiley Online Library: 2017.
2. Bartrina, J.A.; Majem, L.S. Objetivos nutricionales para la población española: Consenso de la Sociedad Española de Nutrición Comunitaria 2011. Revista española de nutrición comunitaria= Spanish journal of community nutrition 2011, 17, 178-199.
3. Calvo, M.S.; Dunford, E.K.; Uribarri, J. Industrial Use of Phosphate Food Additives: A Mechanism Linking Ultra-Processed Food Intake to Cardiorenal Disease Risk? Nutrients 2023, 15, doi:10.3390/nu15163510.
4. Duong, C.N.; Akinlawon, O.J.; Gung, J.; Noel, S.E.; Bigornia, S.; Flanagan, K.; Pourafshar, S.; Lin, P.H.; Davenport, C.A.; Pendergast, J., et al. Bioavailability of phosphorus and kidney function in the Jackson Heart Study. Am J Clin Nutr 2022, 116, 541-550, doi:10.1093/ajcn/nqac116.
5. Itkonen, S.T.; Lamberg-Allardt, C. Phosphorus - a scoping review for Nordic Nutrition Recommendations 2023. Food Nutr Res 2023, 67, doi:10.29219/fnr.v67.10318.
6. Dunford, E.K.; Calvo, M.S. Phosphate-based additives in processed foods: is excess exposure a cause for concern? A cross-sectional examination of the United States packaged food supply. Am J Clin Nutr 2025, 121, 873-881, doi:10.1016/j.ajcnut.2025.01.009.
7. Volk, C.; Schmidt, B.; Brandsch, C.; Kurze, T.; Schlegelmilch, U.; Grosse, I.; Ulrich, C.; Girndt, M.; Stangl, G.I. Acute Effects of an Inorganic Phosphorus Additive on Mineral Metabolism and Cardiometabolic Risk Factors in Healthy Subjects. J Clin Endocrinol Metab 2022, 107, e852-e864, doi:10.1210/clinem/dgab635.

 

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have substantially strengthened the analytical approach, including energy adjustment, multiple-testing correction, non‑linear analyses, and sensitivity analyses excluding supplement users. The revised results are presented with appropriate caution: no dietary mineral–vascular association remained significant after FDR correction, and the nominal zinc finding was attenuated. The discussion appropriately acknowledges the cross‑sectional design and dietary self‑report limitations, avoiding overinterpretation. The manuscript is well structured, conclusions are balanced, and it now meets Nutrients standards. 

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have addressed the issues raised during the first round of peer review and have made the necessary revisions in line with the reviewer’s comments. I have no further comments and suggestions.