Next Article in Journal
New Onset Autoimmune Diseases after the Sputnik Vaccine
Next Article in Special Issue
Evaluation of the Effectiveness of an Innovative Polycomponent Formulation on Adult and Aged Human Dermal Fibroblasts
Previous Article in Journal
Expression of Amine Oxidase Proteins in Adrenal Cortical Neoplasm and Pheochromocytoma
Previous Article in Special Issue
Continuous Glucose Monitoring as an Additional Tool in Early Cystic Fibrosis-Related Diabetes Monitoring and in Evaluation of Short-Term Sitagliptin Response
 
 
Article
Peer-Review Record

Serum Urate Levels and Ultrasound Characteristics of Carotid Atherosclerosis across Obesity Phenotypes

Biomedicines 2023, 11(7), 1897; https://doi.org/10.3390/biomedicines11071897
by Daniela Efremova 1,2, Natalia Ciobanu 1,2, Danu Glavan 1, Pavel Leahu 1,2, Renata Racila 1,2, Tatiana Bălănuță 1, Alexandru Matei 1, Maria Vasilieva 1,2, Cristina Cheptea 1, Paula Bîtcă 1, Cristina Damian 1, Ana Bondarciuc 1, Irina Bejenari 1, Adelina Cojocaru 1, Diana Manea 1, Mihail Ciocanu 1, Eremei Zota 1,2, Dumitru Ciolac 1,2,* and Stanislav A. Groppa 1,2,*
Reviewer 1:
Reviewer 2:
Biomedicines 2023, 11(7), 1897; https://doi.org/10.3390/biomedicines11071897
Submission received: 6 June 2023 / Revised: 28 June 2023 / Accepted: 2 July 2023 / Published: 4 July 2023
(This article belongs to the Special Issue Biomedicines: 10th Anniversary)

Round 1

Reviewer 1 Report

This is a nicely conducted analysis taking advantage of large prospective study. 

My comments are minor, but I’m not convinced that the 4 group (non, general, central and mixed) categories are the best way to analyze the data. The general group is so small, I don’t see it that helpful. The 4 categories combine 2 measures, BMI and waist circumference (WC). The following need not go into the manuscript, but I would like greater reassurance that more thorough bivariate explorations (SU vs. BMI and SU vs. WC) were conducted. Is dichotomization justified at the cut points (rather than continuous measures or other cut points). I appreciate the a priori cut points, but I would like this evaluated. I don’t like the very small general category. The authors might consider separate dichotomous (or continuous) variables (rather than 4 group). Then an interaction term could be considered. Again if general obesity in absence of central obesity is so infrequent, the modeling may not be helpful.

Is renal function available? Since that impacts SU and atherosclerosis, that would be important. Similarly is med list available? It would be nice to include diuretics. (More complicated so maybe it wont’ add clarity.)

Minor point, when you define groups by BMI and WC, I do not think you should include p-values (Table 1). I suggest deleting p-values for these rows.

Other minor point. Serum urate (SU) is the preferred term. Please see G-CAN statement on nomenclature. PMID: 31501138

Minor point. I found this wording confusing. No difference? (But there were and you state p<0.05) “No differences in systolic or diastolic BP were attested across 192 obesity types (all post-hoc p < 0.05).”

A few issues with this statement, “the cut-off values adopted in our and previous studies (i.e. >6 in women and >7 mg/dl in men) are based on the satura- 338 tion point of uric acid. Using lower cut-off values may turn more clinically relevant, since 339 existing evidence suggests that even lower levels of SUA have a negative impact on car- 340 diovascular risk [35].”

I understand using pre-specified cut offs, but if you express a concern and posit an alternative, you should look at that possibility. You are able to do that. Secondly men and women do not have different solubility levels, so the first part of the statement is inaccurate. 6 and 7 are based on solubility, as men have higher SU averages, a higher threshold has been selected. (I’m not sure this is well justified for most scientific hypotheses, but it is a standard.) I’m OK with the cut-offs, but state it is convention and that you did some sensitivity (e.g. changed women’s cutoff, looked at lower levels). Don’t just say it is problematic and leave it there. (Critique not meant to sound grumpy, apologies if it comes across that way)

Overall, excellent analysis and write up.

At discretion of authors, a single sentence on SLGT2 inhibitors might add future direction given pleiotropic effects target your variables of interest.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

The manuscript entitled Serum uric acid levels and ultrasound characteristics of carotid  atherosclerosis across obesity phenotypes is an original article. The authors assessed the interrelations between SUA levels and carotid atherosclerosis (IMT and carotid plaque characteristics) in subjects with general, central and mixed (general-central) types of obesity.

In patients with central obesity, higher SUA levels were significantly related to a higher occurrence of increased IMT and were able to detect those with carotid atherosclerosis.

This is non-invasive, simple, interesting clinical study.

The article is very well written. Methodology is well detailed. The results are well presented.

Minor revision

I would suggest adding inclusion and exclusion criteria.

Thiazide and thiazide like diuretics may have as adverse effect hyperuricemia. It is mandatory putting in table 3 the drugs, which could interfere with SUA level.

Table 3: How do you explain significant p for presence of carotid plaques in non-obese patients?

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Back to TopTop