Get started
Work

Response to Comment on “Soluble Receptor for Advanced Glycation End Products and Incident Hypertension in REGARDS”

Year: 2025
Type: article
Abstract: We thank Dr Shen and colleagues for their thoughtful comments on our recent article, “Soluble Receptor for Advanced Glycation End Products (sRAGE) and Incident Hypertension in REGARDS.”1 They raise 3 key concerns: (1) the treatment of missing dietary data and the role of diet as a potential confounder, (2) kidney function and its impact on sRAGE levels, and (3) statin use and its potential influence on sRAGE levels. Regarding missing dietary data, we used the Life’s Simple 7 dietary metric to quantify the quality of diet in the population as poor (0-1 healthy dietary components), intermediate (2-3 components), or ideal (4-5 components). Nationally, diet quality in the United States was previously estimated to be 76% for poor, 24% for intermediate, at <0.5% for ideal.2 In REGARDS, the distribution among participants with available dietary is similar to national data (80.24%, 19.71%, and 0.05% for poor, intermediate, and ideal, respectively). Dietary data are considered missing not at random in REGARDS given differential completion of the Food Frequency Questionnaire based on sociodemographics. Therefore, conventional non-deterministic imputation methodologies (eg, multiple imputation) cannot be used.3 To assess whether assigning a “poor” score to missing diets over adjusted results, we performed two sensitivity analyses, which are shown in Table 1. The first assigned missing dietary scores a value of “intermediate”, the second excluded persons with missing dietary data. The overall associations between sRAGE and risk of incident hypertension remained consistent with our manuscript when performing these sensitivity analyses, suggesting it is unlikely that our primary analysis was mired by excessive over attenuation. Relative risk of incident hypertension by quartile of sRAGE using alternative methods to account for missing dietary data and also adjusting for measures of renal functiona The “minimally adjusted” model included age, sex and SBP. “Fully adjusted” adjusted for age, sex, SBP, level of education, income, geographic region, physical activity, diet, smoking status, alcohol use, waist circumference, BMI, depressive symptoms, perceived stress scale, and diabetes status at visit 1. These were primary results from our article.1 Reproduced with permission from our authorship group, who are the copyright holders. eGFR calculated using the 2021 CKD-Epi Creatinine-cystatin C equation.5 Relative risk of incident hypertension by quartile of sRAGE using alternative methods to account for missing dietary data and also adjusting for measures of renal functiona The “minimally adjusted” model included age, sex and SBP. “Fully adjusted” adjusted for age, sex, SBP, level of education, income, geographic region, physical activity, diet, smoking status, alcohol use, waist circumference, BMI, depressive symptoms, perceived stress scale, and diabetes status at visit 1. These were primary results from our article.1 Reproduced with permission from our authorship group, who are the copyright holders. eGFR calculated using the 2021 CKD-Epi Creatinine-cystatin C equation.5 Additionally, evidence directly linking diet quality to sRAGE remains limited. Moreover, the authors attribute exposure to dietary AGEs as the driving factor influencing sRAGE as it relates to diet quality; however, the Life’s Simple 7 dietary pattern used in our analysis does not necessarily correspond to dietary AGE content. AGEs are formed during high-heat cooking via the Maillard reaction; 4 thus, even a Mediterranean-style (“good”) diet could lead to higher levels of AGEs (and consequently lower levels of sRAGE). Additionally, REGARDS does not measure AGEs directly, and given the limited evidence linking dietary AGEs with sRAGE, adjusting for this factor was not warranted. Nonetheless, this highlights an important gap for future research, linking dietary patterns, ultra-processed food intake, and serum sRAGE levels. The authors also raise concern about kidney function, citing studies of renal dysfunction with lower sRAGE concentrations. In our sensitivity analysis, adding estimated glomerular filtration rate (eGFR) and urinary albumin-to-creatinine ratio resulted in no differences between the fully adjusted model and the model that also included these covariates (Table 1). Finally, the authors note that statins may influence sRAGE levels, as prior studies have reported increases in plasma sRAGE with statin therapy. Given this observation, it is possible that statins might confound the relationship between sRAGE and incident hypertension. We performed a sensitivity analysis that found no meaningful difference in the association between sRAGE and incident hypertension when adjusting for baseline statin use (Table 1). Funding support for this article was provided by the National Institute of Neurological Disorders and Stroke and the National Institute on Aging at the National Institutes of Health (NS041588). The authors report no relevant disclosures. REGARDS data are not publicly available due to ethical and legal restrictions. To abide by its obligations with NIH/NINDS and the Institutional Review Board of the University of Alabama at Birmingham, REGARDS facilitates data sharing through data use agreements. Any investigator is welcome to access the REGARDS data, including statistical code, through this process. Requests for data access may be sent to [email protected].
DOI: 10.1093/ajh/hpaf229
Language: English
Open Access status: bronze
OpenAlex ID: W4416412167
FWCI:
Cites: 5
Cited by:
Related to: