The authors respond: We thank Kawada1 for his interest in our study2 on cadmium exposure and incident cardiovascular disease (CVD). We evaluated the association of urine cadmium concentration with cardiovascular disease incidence and mortality in the Strong Heart Study, a large population-based cohort with 20 years of follow-up. Rather than evaluate the predictive ability of urine cadmium for incident cardiovascular disease, we aimed to assess the independent association of cadmium as a potential cardiovascular risk factor. In our article, we did not report the association of cardiovascular disease with sex, increased low-density lipoprotein cholesterol, decreased high-density lipoprotein cholesterol, smoking, hypertension, decreased kidney function, and diabetes because these associations have been reported previously.3,4 To respond to Dr. Kawada’s request, we provide a Table showing the association between these established risk factors and incident CVD, with and without adjustment for urine cadmium.TABLE: Association of Traditional Cardiovascular Risk Factors with Cardiovascular Disease Incidence in the Strong Heart Study after 20 Years of Follow-up (n = 3,348)Second, we agree with Dr. Kawada that it is important to report associations between cadmium and CVD stratified by sex. The evaluation of effect modification by sex is important because there are well-established differences in cadmium levels and in CVD incidence by sex, and some epidemiologic studies have found differences in health outcomes by sex.5 In our study, as shown in the eAppendix figures, the magnitude of associations in sex-specific strata were relatively similar for men and women, and the P values for interaction were not statistically significant except for stroke. Consequently, we decided to present the main results in the whole study population (rather than separately for men and women) because it is a more powerful approach. Regarding Dr. Kawada’s comment on our original Table, the overall prevalence for the different determinants cannot be indirectly inferred based on the comparison of their respective associations with urine cadmium. For instance, the prevalence of current smoking in women (30%) was lower compared with men (42%), contrary to what Dr. Kawada anticipated. In our study, smoking was also related to increased urine cadmium levels in women. Finally, we do not have data on N-acetyl-beta-D-glucosaminidase and could not evaluate its association with cadmium in our population. However, we adjusted for estimated glomerular filtration rate, indicating that cadmium is a cardiovascular risk factor independent of measures of renal function. Current cadmium standards are based on kidney markers. Our findings support the idea that cardiovascular disease should be incorporated in risk assessment analyses for establishing cadmium safety standards. Maria Tellez-Plaza Departments of Epidemiology and Environmental Health Sciences Johns Hopkins University Baltimore, MD Area of Epidemiology and Population Genetics National Center for Cardiovascular Research (CNIC) Madrid, Spain Fundacion de Investigacion del Hospital Clinico de Valencia-INCLIVA Valencia, Spain [email protected] Eliseo Guallar Department of Epidemiology Johns Hopkins University Baltimore, MD Area of Epidemiology and Population Genetics National Center for Cardiovascular Research (CNIC) Madrid, Spain Barbara V. Howard MedStar Health Research Institute Hyattsville, MD Georgetown-Howard Universities Center for Clinical and Translational Science Washington, DC Ana Navas-Acien Departments of Epidemiology and Environmental Health Sciences Johns Hopkins University Baltimore, MD
Search from 250 M+ research papersSearch from 250 M+ research papers
Open Access
Jul 1, 2021
Thalia Ponce Dimas 
