AUTHORS
Caterina Rosano, MD, MPH, Enrico M Novelli, MD2, Iva Miljkovic, MD, PhD3, Alex Grant4, Ryan Cvejkus, MPH5, Lynda Little-Ihrig6, Kristine Ruppert, PhD7, Yvette Yien, PhD8
AFFILIATIONS
1 University of Pittsburgh, PIttsburgh, PENNSYLVANIA, United States
2 University of Pittsburgh, PIttsburgh, PENNSYLVANIA, United States
3 University of Pittsburgh, PIttsburgh, PENNSYLVANIA, United States
4 University of Pittsburgh, PIttsburgh, PENNSYLVANIA, United States
5 University of Pittsburgh, PIttsburgh, PENNSYLVANIA, United States
6 University of Pittsburgh, PIttsburgh, PENNSYLVANIA, United States
7 University of Pittsburgh, PIttsburgh, PENNSYLVANIA, United States, University of Pittsburgh, PIttsburgh, PENNSYLVANIA, United States
8 University of Pittsburgh, PIttsburgh, PENNSYLVANIA, United States
ABSTRACT
Iron deficiency (ID) is the most common micronutrient deficiency in the world and disproportionately affects women. Worldwide it is estimated that up to 50% of women, primarily from African ancestry, are affected by ID, which often remains untreated throughout the lifespan, resulting in a glaring health care disparity. An association between ID and lower cognition has been shown in Caucasian populations, but data are missing in AA women.
Methods. We tested serum biomarkers of iron metabolism including iron, ferritin, transferrin saturation (TSAT), hepcidin and soluble transferrin receptor (sTfR) in 206 women (mean age 58.9, 82% postmenopausal) from the Tobago Health Study, a population-based, prospective cohort study of AA women residing in Tobago. Data on Digit Symbol Substitution Test (DSST), a test of psychomotor speed, were obtained concurrently with extensive markers of cardiometabolic health at study entry. Five years later, women 60 years and older were invited to return for follow-up exams, and annualized percent change in DSST was computed. Linear regression models estimated the association of TSAT at study entry with DSST cross-sectionally, or longitudinally; models were first adjusted for age, education, body mass index (BMI) and hypertension and further adjusted for menopausal state and diabetes.
Results. At baseline, about 25% of women met criteria for absolute ID (Ferritin < 30ng/ml), and 27% had low TSAT (<20%). Median ferritin was 56 (IQ range 30.0-99.5), mean TSAT was 24.50 (SD:9.03), Among participants with ferritin ≥30 ng/mL, 18% had low TSAT, a pattern consistent with functional ID, that is low bioavailable iron despite relatively high ferritin. Mean hepcidin (30.4 ng/mL, reference range: 0.07-49.4 ng/mL) and sTfR (1.02 ug/mL and reference range: 0.378-1.51 ug/mL) were within the reference range, and the sTfR/log-ferritin index was <2 in 98% of the sample, arguing against inflammation-driven iron restriction as the predominant mechanism in this cohort despite the high prevalence of obesity.
At study entry, this cohort had similar prevalence of hypertension and diabetes to what observed in other Caribbean islands (48% and 17%, respectively); average BMI (31.5kg/m2) and physical activity (24.4 min/week) indicated obesity and sedentary behavior are also highly prevalent. Average DSST was 38 points (SD: 12) at baseline, and declined by 1.6%/year during follow-up. Lower TSAT was significantly associated with lower DSST cross-sectionally, independent of age, education, hypertension and BMI (β: 0.172, SE = 0.08, p = 0.038); associations were independent of menopausal state, but adjustment for diabetes attenuated the results. Among those returning for a follow-up (N=70), the association of TSAT at study entry with annualized percent change in DSST trended toward significance (β: 0.221, SE = 0.32, p = 0.07), and it depended on age (age*TSAT interaction p<0.02); further adjustment for diabetes attenuated the results.
Conclusions/Discussion. In this cohort of African Caribbean women with a high burden of cardiometabolic risk factors, lower TSAT was associated with poorer processing speed cross-sectionally, independent of age, education, hypertension, and BMI, although the association was attenuated by adjustment for diabetes. Similar trends were seen longitudinally in a smaller sample. These initial findings support a link between absolute and functional ID with dementia risk and support further study of iron status as a potentially modifiable contributor to cognitive aging in AA women. Longitudinal studies in this vulnerable population should investigate whether diabetes masks inflammation-driven iron restriction as a mechanism driving these associations.
DISCLOSURES
Caterina Rosano, MD, MPH: No relevant conflicts of interest to declare.