Your Brain Knows if You’re Iron Deficient: Distinct Brain Dynamics in Iron Deficient vs. Sufficient Females in a Visual Category Learning Task

Abstract

ObjectivesThe present study sought to determine the possibility of identifying someone as either iron deficient or sufficient based solely on brain activity, and, if possible, how quickly (based on processing time on a cognitive task) this could be done. MethodsBoth iron sufficient(IS) and iron deficient non-anemic (IDNA) females (mean age 21.1 y) learned two visual categorization tasks while concurrent EEG was acquired. Both tasks involved classifying gray-scale gabor patches on the basis of spatial frequency and orientation; one task used an easily-verbalized rule (rule-based, RB), the other required complex integration of the information (II). Moving windows (20 ms width) of EEG data from 100 electrodes were used to predict the participant’s iron status using logistic regression; model form was determined using stepwise variable selection. The outcome variable was the area under the curve (AUC) of the receiver operating characteristic for the classification. We set a criterion of AUC ≥ 0.80 for successful classification performance. ResultsFor both tasks, successful classification was possible before 200 ms of processing on the basis of fewer than 12 electrodes. Classification in the RB task suggested some early right lateralization in the selection of electrodes, which became more central as processing proceeded. Classification in the II task did not suggest lateralization, although there was some change to more central electrodes as processing proceeded. At each 20 ms time window, for each of the selected set of electrodes, a measure of neural efficiency was calculated as the ratio of the hazard function of the reaction time distribution and the global field power of the selected set of electrodes. This ratio can be interpreted as the amount of work accomplished per unit of energy expended. In all cases, neural efficiency for the IS females exceeded that for the IDNA females, suggesting that the efficiency with which neural energy is expended in cognitive work differs as a function of iron status. ConclusionsIron deficiency without anemia results in distinct patterns of brain activity early in processing that reflect reduced levels of neural efficiency, relative to females who are iron sufficient. Funding SourcesOU Office of the Vice President for Research.