Low Iron Status and Sleep Disturbance in Children with Autism

Background: Providing a well-balanced diet to children with autism can be difficult due to frequently reported restricted eating habits. Many children, and particularly those with autism, do not meet the recommendations for daily iron intake, putting them at risk for iron deficiency (ID) (Herndon, 2009). ID has been linked to sleep disorders such as restless leg syndrome (RLS) and periodic limb movement (PLM) disorder in both children and adults (Gozal, 2009). Children with autism also exhibit higher rates of sleep problems such as bedtime resistance and disrupted sleep patterns (Stores and Wiggs, 2003). 

Objectives: The purpose of this study was to assess ID in children with autism, and to determine if lower iron status is correlated with more sleep problems in children with autism, developmental delays or typical development.

Methods: 75 children (54 with autism (AUT), 10 typically developing (TD), 11 with non-autism developmental delay (DD) (mean age=4.5y and mean BMI=16.3), underwent both an overnight electroencephalogram (EEG) recording with electro-oculogram, electrocardiogram, and surface chin and anterior tibialis EMG, and blood sample collection for complete iron studies.  A caregiver completed a Children’s Sleep Habits Questionnaire (CSHQ) at the time of the overnight study.  Blood samples taken within ±3 months of the EEG were frozen and later analyzed for serum ferritin and total iron binding capacity (serum iron, transferrin and %transferrin saturation).  A ferritin level below the mean for the study population (<20 ng/ml) or the presence of two or more other low iron indices classified a participant as having “low iron” within the context of this study.

Results: Serum ferritin average over all groups was 20.5 ng/ml ±10.1; there were no significant group differences for any iron measures. There were no significant group or iron status differences for CSHQ subscales - bedtime resistance or sleep disordered breathing.  Fifty one percent of AUT parents reported  too little sleep on the CSHQ, compared to 21% of TD and DD parents (p<0.05).  Parent perceptions were supported by the EEG data; the AUT group had significantly less total sleep time (456±99 min) than either TD (529±56 min) or DD (545±54 min) groups (p<0.01). Overall, serum ferritin was positively correlated with sleep restlessness on the CSHQ (0.277, p<0.05) and within the AUT group this relationship was sustained at a trend level (p<0.07).  Using a categorical approach overall, there was a trend for those with low iron status to have a higher PLM index (p<0.12) but no within-group differences.

Conclusions: In our sample, the iron status of AUT children was similar to TD and DD children; all had serum ferritin levels in the low normal range. An overall low level and narrow range of iron status values across groups may have limited our ability to detect true group differences in iron status or the relationship to other sleep indices. The relationship of iron status to sleep disorders such as RLS and PLMD deserves further study to determine if this interaction may be different in children with autism.