Aims: Iron deficiency has been associated with ADHD and low IQ in epidemiological studies (Froehlich et al., 2011), but this association is likely to be confounded with social class and other demographic variables and health behaviours. Our aim is to use Mendelian randomization to test the hypothesis that iron deficiency is a causal risk factor for ADHD and low IQ. A number of genotypes have shown replicated associations with serum transferrin and/or soluble serum transferrin and related phenotypes (Benyamin et al., 2009; Kamatani et al., 2010; McLaren et al., 2011; Oexle et al., 2011; Pichler et al., 2011) and could be used as instruments for iron status. Although our focus is on ADHD and IQ, we are requesting a broader set of outcome variables that represent related phenotypes (e.g., motor and language abilities, conduct problems).

Hypotheses: We hypothesise that if iron deficiency is a causal risk factor for ADHD and low IQ, then (1) selected gene variants will be associated (in predicted directions) with haemoglobin levels (measured at 8, 12, 18, 31, 43, 61 months and 7 yrs) and/or serum ferritin levels (measured at 8, 12, 18 months); (2) haemoglobin levels (measured at 8, 12, 18, 31, 43, 61 months and 7 years) and serum ferritin levels (measured at 8, 12, 18 months) will be associated with ADHD symptom scores, clinically-signficant levels of ADHD and IQ scores (measured from 7 to 10 yrs); and (3) selected genotypes will be associated with ADHD symptom scores, clinically-significant levels of ADHD, and IQ scores.

Exposure variables: serum ferritin (8, 12, 18 months), haemoglobin (8, 12, 18, 31, 43, 61 months and 7 years), HFE C282Y (rs1800562); TMPRSS6 (rs855791), PCSK7 (rs236918), TF (rs3811647)

Outcome variables: We propose to look at variables in early childhood and later childhood capturing IQ (e.g., WISC, WPPSI), externalizing problem behaviours (e.g., DAWBA ADHD and conduct disorder scores, SDQ scores), attentional problems, motor ability, and language ability.

Confounding variables: We are aware that previous ALSPAC publications have reported associations between demographic variables (e.g., maternal education) and serum ferritin and haemoglobin levels at 18 months (Sherrif et al., 1999). We are requesting the measure of maternal education so as to replicate the association with haemoglobin levels at 7 years (for the purpose of justifying, in part, the need to use MR). We are also requesting gestational age, child sex, child ethnicity, birth weight, childhood weight & height (8, 12, 18, 31, 43, 61 mos; 7 years), recent infection status, child iron supplements (up to 10 yrs), iron in diet (8, 18, 43, 61 mos; 7 & 10 yrs).

Note: We are aware of ALSPAC publications looking at the relationship between Hb and ferritin with neurodevelopmental outcomes in early childhood (e.g., locomotor activity; Sherriff et al., 2001), but not with outcomes in later childhood. If the associations between iron status and later childhood outcomes have already been examined and found to be non-significant, then we would not want to take this application forward.

References

Benyamin, B., McRae, A. E., Zhu, G., Gordon, S., Henders, A. K., Palotie, A. et al. (2009). Variants in TF and HFE explain ~40% of genetic variation in serum-transferrin levels. American Journal of Human Genetics, 84, 60-65.

Froehlich, T. E., Anixt, J. S., Loe, I. M., Chirdkiatgumchai, V., Kuan, L., & Gilman, R. C. (2011). Update on environmental risk factors for attention-deficit/hyperactivity disorder. Current Psychiatry Reports, 13, 333-344.

Kamatani, Y., Matsuda, K., Okada, Y., Kubo, M., Hosono, N., Daigo, Y. et al. (2010). Genome-wide association study of hematological and biochemical traits in a Japanese population. Nature Genetics, 42, 210-U25.

McLaren, C. E., Garner, C. P., Constantine, C. C., McLachlan, S., Vulpe, C. D., Snively, B. M. et al. (2011). Genome-wide association study identifies genetic loci associated with iron deficiency. Plos One, 6.

Oexle, K., Ried, J. S., Hicks, A. A., Tanaka, T., Hayward, C., Bruegel, M. et al. (2011). Novel association to the proprotein convertase PCSK7 gene locus revealed by analysing soluble transferrin receptor (sTfR) levels. Human Molecular Genetics, 20, 1042-1047.

Pichler, I., Minelli, C., Sanna, S., Tanaka, T., Schwienbacher, C., Naitza, S. et al. (2011). Identification of a common variant in the TFR2 gene implicated in the physiological regulation of serum iron levels. Human Molecular Genetics, 20, 1232-1240.