Crespi & Badcock (C&B) are to be congratulated for an outstanding theoretical contribution that may constitute a breakthrough regarding our understanding of autism-spectrum versus schizophrenic-spectrum disorders. Although their approach is theoretically guided, it is also data driven. That is, although supposed imprinted genes accounting for these disorders have yet to be pinpointed, C&B amass a large body of evidence (see their target article's Table 1) showing that the disorder phenotypes are diametrically opposed. Therefore, it is legitimate and important to criticize and refute, or to augment and supplement, any constituents of their evidence put forward.
Inter alia, C&B review some digit ratio (2D:4D) research findings consistent with their proposition (see sect. 7 on “Sex differences”). No less than four references (Arató et al. Reference Arató, Frecska, Beck, An and Kiss2004; Manning et al. Reference Manning, Baron-Cohen, Wheelwright and Sanders2001; Milne et al. Reference Milne, White, Campbell, Swettenham, Hansen and Ramus2006; Walder et al. Reference Walder, Andersson, McMillan, Breedlove and Walker2006a) from this brief passage are listed under their key references, so this clearly has relevance. In what follows, I augment and supplement further evidence from 2D:4D research consistent with the C&B hypothesis.
2D:4D is the length ratio of the second digit (index finger, 2D) to the fourth digit (ring finger, 4D). This trait is sexually differentiated (men have lower 2D:4D than women) and may be a marker for the organizational (permanent) effects of prenatal androgen action (Manning et al. Reference Manning, Scutt, Wilson and Lewis-Jones1998). C&B have missed some important further 2D:4D research references. This is fully excusable, but the gaps in the cited literature should be filled. As of this writing (early 2008), the 2D:4D literature numbers about 230 journal publications and about 50 unpublished theses, and is therefore overviewed only by specialists. Owing to the fast research progress, complete and topical reviews of 2D:4D research are currently unavailable (for partial reviews, see Cohen-Bendahan et al. Reference Cohen-Bendahan, van de Beek and Berenbaum2005; Manning Reference Manning2002; Peters et al. Reference Peters, Mackenzie and Bryden2002; Putz et al. Reference Putz, Gaulin, Sporter and McBurney2004; Voracek & Dressler Reference Voracek and Dressler2007; Voracek & Offenmüller Reference Voracek and Offenmüller2007; Voracek & Pavlovic Reference Voracek and Pavlovic2007; Voracek et al. Reference Voracek, Reimer, Ertl and Dressler2006; Reference Voracek, Dressler and Manning2007).
Ten studies (C&B cite two: Manning et al. Reference Manning, Baron-Cohen, Wheelwright and Sanders2001; Milne et al. Reference Milne, White, Campbell, Swettenham, Hansen and Ramus2006) are informative for 2D:4D in autism-spectrum disorders. Manning et al. Reference Manning, Baron-Cohen, Wheelwright and Sanders(2001) found that children with autism or high-functioning autism (Asperger syndrome) and their first-degree relatives (unaffected siblings, mothers, and fathers), had markedly lower (hypermasculinized) 2D:4D than healthy population controls. This basic finding replicated in a study of autistic patients from Japan (Osawa et al. Reference Osawa, Sugie, Fukuda, Ito, Sugie and Ohzeki2005), was also replicated by Milne et al. Reference Milne, White, Campbell, Swettenham, Hansen and Ramus(2006) and de Bruin et al. Reference de Bruin, Verheij, Wiegman and Ferdinand(2006) for boys with autism, but only partially replicated by Falter et al. (in press) for boys with autism, who had somewhat, but not significantly, lower 2D:4D than normals. Knickmeyer et al. Reference Knickmeyer, Wheelwright, Hoekstra and Baron-Cohen(2006) mention data indicating that girls with autism as well show masculinized (low) 2D:4D. Relatedly, three further papers have reported associations of low 2D:4D with various attention-deficit/hyperactivity disorder (ADHD) subtypes in children (Martel et al., in press; McFadden et al. Reference McFadden, Westhafer, Pasanen, Carlson and Tucker2005; Stevenson et al. Reference Stevenson, Everson, Williams, Hipskind, Grimes and Mahoney2007). Of further note, Voracek and Dressler Reference Voracek and Dressler(2006) have found no correlation of 2D:4D with autistic-like personality traits and other relevant dimensions (empathizing and systemizing) in a large healthy general population sample.
Five studies (C&B cite two: Arató et al. Reference Arató, Frecska, Beck, An and Kiss2004; Walder et al. Reference Walder, Andersson, McMillan, Breedlove and Walker2006a) are informative for 2D:4D in schizophrenic-spectrum disorders. Arató et al. Reference Arató, Frecska, Beck, An and Kiss(2004) focused on distal finger-extent patterns (2D and 4D, both relative to 3D) rather than on 2D:4D and ascertained a smaller distance from 2D tip to 3D tip and a larger distance from 4D tip to 3D tip (i.e., relatively longer 2D and relatively shorter 4D, both broadly equivalent to a high, feminized 2D:4D) in both male and female schizophrenic patients, compared with normals. Walder et al. Reference Walder, Andersson, McMillan, Breedlove and Walker(2006a), using a similar method, reported feminized distal finger extent in male (but not female) adolescents with schizotypal personality disorders, compared with healthy controls. The results of another study (Procopio et al. Reference Procopio, Davies and Marriott2006) are difficult to interpret, as an uncommon measure (height [h] divided through 2D or 4D) was used, which evidently was confounded by patient-control differences in height. Female (but not male) schizophrenic patients had significantly higher h:2D than controls. Weißegger Reference Weißegger(2004) found markedly higher 2D:4D in female schizophrenic patients than in controls. However, the effect was, due to sample size limitations, statistically not significant, though noticeably smaller in males; and a follow-up study (Baumhackl Reference Baumhackl2007) failed to replicate any 2D:4D differences between schizophrenic patients and normals.
On the whole, the literature suggests 2D:4D to be low (masculinized) autism-spectrum disorders and high (feminized) in schizophrenic-spectrum disorders. An earlier conjecture (Manning Reference Manning2002, p. 143), namely that 2D:4D in schizophrenia, as in autism, could be low (male-typed), appears untenable. The evidence is not without inconsistencies and replication failures, and appears stronger and more unequivocal for autism-spectrum than for schizophrenic-spectrum disorders. Whereas associations of 2D:4D and autistic-like traits apparently do not generalize to the normal population (Voracek & Dressler Reference Voracek and Dressler2006), it remains to be seen whether associations of 2D:4D and schizotypal traits do. Additional 2D:4D data pertaining to both autism and schizophrenia would certainly benefit tests of the C&B hypothesis.
Also worth consideration is the substantial heritability (h 2=.60 to .80) of 2D:4D (twin studies: Gobrogge et al. Reference Gobrogge, Breedlove and Klump2008; Paul et al. Reference Paul, Kato, Cherkas, Andrew and Spector2006; Voracek & Dressler Reference Voracek and Dressler2007; family studies: Dressler et al. Reference Dressler, Voracek and Manning2004; Marshall Reference Marshall2000; Ramesh & Murty Reference Ramesh and Murty1977), along with evidence for assortative mating on the trait (spousal correlations: about r=.20; Voracek et al. Reference Voracek, Dressler and Manning2007). Further, 2D:4D is a correlate of androgen sensitivity, as shown through correlations with functional polymorphisms of the androgen receptor gene, located on the X chromosome, in men (Manning et al. Reference Manning, Bundred, Newton and Flanagan2003). Surprisingly, 2D:4D family studies findings seem more consistent with hypothetical Y-linked inheritance (i.e., strong father-son and brother-brother correlations) than with X-linked inheritance, but identification of Y-linked inheritance is known to be elusive. All of this points to possible direct sex chromosome and imprinted gene effects on 2D:4D expression. If so, this in turn would give clearer support of the C&B hypothesis, as far as 2D:4D is concerned. It may also shed more light on further intriguing differences (e.g., age of onset, sex ratio of prevalence) between autism-spectrum and schizophrenic-spectrum disorders.
