On 2014 Oct 14, GEORGE ANDERSON commented:

The paper that Dr. Carter comments on should have been retracted 9 years ago. The urinary oxytocin (and AVP) data that provide the foundation of the research are fundamentally flawed, with the reported values being approximately one million times higher than prior (and subsequent) reports (see Anderson GM. Report of altered urinary oxytocin and AVP excretion in neglected orphans should be reconsidered. Journal of Autism and Developmental Disorders, 36:829-30, 2006). This was NOT due to a simple error in calculation or a misplaced decimal points. Rather, the analytical method used was non-specific. At the time, the authors defended their results as accurate and even provided some mass spectral data purporting to show that they were indeed measuring correct amounts of urinary oxytocin. In more recent research they have used a more specific method and now obtain results consistent with all other published work. However, they continue to reference this paper without mentioning the huge discrepancy. The paper and any of its conclusions should be disregarded. Just as regretable as the continued presence in the literature and continued citing of this misleading research is the fact that the editiorial board of PNAS has not seen fit to perform the retraction which is a necessary part of self-correcting science. This latter fact reflects very poorly on all papers published in PNAS as there does not seem to be any standard or threshold for retraction; the low quality of review that the paper received in the first place is also not reassuring when considering how much credence to give PNAS papers.

On 2014 Oct 14, GEORGE ANDERSON commented:

The paper that Dr. Carter comments on should have been retracted 9 years ago. The urinary oxytocin (and AVP) data that provide the foundation of the research are fundamentally flawed, with the reported values being approximately one million times higher than prior (and subsequent) reports (see Anderson GM. Report of altered urinary oxytocin and AVP excretion in neglected orphans should be reconsidered. Journal of Autism and Developmental Disorders, 36:829-30, 2006). This was NOT due to a simple error in calculation or a misplaced decimal points. Rather, the analytical method used was non-specific. At the time, the authors defended their results as accurate and even provided some mass spectral data purporting to show that they were indeed measuring correct amounts of urinary oxytocin. In more recent research they have used a more specific method and now obtain results consistent with all other published work. However, they continue to reference this paper without mentioning the huge discrepancy. The paper and any of its conclusions should be disregarded. Just as regretable as the continued presence in the literature and continued citing of this misleading research is the fact that the editiorial board of PNAS has not seen fit to perform the retraction which is a necessary part of self-correcting science. This latter fact reflects very poorly on all papers published in PNAS as there does not seem to be any standard or threshold for retraction; the low quality of review that the paper received in the first place is also not reassuring when considering how much credence to give PNAS papers.