Conflicting reports appeared on this subject in this particular issue of The Lancet (which by the late 1990s had become a bit notorious for uncritically accepting articles critical of IVF), one from Australia and one from Belgium.

In both studies, Bayley Scales were used to work out a mental development index (MDI) that assesses memory, problem solving and language skills (generally in the predominant language of the country) and a psychomotor development index (PDI) that assesses control of fine and gross muscle function. The average for MDI and PDI is set at 100 (as in an I.Q. test).

In a poorly controlled study of 89 ICSI children from the Royal North Shore Hospital in Sydney, children at one year of age mostly showed normal development (an MDI between 85 and 114), but as a group showed a lower average MDI than 84 routine IVF children and 80 children conceived naturally; the PDI was the same in all three groups; and the chance of a birth defect was the same in all three groups.

The heart of the MDI analysis problem (at best only partly corrected by the authors) was that the 80 naturally conceived children came from a single upmarket obstetric practice in Sydney's richest suburban area; the "routine IVF" patients came mostly from the same part of Sydney; while the ICSI referrals came from all over metropolitan Sydney, with many more low-skilled occupations among the parents and many more patients with English as a second language (if they spoke English at all).

The authors did not assess the intelligence of the parents in any of the groups, so controlling for this critical variable was not carried out. 35% of ICSI children had a father with an unskilled occupation or education of less than 10 years; this parameter was 15% for the IVF children and 15% for the naturally conceived children.

Furthermore, the medical alternative to ICSI -- donor insemination -- is often completely unacceptable among some non-native-English-speaking ethnic groups in Sydney, who are systematically disadvantaged in English-based MDI assessments of their children.

Whereas agreement to take part in the Royal North Shore study was virtually compulsory for the ICSI parents (100% agreed!), just 80% of IVF parents agreed and only 70% of natural conception parents approached took part.

Meanwhile from the world's largest ICSI experience (in Brussels), a study of 201 ICSI children and 131 IVF children at 2 years of age showed no difference in MDI scores.

Although the acceptance rate for MDI assessment in this study was lower in this internationally active clinic, it was for practical reasons of geography, without the systematic bias evident in the Royal North Shore Hospital study.

All the authors agree that further studies will be important ... and who can disagree with that! In the meantime, we need to stick to the rule that powerful medical techniques can carry hazards as well as benefits, so they should only be used when the alternative of not using them is demonstrably worse. The ultimate decision should be in the hands of the consumer.

There's more on birth outcomes after IVF on WebPages 20 and 24 of Overcoming Infertility.

Order the U.S. version of the book from Amazon Books or Barnes and Noble.