Editor—We read with great interest the article by our colleagues de Vries et al 1on fragile X syndrome, which is generally informative and for the most part accurate. We do, however, take issue with two points raised in the article which should be clarified or corrected.

Firstly, it is stated that “although in female premutation carriers an increased incidence of premature menopause has been reported, a recent study indicated that the premutation does not seem to be a major risk for early menopause”. This would indicate to most readers that there is insufficient or equivocal evidence supporting a link between fragile X premutations and premature ovarian failure. In fact, the authors are making a subtle distinction between “premature menopause” (cessation of menses before the age of 40) and “early menopause” (before the age of 47), neither of which they define. As our paper published in a subsequent issue shows,2 there is a highly significant association between fragile X premutations in women and menopause before the age of 40. In 147 women referred for premature ovarian failure (POF), four out of 25 patients with familial POF and two out of 122 patients with sporadic POF were found to carry a FRAXA premutation. Thus, 2% of all X chromosomes in the POF group bear a premutation, rising to 8% in the familial POF group; by comparison, no premutations were found in 1268 control chromosomes from mothers of boys with learning difficulties.

This evidence is corroborated by as yet unpublished data where fragile X status is the independent variable. A recent collaborative survey3 of 746 women referred for fragile X testing found 23% of women over the age of 40 with a premutation to have recorded premature ovarian failure, whereas no full mutation carriers or non-carriers in this age group had experienced POF (presented at XLMR8 meeting, Picton, 1997). Hence, we feel that premutations must indeed be considered a major risk factor for POF.

The second statement by de Vries et al which we dispute is their assertion that “unstable alleles ranging from 43 to 200 repeat units are regarded as premutation alleles”, a definition which is reiterated pictorially in their fig 4. A true premutation is, we believe, defined as an allele which exhibits meiotic instability in most of its transmissions, and whose ultimate destiny is to become a full mutation. The overwhelming majority of such alleles which we can observe are over 60 repeats in size and even a conservative lower size limit for the premutation would be around 55 repeats. Alleles between 40 and 55 repeats, by contrast, are best defined as “intermediate”, with their stability less certain than those under 40 repeats but still likely to remain unchanged in size over a number of generations. This is far from being merely a pedantic point as many patients with intermediate alleles already misinterpret their risk of having an affected child, believing themselves to have premutations.

This is not to say that rare instances of minor allelic instability at 43 repeats (changes of one or two repeats in a single generation) do not exist, but this is not grounds for classifying these as premutations. Our data from an unbiased survey of families with intermediate alleles showed that of 104 transmissions involving alleles between 41 and 58 repeats, in only three cases was a change in size recorded, and then only by a single repeat.4 Although it may be a plausible assumption that many or most premutation alleles are derived from a pool of intermediate alleles, this has never been observed in practice and we certainly cannot yet ascribe an equivalent risk to all (CGG)_n alleles in this size range. At this level, the internal structure of the repeat in terms of the number of AGG interspersions may be a crucial determinant of stability, as may unidentified elements/genes outside the repeat. Thus, we would argue that the intermediate range or “grey zone” does not, with our current level of understanding, constitute a subset of the “premutation” category as implied by de Vrieset al, but rather justifies consideration in a category of its own.