A controlled study of longitudinal IQ changes in females and males with fragile X syndrome.

The aim is this study is to compare the longitudinal changes in IQ scores of females and males with fragile X syndrome and controls and to assess the impact on IQ of molecular variations of the FMR-1 gene in males. Medical records from the child development unit at a university-affiliated children's hospital were retrospectively reviewed. Chart review yielded 35 males with fragile X (19 with a fully methylated full mutation, 9 with a mosaic pattern, and 7 with a partially unmethylated full mutation) 16 females with fragile X and a full mutation, 9 female controls, and 9 male controls who had repeated standardized IQ testing separated by 7 months to 13 years. The differences between the first and last IQ scores from the same IQ test were compared by t tests and subsequently by analysis of variance. Overall, a significant IQ decline was seen in 10/35 (28%) of fragile X males, 0/9 (0%) of control males, 6/16 (36%) of fragile X females, and 1/9 (11%) of control females. The initial t tests and analysis of variance showed a significant difference in IQ (p = 0.02) between fragile X males and control males but did not show a significant difference between males and females with fragile X syndrome or between fragile X and control females. When an analysis of covariance was carried out with the initial IQ as a covariable, a significant difference persisted between fragile X and control males, with a greater IQ decline in fragile X males. There were limitations in using the same IQ test. A comparison among the molecular subgroups of males yielded a significant IQ decline in 3/9 (33%) of mosaic males, 6/19 (32%) of fully methylated full mutation males, and 1/7 (14%) of partially methylated full mutation males. An analysis of covariance using the initial IQ and the intertest interval as covariables demonstrated significant differences between the fragile X molecular subgroups and the controls. Our findings show that a substantial percentage of both male and female fragile X patients and female control patients demonstrated significant IQ decline. There was a significant difference in the IQ change between fragile X and control males. There were no significant differences between fragile X and female controls. There were also significant differences in IQ decline among males with different molecular patterns compared with controls. Males with a mosaic pattern versus control males had the most significant decline of the molecular subtypes. Although the numbers were limited, there was no significant IQ decline in males with less than 50% methylation of the full mutation. This suggests that a small amount of FMR-1 protein production, which is often seen in males with less than 50% methylation, protects against significant IQ decline.