Skip to main content
PMC home page
HHS Author Manuscripts logoLink to HHS Author Manuscripts
. Author manuscript; available in PMC: 2024 Mar 18.
Published in final edited form as: Am J Med Genet A. 2021 Dec 8;188(3):858–866. doi: 10.1002/ajmg.a.62594

The association between mosaicism type and cognitive and behavioral functioning among males with fragile X syndrome

Lu Meng 1, Walter E Kaufmann 2, Richard E Frye 3,4, Katherine Ong 1, Jennifer W Kaminski 1, Milen Velinov 5, Elizabeth Berry-Kravis 6
PMCID: PMC10948005  NIHMSID: NIHMS1971040  PMID: 35148024

Abstract

Mosaicism in fragile X syndrome (FXS) refers to two different FMR1 allele variations: size mosaicism represents different numbers of CGG repeats between the two alleles, such that in addition to a full mutation allele there is an allele in the normal or premutation range of CGG repeats, while methylation mosaicism indicates whether a full-mutation allele is fully or partially methylated. The present study explored the association between mosaicism type and cognitive and behavioral functioning in a large sample of males 3 years and older (n = 487) with FXS, participating in the Fragile X Online Registry with Accessible Research Database. Participants with methylation mosaicism were less severely cognitively affected as indicated by a less severe intellectual disability rating, higher intelligence quotient and adaptive behavior score, and lower social impairment score. In contrast, the presence of size mosaicism was not significantly associated with better cognitive and behavioral outcomes than full mutation. Our findings suggest that methylation mosaicism is associated with better cognitive functioning and adaptive behavior and less social impairment. Further research could assess to what extent these cognitive and behavioral differences depend on molecular diagnostic methods and the impact of mosaicism on prognosis of individuals with FXS.

Keywords: fragile X syndrome, methylation mosaicism, size mosaicism

1 ∣. INTRODUCTION

Fragile X syndrome (FXS) is the most prevalent inherited form of intellectual disability and is associated with autism spectrum disorder (Hunter et al., 2014). FXS results from an unstable expansion of a cytosine–guanine–guanine (CGG) nucleotide sequence in the promoter region of the FMR1 gene (Tassanakijpanich et al., 2021). The FMR1 gene product is FMRP, an RNA-binding protein that regulates protein synthesis at synapses (Bagni & Zukin, 2019) and, therefore, plays a critical role in brain development and synaptic plasticity (Martin & Huntsman, 2012). In the vast majority of cases, deficient or absent FMRP is the basis for the FXS phenotype (Bassell & Warren, 2008). FMR1 alleles containing <45 CGG repeats are considered normal. Alleles with CGG repeat expansions in the 55–200 range are termed “premutation” alleles and are associated with FMR1 expression and FMRP synthesis, although mRNA levels may be elevated and FMRP reduced, particularly for larger allele sizes within the premutation range. Premutation alleles are associated with fragile X-associated primary ovarian insufficiency (FXPOI), fragile X-associated tremor/ataxia syndrome (FXTAS), and other less distinctive neurologic phenotypes (Hagerman et al., 2009; Hagerman & Hagerman, 2021). Alleles with CGG repeat expansions of >200 are termed “full mutation” and are associated with atypical FMR1 methylation and the resulting partial or complete silencing of the gene, which leads to decreased or absent FMRP and the clinical features of FXS (Hagerman et al., 2009; Jin & Warren, 2000).

Because of the instability of the CGG repeat during transmission across generations, individuals can be comprised of a mixture of cells in which a proportion have a normal or premutation FMR1 allele and the remaining cells have a full-mutation allele. This situation is termed size mosaicism and, while FMRP levels in individuals with size mosaicism may be higher than in individuals with only full-mutation alleles, FMRP production only occurs in those cells with the normal or premutation allele (Kumari & Usdin, 2020). Furthermore, FMR1 mRNA gain-of-function or repeat-associated non-AUG translation-related toxicity can occur, as seen in FXPOI and FXTAS (Glineburg et al., 2018; Rajaratnam et al., 2017). The process of FMR1 silencing through gene methylation may not occur equally in all cells. This situation is the basis for a second type of FMR1 mosaicism: methylation mosaicism, in which the full-mutation allele escapes methylation and can produce FMRP in a variable proportion of cells (Hagerman et al., 2009; Kumari & Usdin, 2020). Both size and methylation mosaicism can be present, but regardless of mosaicism status, individuals with an FMR1 full mutation will almost universally show a reduction in FMRP level and are categorized as having FXS.

Mosaicism is a source of molecular, systemic, and neurobehavioral phenotypical variability that can be observed in both sexes. The reported mosaicism incidence in males with FXS varies widely from 12% to 41%, most likely reflecting the sensitivity of the method used for detecting mosaicism (Nolin et al., 1994). While size or methylation mosaicism is not always incorporated into the assessment of clinical prognosis due to wide inter-individual variation of the impact of mosaicism on the phenotype, it could have consequences on the overall severity of the disorder in terms of physical impairment and neurobehavioral functioning across a large group of individuals with FXS. Literature findings on the relationship between FMR1 mosaicism and cognitive and behavioral outcomes are mixed. Some publications have reported higher cognitive functioning and more advanced adaptive skills in patients with FXS who have mosaicism compared to nonmosaic FXS (Cohen et al., 1996; Merenstein et al., 1996; Pretto, Yrigollen, et al., 2014; Staley et al., 1993), while other studies did not show differences between males with mosaic and non-mosaic FXS (Backes et al., 2000; de Vries et al., 1993; Harris et al., 2008; Rousseau et al., 1994). Previous studies have also reported the associations of mosaicism type with cognitive and behavioral outcomes. Data consistently show an inverse correlation between methylation level and intellectual functioning, (Basuta et al., 2015; Hagerman et al., 1994; Pandelache et al., 2019; Pretto, Yrigollen, et al., 2014; Wöhrle et al., 1998) with males with FXS having completely or near completely unmethylated full-mutation alleles showing typical intellectual development (Basuta et al., 2015; Hagerman et al., 1994; Wöhrle et al., 1998). On the other hand, studies on size mosaicism have arrived at mixed conclusions. Some studies suggest that the presence of a normal or premutation allele does not compensate for cognitive or behavioral dysfunction associated with FMR1 full mutation (de Vries et al., 1993; Jiraanont et al., 2017), whereas others suggest that males with FXS who have size mosaicism perform better on tests of intelligence than individuals without this type of mosaicism (Baker et al., 2019; Merenstein et al., 1996). These varied findings are likely dependent on the percentage of cells in the brain that are expressing the normal or premutation allele.

Research on a large sample of individuals with FXS could provide a clearer understanding of the impact of FMR1 full-mutation mosaicism on cognitive, behavioral, and other types of functioning. Thus, the purpose of the present investigation is to characterize and compare the impact of size mosaicism and methylation mosaicism on cognitive and behavioral outcomes, including intelligence, adaptive behavior, aberrant behaviors, and autism-related social impairment in males with FXS using clinicians' evaluations and standardized assessments. In addition, the present study also investigated whether males with FXS and mosaicism have better cognitive and behavioral outcomes than those without mosaicism.

2 ∣. METHODS

2.1 ∣. Population and procedures

Data analyzed for this study were from Fragile X Online Registry with Accessible Research Database (FORWARD), a registry and longitudinal database funded by the Centers for Disease Control and Prevention. FORWARD includes standardized clinician- and parent-report forms and standardized caregiver-reported instruments (e.g., Aberrant Behavior Checklist—Community, Social Responsiveness Scale, etc.) submitted by 25 FXS specialty clinics across the United States participating in the Fragile X Clinical and Research Consortium from 2009 through 2018 (Liu et al., 2016). Full details on the creation, enrollment, and data collection for FORWARD are reported in an earlier publication (Sherman et al., 2017).

The analyses for this study utilized baseline data from FORWARD Version 4, obtained from 1471 individuals with FXS (i.e., FMR1 full-mutation allele) evaluated from 2012 through 2019. The FORWARD Version 4 data are currently housed at the Centers for Disease Control and Prevention and are not available for public use. The study was approved by the institutional review board for each participating FXS clinic where data were collected, and written informed consent was obtained from primary caregivers or adult patients who were their own guardians. Due to the focus of this study, only male participants 3 years and older (n = 954) for whom both size and methylation mosaic status were available were included in analyses (n = 487).

2.2 ∣. Measures

2.2.1 ∣. Predictor variables

Size and methylation mosaic status

Size mosaicism was determined with two items: (1) “Is the individual a repeat allele mosaic (e.g., pre/full, intermediate/full, normal/full)?” with answers “0 = no,” “1 = yes,” and “2 = not available;” and (2) “What is the repeat size of the non-full mutation allele?” with a response between 1 and 200. Methylation mosaic status was determined by one clinician-reported item: “What is the methylation status of the full mutation?” with response options “0=fully methylated (without methylation mosaic),” “1 = methylation mosaic,” “2=unspecified abnormal methylation,” and “3 = not available.” Individuals with responses “2” and “3” were omitted from the analysis. Mosaicism in the study was assayed in blood samples, the information for which clinicians gathered from the FMR1 genetic report produced for FXS diagnostic purposes. When the information was not specified in the test report, mosaicism was indicated as “don't know/not available.”

2.2.2 ∣. Outcome variables

Severity of intellectual disability and ASD diagnoses

Clinician-reported severity of intellectual disability (ID) and autism spectrum disorder (ASD) diagnosis, including age at evaluation, were included as outcome variables. Severity of ID was assessed using one item: “Which of these terms best describes the intellectual function of the child currently?” The response options include “1 = no ID,” “2 = borderline ID,” “3 = mild ID,” “4 = moderate ID,” “5 = severe ID,” and “6 = profound ID.” Thus, severity of ID was treated as a continuous outcome variable in analyses. Among 487 participants with reported mosaic status, 428 participants 3 years of age and older with clinician-reported severity of ID were included in the ANCOVA model. ASD diagnosis was performed by the clinician using Diagnostic and Statistical Manual of Mental Disorders, fourth edition, text revision (DSM-IV-TR) or DSM, fifth edition (DSM-V) criteria and reported using one item: “Based on this clinic assessment, does this child currently have a diagnosis of ASD?” with binary answers “0 = no” and “1 = yes.” Among 487 participants with reported mosaic status, 448 participants had ASD diagnosis reported.

Intelligence quotient

Intelligence quotient (IQ) scores were obtained from one of two age-appropriate batteries reported by clinicians: the Stanford-Binet Scale—Fifth Edition (SB5) or the Wechsler Intelligence Scales (Groth-Marnat et al., 2000; Roid & Pomplun, 2012). Of the 487 males with FXS, 172 participants had age of testing and full-scale IQ (FSIQ) scores reported. Based on classifications of ID and distribution of IQ scores, the latter were treated in the analyses as a dichotomous outcome variable with two categories: “0=IQ test scores under 55” and “1=IQ test scores above or equal to 55.” This IQ level corresponds to the cut-off for mild ID (55–69) on the SB5 (Matthews et al., 2015).

Adaptive skills

Adaptive skills were measured by the Vineland Adaptive Behavior Scales—2nd or 3rd Edition (Vineland-II, Vineland-3; here termed Vineland) published by Pearson (Sparrow et al., 1984). The Vineland changed versions in 2015, and our data are a mix of Vineland-II and Vineland-3 depending on when the assessments were done over the 7-year period. Moderate concordance was observed between Vineland-II and Vineland-3 (Farmer et al., 2020). Composite scores and four domain scores were reported by participants' clinicians. Out of the 487 males with FXS, 173 had their Vineland composite scores reported. Age when an informant, typically caregiver, completed the questionnaire was also documented. The Vineland is a semi-structured interview to evaluate adaptive behavior in four domains: communication, socialization, daily living skills, and motor skills. A higher score indicates a higher level of adaptive behavior. The Vineland composite score was treated as a continuous outcome variable in the analyses.

Problem behaviors

The Aberrant Behavior Checklist—Community Edition (ABC-C), a questionnaire published by Slosson, was used to evaluate a wide range of problem behaviors (Aman & Singh, 1986). Out of the 487 males with FXS, 414 participants had a completed ABC-C questionnaire. Age of the participant when an informant, typically a caregiver, completed the questionnaire was also documented. The ABC-C is the most widely used measure of aberrant behavior in individuals with ID, with higher scores indicating more aberrant behavior (Kaat et al., 2014; Schmidt et al., 2013). While applied in its original version in multiple FXS studies, the ABC-C has been adapted for the disorder using a scoring method based on 55 out of the 58 original items (ABCFX) (Sansone et al., 2012). The original ABC-C consists of five subscales: Irritability, Social Withdrawal, Stereotypic Behavior, Hyperactivity/Noncompliance, and Inappropriate Speech. The ABCFX includes an additional subscale: Social Avoidance. Both total and the 6 ABCFX subscale scores were analyzed and were treated as continuous outcome variables in the analysis.

Social skills impairment

Social impairment was measured by the Social Responsiveness Scale—Second Edition (SRS-2) published by Western Psychological Services (Constantino & Gruber, 2012). Out of the 487 males with FXS, 352 had completed SRS-2 questionnaires. Age of the participant when the questionnaire was completed was also documented. The SRS-2 provides a continuous measure of social ability, with higher scores indicating more severe social impairment. The SRS-2 is an ASD screening instrument that includes five subscales: Social Awareness, Social Cognition, Social Communication, Social Motivation, and Restricted Interests and Repetitive Behaviors. SRS-2 total scores are highly correlated with those from the Autism Diagnostic Observation Schedule (ADOS) (Morrier et al., 2017). ADOS is considered a gold standard tool for diagnosing ASD. The present study utilized an SRS-2 scoring method optimized for the FXS population (SRSFX) to provide the highest combination of sensitivity and specificity (Kidd et al., 2020), with a total raw score calculated with 46 out of the 65 original items. The SRSFX total score was treated as a continuous outcome variable in the analysis.

2.3 ∣. Statistical analyses

All statistical analyses were performed using SAS 9.4. In total, data from 487 participants were included in the analyses. Frequencies were calculated for all major study variables. To compare the impact of the two types of FMR1 mosaicism, a series of 2 by 2 (size mosaicism [with vs without repeat allele] by methylation [full vs mosaic]) analyses of covariance were performed on each of the cognitive and behavioral standardized measures, with age at assessment as the covariate. The exact number of participants included in each ANCOVA analysis varied depending on the available data on the cognitive and behavioral measures. To compare the impact of the two types of mosaicism on the level of ID severity (IQ above/below 55) and ASD status (yes/no), logistic regression models controlling for age were developed. A p-value less than 0.05 was considered statistically significant.

3 ∣. RESULTS

Data from 487 males with FXS and mosaicism status who were 3 years and older were included in this analysis. Seventy-six percent of the participants were White, 11% were Hispanic, 8% were African American, 4% were Asian, and 1% were other races/ethnicities. Participants' age ranged from 3 to 60 years with a mean (± SD) age of 13.6 years (±8.30). Approximately 25% of participants are adults and the median age is 12 years. Based on reported genetic testing, 69% (n = 338) had no mosaicism, 11% (n = 52) had size mosaicism only, 12% (n = 57) had methylation mosaicism only, and 8% (n = 40) had both size and methylation mosaicism (Table 1). Table SS1 provides group means by four combinations of the two types of mosaicism.

TABLE 1.

Distribution of mosaicism types among male participants, Fragile X Online Registry with Accessible Research Database, 2012–2019

Size mosaicism
Methylation mosaicism No Yes Total
No 338 (69.4%) 52 (10.7%) 390 (80.1%)
Yes 57 (11.7%) 40 (8.2%) 97 (19.9%)
Total 395 (81.1%) 92 (18.9%) 487 (100.0%)
Open in a new tab

3.1 ∣. Severity of intellectual disability

Most participants were identified as having moderate ID (59.8%), followed by mild ID (24.5%), severe ID (9.8%), borderline ID (3.3%), no ID (2.1%), and profound ID (0.5%). As presented in Table 2, controlling for age, methylation mosaicism had a significant association with ID severity (F = 12.74; df = 1424; p < 0.001). Individuals with methylation mosaicism had significantly less severe ID (N = 87, mean ID severity = 3.41) when compared with individuals who were fully methylated (N = 341, mean ID severity = 3.82). In contrast, there was no association between size mosaicism and severity of ID.

TABLE 2.

Two by two ANCOVA model (controlling for age) analysis results for outcome variable: Clinician reported intellectual disability (ID) severity, Vineland Adaptive Behavior Scales (VABS) composite score, aberrant behavior checklist adapted for fragile X (ABCFX) total score, and Social Responsiveness Scale adapted for Fragile X (SRSFX) total score, Fragile X Online Registry with Accessible Research Database, 2012–2019

No methylation
mosaicism
 With methylation
mosaicism
 p No size mosaicism
 With size mosaicism
 p
N Mean (± SD) N Mean (± SD) N Mean (± SD) N Mean (± SD)
Severity of IDa 341 3.82 (±0.74) 87 3.41 (±0.87) <0.001 344 3.78 (±0.76) 84 3.52 (±0.84) 0.140
VABSb 130 55.19 (±15.60) 43 63.84 (±17.76) 0.002 139 55.88 (±16.21) 34 63.29 (±16.79) 0.481
ABCFXc 330 100.33 (±28.19) 84 96.43 (±28.98) 0.337 334 100.13 (±28.63) 80 97.06 (±27.22) 0.444
SRSFXd 280 68.77 (±20.70) 72 60.85 (±19.01) 0.020 279 68.49 (±20.44) 73 62.04 (±20.49) 0.120
Open in a new tab
a

Severity of ID was categorized based on clinician report as 1 = no ID, 2 = borderline ID, 3 = mild ID, 4 = moderate ID, 5 = severe ID, and 6 = profound ID.

b

VABS measure adaptive skills, with higher scores indicating higher adaptive skills.

c

ABCFX measures 6 problem behaviors: irritability, social withdrawal, stereotypic behavior, hyperactivity/noncompliance, inappropriate speech, and social avoidance, with higher scores indicating more aberrant behavior.

d

SRSFX measures social ability, with higher scores indicating more severe social impairment.

3.2 ∣. Intelligence quotient

IQ scores were from either the Stanford-Binet (N = 131, range 36–79, mean [± SD] score of 45.08 [±7.35]) or the Wechsler Intelligence Scale (N = 41, FSIQ range 34–95, mean [± SD] score of 58.10 [±12.80]). As presented in Table 3, 135 (78%) participants had an IQ score less than 55 (i.e., IQ below mild ID), and 37 (22%) participants had an IQ score above or equal to 55 (i.e., equal or higher level than mild ID). As presented in Table 4, controlling for age and type of IQ test, methylation mosaicism was a significant predictor of having IQ of 55 or above (odds ratio = 3.93 [95% confidence interval:1.48–10.44], p < 0.01). In contrast, size mosaicism was not a significant predictor of IQ.

TABLE 3.

The number of participants in intelligence quotient (IQ) score categories and autism spectrum disorder (ASD)a diagnosis by types of mosaicism, Fragile X Online Registry with Accessible Research Database, 2012–2019

Total
N (%)		 No size mosaic
 With size mosaic
No methylation
mosaic N 		With methylation
mosaic N 		No methylation
mosaic N 		With methylation
mosaic N
IQ test
 IQ lower than 55 135 (78%) 98 17 14 6
 IQ higher than 55 37 (22%) 15 10 6 6
ASD diagnosis
 No ASD diagnosis 223 (50%) 149 28 28 18
 With ASD diagnosis 225 (50%) 165 22 21 17
Open in a new tab
a

ASD diagnosis was documented based on clinician report.

TABLE 4.

Logistic regression model (controlling for age and type of intelligence quotient (IQ) test) analysis results for outcome variable, IQ levela (≥55 vs. <55), Fragile X Online Registry with Accessible Research Database, 2012–2019

Variable Df Odds ratio SE p
Intercept 1 — 0.7597 0.0097
Methylation mosaicism 1 3.928 0.4988 0.0061
Size mosaicism 1 0.465 0.5207 0.1411
Type of IQ testb 1 0.869 0.0549 0.0105
Age (in years) 1 0.082 0.4755 <0.0001
Open in a new tab
a

N = 172 participants with IQ scores were included in the logistic regression model.

b

Two types of IQ tests were included: (1) the Stanford-Binet Scale – Fifth Edition (SB5) and (2) the Wechsler Intelligence Scales.

3.3 ∣. Adaptive behavior skills

Participants' Vineland composite scores ranged from 20 to 92 with a mean (± SD) score of 57.34 (±16.54). As presented in Table 2, controlling for age, methylation mosaicism was significantly associated with Vineland composite score (F = 10.04; df = 1169; p = 0.002). Individuals with methylation mosaicism received significantly higher Vineland composite scores (N = 43, mean [± SD] score of 63.84 [±17.76]) compared with individuals who were fully methylated (N = 130, mean [± SD] score of 55.19 [±15.60]), indicating higher adaptive skills among participants with methylation mosaicism. In contrast, there was no association between size mosaicism and Vineland composite scores.

3.4 ∣. Problem behaviors

ABCFX total scores ranged from 55 to 193 with a mean (± SD) score of 99.54 (±28.36). As presented in Table 2, controlling for age, neither methylation mosaicism (p = 0.337) nor size mosaicism (p = 0.444) was significantly associated with ABCFX total scores (Table 2). Similar analyses for the six ABCFX subscales also showed no significant association with mosaicism status.

3.5 ∣. Social skills impairment

SRSFX scores ranged from 14 to 115 with a mean (± SD) score of 67.15 (±20.59). As presented in Table 2, controlling for age, methylation mosaicism was significantly associated with SRSFX total scores (F = 5.42; df = 1348; p = 0.020). Individuals with methylation mosaicism received significantly lower SRSFX scores (N = 72, mean SRSFX score = 60.85 [±19.01]) compared with individuals who were fully methylated (N = 280, mean [±SD] SRSFX score = 68.77 [±20.70]). In contrast, there was no association between size mosaicism and SRSFX total scores.

3.6 ∣. ASD diagnosis

Among 487 participants with reported mosaic status, 448 participants with information on current ASD diagnosis were included in the logistic regression model. Of these, 225 participants (50.2%) were diagnosed with ASD (Table 3). Table 5 displays the results of the logistic regression. Controlling for age of ASD diagnosis, neither methylation mosaicism nor size mosaicism was a predictor of ASD diagnosis.

TABLE 5.

Logistic regression model (controlling for age) analysis results for outcome variable, autism spectrum disorder (ASD) diagnosisa, Fragile X Online Registry with Accessible Research Database, 2012–2019

Variable Df Odds ratio SE p
Intercept 1 — 0.2815 0.4414
Methylation mosaicism 1 1.170 0.2517 0.5326
Size mosaicism 1 0.825 0.2534 0.4475
Age (in years) 1 0.993 0.0112 0.5197
Open in a new tab
a

ASD diagnosis was documented based on clinician report, and N = 448 participants with ASD diagnosis were included in the logistic regression model.

4 ∣. DISCUSSION

This study assessed the association between two types of mosaicism and cognitive and behavioral outcomes from a large sample of males age 3 years and older with full-mutation FXS from specialty clinics across the United States. We found that methylation mosaicism had a significant positive association with cognitive and behavioral outcomes among males with FXS. Compared to participants without methylation mosaicism, those with methylation mosaicism had less severe intellectual disability, higher mean intelligence test scores, and adaptive behavior scores, and better social skills. In contrast, the presence of size mosaicism was not significantly associated with cognitive and behavioral outcomes. The positive association with methylation mosaicism was limited to a similar range of cognitive functioning (e.g., mild–moderate ID). However, the difference in the cognitive functioning may be sufficient to significantly impact life functioning in some individual cases (Basuta et al., 2015; Hagerman et al., 1994; Wang et al., 1996; Wöhrle et al., 1998).

Although our findings suggest that males with FXS and methylation mosaicism have significantly lower scores on the SRS, methylation mosaicism was not associated with a reduced odds of having an ASD diagnosis. The SRS, even when scoring is modified for FXS (SRSFX), still might not accurately predict ASD in individuals with FXS (Aldridge et al., 2012). In addition, ASD diagnosis in this analysis is a dichotomous variable, and may not be as sensitive to smaller differences in social functioning that are captured in the SRSFX score (Kidd et al., 2020). Baker et al. (2019) found that the presence of detectable FMR1 mRNA was associated with increased features of ASD in their cohort. This is not supported by our results; however, it should be noted that increased FMR1 mRNA could be detected in either size or methylation mosaicism and would not distinguish the two genotypes. Thus, it is difficult to compare the two results. There may also be differences in the two cohorts with respect to the methods for diagnosing ASD that underly the differing findings. Neither methylation mosaicism nor size mosaicism were significantly associated with problem behaviors as measured by the ABC scores. Both ABCFX total scores and six ABCFX sub-domain scores, covering a broad range of behavioral functioning, were not significantly associated with either type of mosaicism. Behavior in FXS may not always correspond to cognitive function and both behavior itself and the rating of the behavior by parents could be more variable than performance-based measures, such as IQ tests. As such, a significant relationship between problem behavior and methylation mosaicism may be more difficult to identify, even though irritable behavior is seen with higher frequency in more severely cognitively impaired individuals with FXS (Eckert et al., 2019). A recent report by the aforementioned research group (Baker et al., 2020) found complex differences in ABC scores between FXS mosaic groups. For instance, increased FMR1 mRNA levels associated with greater irritability in individuals with incompletely silenced full-mutation alleles. These data are difficult to compare with our findings. Nonetheless, it underscores the potential value of follow-up investigations of the relationship between FMR1 mosaicism and behavioral abnormalities that include assessments of FMR1 mRNA levels.

Similar to previous findings, methylation mosaicism was associated with higher cognitive scores when compared to males with a fully methylated full mutation. Positive correlations between methylation status and FMRP levels have been demonstrated previously (De Vries et al., 1996; Pretto, Mendoza-Morales, et al., 2014; Tassone et al., 1999). Methylation is the process by which the FMR1 gene is silenced, although in males (females methylate the gene as a part of the X-inactivation process) methylation usually does not occur unless the expansion to the full mutation has taken place (Hagerman et al., 1994). The unmethylated full-mutation allele can produce FMRP, most likely at lower levels than the normal allele. Thus, even low levels of FMRP may be important for early fetal brain development and ongoing synaptic plasticity and function throughout life (Abitbol et al., 1993; Hinds et al., 1993). Furthermore, perhaps the lack of methylation of an expansion that would typically be methylated may be a sign of a general tendency for incomplete methylation in many cells leading to low levels of FMRP in a significant percent of cells and higher functioning (Wang et al., 1996; Wöhrle et al., 1998). This relationship contrasts with size mosaicism, where the fraction of cells making FMRP will be only those with the premutation or normal allele, which may be a very small percent of cells. These differences may help to explain the finding in our study of a relationship between multiple areas of function with methylation but not size mosaicism. Future research could investigate the biological basis of the two different types of mosaicism and examine whether methylation mosaicism is a signal of a more general cellular effect in all or many cells, and whether there are fundamentally different effects on FMRP between the two types of mosaicism that then have an impact on phenotype.

Published studies have reported mixed conclusions about the functioning of individuals with mosaicism, but none have separated mosaicism by type. Most previous studies did not distinguish size mosaicism and methylation mosaicism, which may explain the variation in findings. In our study, we have 40 participants with both methylation and size mosaicism that showed slightly better performance and functioning than the ones with single type of mosaicism or no mosaicism. However, when separating the effects of each type of mosaicism, only methylation mosaicism was significant. Compared with methylation mosaicism, size mosaic alleles in the premutation or normal range can be easily detected by polymerase chain reaction (PCR) even when they only represent a small percent of the cells, because of the relatively small size and preferential amplification relative to a full mutation (Jiraanont et al., 2017). Thus, individuals with size mosaicism may have very small percentages of FMRP-producing cells, which might explain the similarities in functioning between those with size mosaicism and a fully methylated full mutation observed in this study. Until recently, methylation mosaicism had been mainly detected by Southern blotting, a technique that is not as sensitive as PCR and likely requires a larger percent of cells with mosaicism for its detection (Berry-Kravis et al., 2021). Although methylation PCR has recently become available, this method does not allow determination of incomplete methylation unless at least 10%–20% of alleles are unmethylated, a percent likely much higher than the sensitivity for detection of size mosaic alleles (Aliaga et al., 2016; Filipovic-Sadic et al., 2010).

Although we utilized a large clinic-based sample from a national registry, participation in FORWARD is completely voluntary and only half of participants had complete data and, thus, selection bias is possible, and the sample may not be representative of all individuals with FXS. Similar to previously published studies of mosaicism in FXS, our study was limited by lack of data on the exact percent methylation of full-mutation alleles, percent of size mosaicism, FMRP production, and FMR1 mRNA levels. Potential inaccuracy in clinician interpretation of the DNA report could be a potential limitation; however, FXS DNA reports from standard molecular diagnostic labs are most likely quite accurate. The type of mosaicism is usually specified and, in the case of size mosaicism the allele sizes are specified, so there is not extensive interpretation needed. Also, if the report did not specify methylation or size (premutation/normal allele) mosaicism, the clinicians could just report that mosaicism status was unknown. Furthermore, many of the FXS clinicians at the clinics participating in FORWARD are geneticists or work with a genetic counselor or a geneticist to read the FXS DNA report. So, it is unlikely that inaccuracy in genetic result reporting occurs with sufficient frequency to affect the findings. Future studies with direct molecular analyses could further expand and contribute to an explanation of our findings, since FMR1 mRNA levels cannot be estimated from diagnostic reports of mosaicism. In addition, our study focused exclusively on comparing the two types of mosaicisms separately in the models; future studies could also examine the interaction between mosaicism types and the effect of age on cognitive and behavioral outcomes.

The ASD diagnosis we collected is based on clinician report and there can be variability in assessment and interpretation. Future investigations could also explore how FMR1 mosaicism affects physical/systemic phenotypes of FXS. In addition, the present study used two measures for IQ and two versions of the Vineland scales (Vineland-II and Vineland-3) to maximize sample size for analysis. The IQ tests employed in this study may not be sensitive to the full range of IQ functioning in FXS due to floor effects; thus, z-deviation methods are needed to accurately characterize IQ in FXS (Sansone et al., 2014). Future studies expanding on the biology of mosaicism could evaluate larger cohorts with z-deviation IQ scores on a single measure, such as the SB5, and Vineland scores collected with a single version of the measure. In theory, clinicians (and caregivers) reporting on the outcome measures would not have been necessarily blinded to the methylation/mosaic status of the participants, and this could have influenced their reporting, introducing a potential bias. However, for IQ, the bias is not likely to be significant, as this is a direct performance-based measure, and it is unlikely that the participants with FXS who are being tested on the measure would have the faintest idea what their mosaicism status is or what it means. The psychologists performing the IQ test would not have known the DNA result either. For the Vineland assessment, as this is in most cases a clinician interview, the interviewed families would have not been biased to mosaicism status if the interview was performed correctly. Furthermore, only a small proportion of caregivers knows and understands their methylation status. They may have known the participant was mosaic, but most would not know the difference between size and methylation mosaicism. Thus, it is unlikely that this and other outcomes were significantly biased by the lack of blinding.

In conclusion, methylation mosaicism, but not size mosaicism, seems to be associated with cognitive/adaptive/social functioning in FXS based on multiple measures. This study suggests that methylation mosaicism may be important as a marker for population stratification in clinical studies and a factor in clinical trial design that could impact response to interventions. Determination of mosaicism and methylation status may be important in establishing ways of managing the disorder, for anticipatory guidance in clinic, and future research could assess to what extent these cognitive and behavioral differences due to mosaicism affect the prognosis of individuals with FXS. Thus, this study identifies a key genetic parameter in FXS and it provides directions for future studies with more stringent designs to examine the effect of methylation mosaicism on phenotype and prognosis of individuals with FXS.

Supplementary Material

SUP - Meng - The association between mosaicism type and cognitive and behavioral

Funding information

Centers for Disease Control and Prevention, Grant/Award Numbers: 5U01DD000231, 5U19DD000753-02

Footnotes

CONFLICT OF INTEREST

The authors declare that there is no conflict of interest.

The findings and conclusions in this report are those of the authors and do not necessarily reflect the official position of the Centers for Disease Control and Prevention.

SUPPORTING INFORMATION

Additional supporting information may be found in the online version of the article at the publisher's website.

DATA AVAILABILITY STATEMENT

The analyses for this study utilized baseline data from FORWARD Version 4, obtained from 1471 individuals with FXS (i.e., FMR1 full mutation allele) evaluated from 2012 through 2019. The FORWARD Version 4 data are currently housed at the Centers for Disease Control and Prevention and are not available for public use.

REFERENCES

  1. Abitbol M, Menini C, Delezoide A-L, Rhyner T, Vekemans M, & Mallet J (1993). Nucleus basalis magnocellularis and hippocampus are the major sites of FMR-1 expression in the human fetal brain. Nature Genetics, 4(2), 147–153. [DOI] [PubMed] [Google Scholar]
  2. Aldridge FJ, Gibbs VM, Schmidhofer K, & Williams MJ (2012). Investigating the clinical usefulness of the Social Responsiveness Scale (SRS) in a tertiary level, autism spectrum disorder specific assessment clinic. Journal of Autism and Developmental Disorders, 42(2), 294–300. 10.1007/s10803-011-1242-9 [DOI] [PubMed] [Google Scholar]
  3. Aliaga SM, Slater HR, Francis D, Du Sart D, Li X, Amor DJ, Alliende AM, Santa Maria L, Faundes V, Morales P, Trigo C, Salas I, Curotto B, & Godler DE (2016). Identification of males with cryptic fragile X alleles by methylation-specific quantitative melt analysis. Clinical Chemistry, 62(2), 343–352. 10.1373/clinchem.2015.244681 [DOI] [PubMed] [Google Scholar]
  4. Aman MG, & Singh NN (1986). Aberrant behavior checklist. Slosson. [Google Scholar]
  5. Backes M, Genç B, Schreck J, Doerfler W, Lehmkuhl G, & von Gontard A (2000). Cognitive and behavioral profile of fragile X boys: Correlations to molecular data. American Journal of Medical Genetics, 95(2), 150–156. [PubMed] [Google Scholar]
  6. Bagni C, & Zukin RS (2019). A synaptic perspective of fragile X syndrome and autism spectrum disorders. Neuron, 101(6), 1070–1088. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Baker EK, Arpone M, Aliaga SM, Bretherton L, Kraan CM, Bui M, Slater HR, Ling L, Francis D, Hunter MF, Elliott J, Rogers C, Field M, Cohen J, Cornish K, Santa Maria L, Faundes V, Curotto B, Morales P, … Godler DE (2019). Incomplete silencing of full mutation alleles in males with fragile X syndrome is associated with autistic features. Molecular Autism, 10, 21–21. 10.1186/s13229-019-0271-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Baker EK, Arpone M, Kraan C, Bui M, Rogers C, Field M, Bretherton L, Ling L, Ure A, Cohen J, Hunter MF, Santa María L, Faundes V, Curotto B, Morales P, Trigo C, Salas I, Alliende A, Amor DJ, & Godler DE (2020). FMR1 mRNA from full mutation alleles is associated with ABC-C FX scores in males with fragile X syndrome. Scientific Reports, 10(1), 11701. 10.1038/s41598-020-68465-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Bassell GJ, & Warren ST (2008). Fragile X syndrome: Loss of local mRNA regulation alters synaptic development and function. Neuron, 60(2), 201–214. 10.1016/j.neuron.2008.10.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Basuta K, Schneider A, Gane L, Polussa J, Woodruff B, Pretto D, Hagerman R, & Tassone F (2015). High functioning male with fragile X syndrome and fragile X-associated tremor/ataxia syndrome. American Journal of Medical Genetics. Part A, 167A(9), 2154–2161. 10.1002/ajmg.a.37125 [DOI] [PubMed] [Google Scholar]
  11. Berry-Kravis E, Zhou L, Jackson J, & Tassone F (2021). Diagnostic profile of AmplideX fragile X dx and carrier screen kit for diagnosis and screening of fragile X syndrome and other FMR1-related disorders. Expert Review of Molecular Diagnostics, 21(3), 255–267. 10.1080/14737159.2021.1899812 [DOI] [PubMed] [Google Scholar]
  12. Cohen IL, Nolin SL, Sudhalter V, Ding XH, Dobkin CS, & Brown WT (1996). Mosaicism for the FMR1 gene influences adaptive skills development in fragile X-affected males. American Journal of Medical Genetics, 64(2), 365–369. [DOI] [PubMed] [Google Scholar]
  13. Constantino JN, & Gruber CP (2012). Social responsiveness scale second edition (SRS-2): Manual. Western Psychological Services (WPS). [Google Scholar]
  14. De Vries B, Jansen C, Duits AA, Verheij C, Willemsen R, van Hemel JO, van den Ouweland AM, Niermeijer MF, Oostra BA, & Halley DJ (1996). Variable FMR1 gene methylation of large expansions leads to variable phenotype in three males from one fragile X family. Journal of Medical Genetics, 33(12), 1007–1010. 10.1136/jmg.33.12.1007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. de Vries BB, Wiegers AM, de Graaff E, Verkerk AJ, Van Hemel JO, Halley DJ, Fryns JP, Cufts LM, Niermeijer MF, & Oostra BA (1993). Mental status and fragile X expression in relation to FMR-1 gene mutation. European Journal of Human Genetics, 1(1), 72–79. 10.1159/000472389 [DOI] [PubMed] [Google Scholar]
  16. Eckert EM, Dominick KC, Pedapati EV, Wink LK, Shaffer RC, Andrews H, Choo T-S, Chen C, Kaufmann WE, Tartaglia N, Berry-Kravis EM, & Erickson CA (2019). Pharmacologic interventions for irritability, aggression, agitation and self-injurious behavior in fragile X syndrome: An initial cross-sectional analysis. Journal of Autism and Developmental Disorders, 49(11), 4595–4602. 10.1007/s10803-019-04173-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Farmer C, Adedipe D, Bal V, Chlebowski C, & Thurm A (2020). Concordance of the Vineland adaptive behavior scales, second and third editions. Journal of Intellectual Disability Research, 64(1), 18–26. 10.1111/jir.12691 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Filipovic-Sadic S, Sah S, Chen L, Krosting J, Sekinger E, Zhang W, Hagerman PJ, Stenzel TT, Hadd A, Latham GJ, & Tassone F (2010). A novel FMR1 PCR method that reproducibly amplifies fragile X full mutations in concordance with southern blotting and reliably detects low abundance expanded alleles. Clinical Chemistry, 56(3), 399–408. 10.1373/clinchem.2009.136101 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Glineburg MR, Todd PK, Charlet-Berguerand N, & Sellier C (2018). Repeat-associated non-AUG (RAN) translation and other molecular mechanisms in fragile X tremor ataxia syndrome. Brain Research, 1693(Pt A), 43–54. 10.1016/j.brainres.2018.02.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Groth-Marnat G, Gallagher RE, Hale JB, & Kaplan E (2000). The Weschler Intelligence Scales. In Groth-Marnat G (Ed.), Neuropsychological assessment in clinical practice: A guide to test interpretation and integration (pp. 129–194). John Wiley & Sons, Inc. [Google Scholar]
  21. Hagerman R, & Hagerman P (2021). Fragile X-associated tremor/ataxia syndrome: Pathophysiology and management. Current Opinion in Neurology, 34(4), 541–546. 10.1097/wco.0000000000000954 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Hagerman RJ, Berry-Kravis E, Kaufmann WE, Ono MY, Tartaglia N, Lachiewicz A, Kronk R, Delahunty C, Hessl D, Visootsak J, Picker J, Gane L, & Tranfaglia M (2009). Advances in the treatment of fragile X syndrome. Pediatrics, 123(1), 378–390. 10.1542/peds.2008-0317 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Hagerman RJ, Hull CE, Safanda JF, Carpenter I, Staley LW, O'Connor RA, Seydel C, Mazzocco MMM, Snow K, Thibodeau SN, Kuhl D, Nelson DL, Caskey CT, & Taylor AK (1994). High functioning fragile X males: Demonstration of an unmethylated fully expanded FMR-1 mutation associated with protein expression. American Journal of Medical Genetics, 51(4), 298–308. 10.1002/ajmg.1320510404 [DOI] [PubMed] [Google Scholar]
  24. Harris SW, Hessl D, Goodlin-Jones B, Ferranti J, Bacalman S, Barbato I, Tassone F, Hagerman PJ, Herman H, & Hagerman RJ (2008). Autism profiles of males with fragile X syndrome. American Journal of Mental Retardation, 113(6), 427–438. 10.1352/2008.113:427-438 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Hinds HL, Ashley CT, Sutcliffe JS, Nelson DL, Warren ST, Housman DE, & Schalling M (1993). Tissue specific expression of FMR-1 provides evidence for a functional role in fragile X syndrome. Nature Genetics, 3(1), 36–43. 10.1038/ng0193-36 [DOI] [PubMed] [Google Scholar]
  26. Hunter J, Rivero-Arias O, Angelov A, Kim E, Fotheringham I, & Leal J (2014). Epidemiology of fragile X syndrome: A systematic review and meta-analysis. American Journal of Medical Genetics, Part A, 164A(7), 1648–1658. 10.1002/ajmg.a.36511 [DOI] [PubMed] [Google Scholar]
  27. Jin P, & Warren ST (2000). Understanding the molecular basis of fragile X syndrome. Human Molecular Genetics, 9(6), 901–908. 10.1093/hmg/9.6.901 [DOI] [PubMed] [Google Scholar]
  28. Jiraanont P, Kumar M, Tang H-T, Espinal G, Hagerman PJ, Hagerman RJ, Chutabhakdikul N, & Tassone F (2017). Size and methylation mosaicism in males with fragile X syndrome. Expert Review of Molecular Diagnostics, 17(11), 1023–1032. 10.1080/14737159.2017.1377612 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Kaat AJ, Lecavalier L, & Aman MG (2014). Validity of the aberrant behavior checklist in children with autism spectrum disorder. Journal of Autism and Developmental Disorders, 44(5), 1103–1116. [DOI] [PubMed] [Google Scholar]
  30. Kidd SA, Berry-Kravis E, Choo TH, Chen C, Esler A, Hoffmann A, Andrews HF, & Kaufmann WE (2020). Improving the diagnosis of autism spectrum disorder in fragile X syndrome by adapting the social communication questionnaire and the social responsiveness Scale-2. Journal of Autism and Developmental Disorders, 50(9), 3276–3295. 10.1007/s10803-019-04148-0 [DOI] [PubMed] [Google Scholar]
  31. Kumari D, & Usdin K (2020). Molecular analysis of FMR1 alleles for fragile X syndrome diagnosis and patient stratification. Expert Review of Molecular Diagnostics, 20(4), 363–365. 10.1080/14737159.2020.1729744 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Liu JA, Hagerman RJ, Miller RM, Craft LT, Finucane B, Tartaglia N, Berry-Kravis EM, Sherman SL, Kidd SA, & Cohen J (2016). Clinicians' experiences with the fragile X clinical and research consortium. American Journal of Medical Genetics, Part A., 170(12), 3138–3143. [DOI] [PubMed] [Google Scholar]
  33. Martin BS, & Huntsman MM (2012). Pathological plasticity in fragile X syndrome. Neural Plasticity, 2012, 275630. 10.1155/2012/275630 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Matthews NL, Pollard E, Ober-Reynolds S, Kirwan J, Malligo A, & Smith CJ (2015). Revisiting cognitive and adaptive functioning in children and adolescents with autism spectrum disorder. Journal of Autism and Developmental Disorders, 45(1), 138–156. 10.1007/s10803-014-2200-0 [DOI] [PubMed] [Google Scholar]
  35. Merenstein SA, Sobesky WE, Taylor AK, Riddle JE, Tran HX, & Hagerman RJ (1996). Molecular-clinical correlations in males with an expanded FMR1 mutation. American Journal of Medical Genetics, 64(2), 388–394. [DOI] [PubMed] [Google Scholar]
  36. Morrier MJ, Ousley OY, Caceres-Gamundi GA, Segall MJ, Cubells JF, Young LJ, & Andari E (2017). Brief report: Relationship between ADOS-2, module 4 calibrated severity scores (CSS) and social and non-social standardized assessment measures in adult males with autism spectrum disorder (ASD). Journal of Autism and Developmental Disorders, 47(12), 4018–4024. 10.1007/s10803-017-3293-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Nolin SL, Glicksman A, Houck GE Jr., Brown WT, & Dobkin CS (1994). Mosaicism in fragile X affected males. American Journal of Medical Genetics, 51(4), 509–512. 10.1002/ajmg.1320510444 [DOI] [PubMed] [Google Scholar]
  38. Pandelache A, Baker EK, Aliaga SM, Arpone M, Forbes R, Stark Z, Francis D, & Godler DE (2019). Clinical and molecular differences between 4-year-old monozygous male twins mosaic for normal, premutation and fragile X full mutation alleles. Genes, 10(4), 279. 10.3390/genes10040279 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Pretto D, Yrigollen CM, Tang H-T, Williamson J, Espinal G, Iwahashi CK, Durbin-Johnson B, Hagerman R, Hagerman PJ, & Tassone F (2014). Clinical and molecular implications of mosaicism in FMR1 full mutations. Frontiers in Genetics, 5, 318. 10.3389/fgene.2014.00318 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Pretto DI, Mendoza-Morales G, Lo J, Cao R, Hadd A, Latham GJ, Durbin-Johnson B, Hagerman R, & Tassone F (2014). CGG allele size somatic mosaicism and methylation in FMR1 premutation alleles. Journal of Medical Genetics, 51(5), 309–318. 10.1136/jmedgenet-2013-102021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Rajaratnam A, Shergill J, Salcedo-Arellano M, Saldarriaga W, Duan X, & Hagerman R (2017). Fragile X syndrome and fragile X-associated disorders. F1000Research, 6, 2112. 10.12688/f1000research.11885.1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Roid GH, & Pomplun M (2012). The Stanford-Binet intelligence scales, fifth edition. In Flanagan DP & Harrison PL (Eds.), Contemporary intellectual assessment: Theories, tests, and issues (pp. 249–268). The Guilford Press. [Google Scholar]
  43. Rousseau F, Heitz D, Tarleton J, MacPherson J, Malmgren H, Dahl N, Barnicoat A, Mathew C, Mornet E, & Tejada I (1994). A multicenter study on genotype-phenotype correlations in the fragile X syndrome, using direct diagnosis with probe StB12.3: The first 2,253 cases. American Journal of Human Genetics, 55(2), 225–237. [PMC free article] [PubMed] [Google Scholar]
  44. Sansone SM, Schneider A, Bickel E, Berry-Kravis E, Prescott C, & Hessl D (2014). Improving IQ measurement in intellectual disabilities using true deviation from population norms. Journal of Neurodevelopmental Disorders, 6(1), 16. 10.1186/1866-1955-6-16 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Sansone SM, Widaman KF, Hall SS, Reiss AL, Lightbody A, Kaufmann WE, Berry-Kravis E, Lachiewicz A, Brown EC, & Hessl D (2012). Psychometric study of the aberrant behavior checklist in fragile X syndrome and implications for targeted treatment. Journal of Autism and Developmenal Disorders, 42(7), 1377–1392. 10.1007/s10803-011-1370-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Schmidt JD, Huete JM, Fodstad JC, Chin MD, & Kurtz PF (2013). An evaluation of the aberrant behavior checklist for children under age 5. Research in Developmental Disabilities, 34(4), 1190–1197. 10.1016/j.ridd.2013.01.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Sherman SL, Kidd SA, Riley C, Berry-Kravis E, Andrews HF, Miller RM, Lincoln S, Swanson M, Kaufman WE, & Brown W (2017). FORWARD: A registry and longitudinal clinical database to study fragile X syndrome. Pediatrics, 139(Supplement 3), S183–S193. 10.1542/peds.2016-1159E [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Sparrow S, Balla D, & Cicche H (1984). Vineland adaptive behavior scales-interview edition survey form manual. American Guidance Service, Inc. [Google Scholar]
  49. Staley LW, Hull CE, Mazzocco MMM, Thibodeau SN, Snow K, Wilson VL, Taylor A, McGavran L, Weiner D, Riddle J, & Hagerman RJ (1993). Molecular-clinical correlations in children and adults with fragile X syndrome. American Journal of Diseases of Children, 147(7), 723–726. 10.1001/archpedi.1993.02160310025011. [DOI] [PubMed] [Google Scholar]
  50. Tassanakijpanich N, Hagerman RJ, & Worachotekamjorn J (2021). Fragile X premutation and associated health conditions: A review. Clinical Genetics, 99(6), 751–760. 10.1111/cge.13924 [DOI] [PubMed] [Google Scholar]
  51. Tassone F, Hagerman RJ, Iklé DN, Dyer PN, Lampe M, Willemsen R, Oostra BA, & Taylor AK (1999). FMRP expression as a potential prognostic indicator in fragile X syndrome. American Journal of Medical Genetics, 84(3), 250–261. [PubMed] [Google Scholar]
  52. Wang Z, Taylor AK, & Bridge JA (1996). FMR1 fully expanded mutation with minimal methylation in a high functioning fragile X male. Journal of Medical Genetics, 33(5), 376–378. 10.1136/jmg.33.5.376 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Wöhrle D, Salat U, Gläser D, Mücke J, Meisel-Stosiek M, Schindler D, Vogel W, & Steinbach P (1998). Unusual mutations in high functioning fragile X males: Apparent instability of expanded unmethylated CGG repeats. Journal of Medical Genetics, 35(2), 103–111. 10.1136/jmg.35.2.103 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

SUP - Meng - The association between mosaicism type and cognitive and behavioral
NIHMS1971040-supplement-SUP_-_Meng_-_The_association_between_mosaicism_type_and_cognitive_and_behavioral.docx (15.5KB, docx)

Data Availability Statement

The analyses for this study utilized baseline data from FORWARD Version 4, obtained from 1471 individuals with FXS (i.e., FMR1 full mutation allele) evaluated from 2012 through 2019. The FORWARD Version 4 data are currently housed at the Centers for Disease Control and Prevention and are not available for public use.

RESOURCES