3q29 Deletion syndrome (3q29del) is a rare (~ 1:30,000) (Kendall et al.,
2017; Stefansson et al.,
2014) genomic disorder caused by a typically de novo 1.6 Mb deletion on the long arm of chromosome 3 (hg19, chr3:195725000–197350000) (Ballif et al.,
2008; Glassford et al.,
2016; Willatt et al.,
2005), encompassing 21 protein-coding genes, three antisense transcripts, one long noncoding RNA, and one microRNA. 3q29del is associated with a wide array of neurodevelopmental and neuropsychiatric phenotypes, including a greater than 40-fold increased risk for schizophrenia (SZ) (Kirov et al.,
2012; Marshall et al.,
2017; Mulle, 2015; Mulle et al.,
2010; Szatkiewicz et al.,
2014) and a 19-fold increased risk for autism spectrum disorder (ASD) (Itsara et al.,
2009; Pollak et al.,
2019Sanders et al., 2015). Mild to moderate intellectual disability (ID), global developmental delay (GDD), attention deficit/hyperactivity disorder (ADHD), and anxiety are also commonly associated with 3q29del (Ballif et al.,
2008; Cox & Butler,
2015; Girirajan et al.,
2012; Glassford et al.,
2016; Klaiman et al.,
2022; Sanchez Russo et al.,
2021; Willatt et al.,
2005).
Previous work by our team used standardized direct assessment tools to evaluate individuals with 3q29del for cognitive, neurodevelopmental, neuropsychiatric, and somatic phenotypes (Klaiman et al.,
2022; Murphy et al.,
2018; Sanchez Russo et al.,
2021). Individuals with 3q29del had multiple deficits in cognitive and adaptive functioning, high rates of ID and ASD, and clinically significant deficits in executive function (Klaiman et al.,
2022; Murphy et al.,
2018; Sanchez Russo et al.,
2021). Further, an exceptionally high proportion of individuals with 3q29del (78%) qualified for a clinical diagnosis of graphomotor weakness (Sanchez Russo et al.,
2021). Graphomotor weakness had not previously been associated with 3q29del, and the nuances of this specific functional deficit remain to be explored. Graphomotor difficulties have been documented in genomic disorders with phenotypic similarities to 3q29del, including 22q11.2 deletion syndrome and Williams syndrome (Bellugi,
1988; Bertrand et al.,
1997; Duijff et al.,
2012; Dykens et al.,
2001; Fu et al.,
2015; Heiz & Barisnikov,
2016; Lajiness-O’Neill et al.,
2006; Niklasson & Gillberg,
2010; Pagon et al.,
1987; Pezzini et al.,
1999; Roizen et al.,
2010; Van Aken et al.,
2009; Wang et al.,
1995). It is unclear whether graphomotor weakness is a feature common to most neurodevelopmental disorders, or whether the degree of impairment is specific to 3q29 deletion syndrome only.
Graphomotor skills are the foundational fine motor skills required for more advanced activities like writing and drawing (Suggate et al.,
2018; Ziviani & Wallen,
2006). Graphomotor skills begin to develop in early childhood with scribbles on paper, and gradually progress to more complex skills like legible handwriting (Ziviani & Wallen,
2006). Graphomotor weakness, in turn, is a clinically significant deficit in these foundational skills. In addition to poor handwriting, graphomotor skills are also strongly associated with the development of basic reading skills (Suggate et al.,
2016,
2018) and math skills (Luo et al.,
2007). Graphomotor ability can be assessed in early elementary school-aged children (Volman et al.,
2006). It has been shown that early intervention can substantially improve graphomotor skills in typically developing young children (Ratzon et al.,
2007); while the impact of early intervention on graphomotor skills in populations with mild to moderate ID is unknown, incorporating graphomotor skills training into an early intervention program is unlikely to exacerbate existing deficits. Based on these data and the potentially positive impact of graphomotor skills training, it is critically important to identify graphomotor difficulties early, especially in high-risk populations.
In the present work, we further define the graphomotor phenotype of 3q29del, including possible associations with cognitive, neuropsychiatric, and neurodevelopmental comorbidities within this population. This study provides critical insight into this previously unexplored deficit in 3q29del, with important implications for future studies to assess therapeutic interventions and educational supports. These data ultimately contribute to an improved understanding of this patient population.
Discussion
This is the first study to examine aspects of visual-motor integration phenotypes associated with 3q29 deletion syndrome. In previous work by our team, 78% of 3q29del subjects were found to have clinically significant deficits in overall visual-motor integration as assessed by the Beery-Buktenica Developmental Test of Visual-Motor Integration (Sanchez Russo et al.,
2021). Here, we explored nuances of that deficit, and found substantial deficits in both visual processing and Motor Coordination. Further, we found no effect of age, but significant sex-dependent differences in performance, with males performing worse than females. VMI performance was significantly associated with ADHD diagnosis, where individuals with 3q29del and ADHD scored
higher than individuals with 3q29del and no ADHD. Finally, VMI performance was positively correlated with cognitive and spatial ability, but was not associated with executive function, daily living skills, or school performance. VMI performance in 3q29del is considerably worse than what has been reported in 22q11.2 deletion syndrome, despite the similar overall phenotypic profile in both syndromes. VMI performance in 3q29del is also qualitatively different than the deficits reported in Williams syndrome. Taken together, this analysis reveals a unique profile of VMI deficits in 3q29 deletion syndrome.
The 3q29 deletion is associated with a variety of neuropsychiatric and neurodevelopmental phenotypes, including ASD, ADHD, anxiety, and executive function deficits (Itsara et al.,
2009; Kirov et al.,
2012; Marshall et al.,
2017; Mulle, 2015; Mulle et al.,
2010; Pollak et al.,
2019; Sanchez Russo et al.,
2021; Sanders et al., 2015; Szatkiewicz et al.,
2014). In the general population, idiopathic ADHD and ASD have independently been associated with significantly worse performance on the VMI (Geurts et al.,
2005; Green et al.,
2016; Sutton et al.,
2011). However, in our study sample ADHD was associated with
better VMI performance, with individuals with 3q29del and ADHD scoring significantly higher on the VMI; however, this may be due to the relatively small sample size of the present study. Additionally, ASD and anxiety diagnoses were not associated with VMI performance. We also found that VMI performance was not related to executive function deficits in our study sample. Together, these data indicate that poor visual-motor integration is independent from executive function and attentional deficits within our 3q29del study population. Further studies with larger sample size are needed to better understand the inter- and cross-disorder profiles and if specific syndromic factors play a role in symptom expression, and to confirm any relationship between ADHD and visual-motor integration ability in 3q29del.
Mild to moderate ID is observed in the majority of individuals with 3q29del (Ballif et al.,
2008; Cox & Butler,
2015; Girirajan et al.,
2012; Glassford et al.,
2016; Klaiman et al.,
2022; Sanchez Russo et al.,
2021; Willatt et al.,
2005), and previous studies have shown that low IQ can contribute to poor VMI performance in other genomic disorders (Bertrand et al.,
1997; Duijff et al.,
2012; Lo et al.,
2015; Van Aken et al.,
2009; Wang et al.,
1995). Indeed, composite IQ, verbal IQ, nonverbal IQ, and spatial ability were all positively associated with VMI performance in our study sample. However, none of these variables were perfectly correlated with VMI performance, which suggests that ID alone cannot fully explain the visual-motor integration or motor coordination deficits observed in our study population. Studies of VMI performance in other genomic disorders have found that children with 22q11.2 deletion syndrome or Williams syndrome perform significantly worse than IQ-matched controls (Bertrand et al.,
1997; Van Aken et al.,
2009). Another study compared VMI performance between individuals with Williams syndrome and Down syndrome and found that the individuals with Williams syndrome had a significantly lower mean score (Wang et al.,
1995). These studies indicate that the genomic insults in Williams syndrome and 22q11.2 deletion syndrome impact visual-motor integration ability above and beyond the insult to cognitive ability; the present study suggests the same may be true in the context of the 3q29 deletion. Indeed, previous work by our team has identified a significant association between VMI performance and cerebellar white matter volume (Sefik et al.,
2022), suggesting that VMI performance in 3q29del is related to a brain region distinct from those canonically involved in cognition, but likely to contribute to visuo-spatial function. Further research is needed to fully define the relationship between cognitive ability and visuo-spatial function in individuals with 3q29del; identifying an IQ-matched comparison group would be a particularly informative line of future investigation.
Poor VMI performance has been associated with deficits in fine motor skills and school problems due to difficulties with handwriting and other critical classroom skills (Daly et al.,
2003; Kılıçöz et al.,
2022; Sortor & Kulp,
2003; Weil & Amundson,
1994). We found that individuals with 3q29del scored lowest on the Motor Coordination subtest, with multiple individuals earning the lowest possible standardized score. Case studies of 3q29del have reported some gross motor phenotypes, including gait abnormalities (Cox & Butler,
2015), and previous work by our team identified mild to moderate fine motor challenges associated with the 3q29 deletion (Sanchez Russo et al.,
2021). However, these challenges do not appear to drive poor VMI performance in our study population (Table S1). Additionally, VMI performance was not related to daily living skills measured by the Vineland-3 or school performance measured by the CBCL. It is possible that visual-motor integration deficits in 3q29del are truly not associated with challenges in daily living skills or school performance, but it is also possible that the measures we used did not adequately capture the challenges that individuals with 3q29del experience in these arenas.
We identified a significant sex difference in our 3q29del study population, with males performing worse than females on the VMI, the Visual Perception subtest, and the Motor Coordination subtest. Additionally, consistent with prior findings in the general population (Maccoby & Jacklin,
1972), females with 3q29del have a higher mean verbal IQ compared to males (female 3q29del mean = 86.33 ± 10.25, male 3q29del mean = 75.55 ± 23.19, p = 0.0009; Fig. S7). However, the weak correlation between verbal IQ and VMI performance suggests that this difference is not driving sex differences in VMI performance; rather, sex differences in VMI performance in our 3q29del study subjects appear to be an independent phenomenon. While the mean composite IQ for males with 3q29del is lower than the mean for females, this difference is not statistically significant (female 3q29del mean = 78.75 ± 10.01, male 3q29del mean = 69.60 ± 15.40, p = 0.051). However, due to the strong positive correlation between composite IQ and VMI performance, it is possible that this marginal sex difference in IQ may moderate some of the relationship between VMI performance and sex. Sex differences in VMI performance have been sporadically reported in the literature, both in studies of individuals without a genomic disorder and in studies of individuals with Down syndrome and 22q11.2 deletion syndrome (Coallier et al.,
2014; Duijff et al.,
2012; Lachance & Mazzocco,
2006; Munro et al.,
2012; Niklasson & Gillberg,
2010; Rihtman et al.,
2010). Studies of children in the general population have identified better performance in females relative to males (Coallier et al.,
2014; Lachance & Mazzocco,
2006), and studies of adults in the general population have found the reverse, with males performing significantly better than females (Munro et al.,
2012). Studies of Down syndrome and 22q11.2 deletion syndrome that have examined sex differences in VMI performance have found that females perform significantly better than males (Duijff et al.,
2012; Niklasson & Gillberg,
2010; Rihtman et al.,
2010); however, the samples were overwhelmingly children. It is unclear whether the change in relative performance with increasing age, observed in some general population samples, would also occur in populations with genomic disorders. The sample size in the current study is not large enough to interrogate sex by age interactions; future studies with larger sample sizes will be needed to explore these relationships in the 3q29del population.
There are several aspects of VMI performance that are similar between our 3q29del study sample and reports on other genomic disorders. Visual-motor integration deficits have been identified in individuals with 22q11.2 deletion syndrome (Duijff et al.,
2012; Lajiness-O’Neill et al.,
2006; Niklasson & Gillberg,
2010; Roizen et al.,
2010; Van Aken et al.,
2009), Prader-Willi syndrome (Dykens,
2002; Lo et al.,
2015), and Williams syndrome (Bellugi,
1988; Bertrand et al.,
1997; Dykens et al.,
2001; Fu et al.,
2015; Heiz & Barisnikov,
2016; Pagon et al.,
1987; Pezzini et al.,
1999; Wang et al.,
1995). While the degree of deficit varies between genomic syndromes, there are some qualitative similarities with 3q29del. Similar to our findings in 3q29del, a significant sex difference has been reported in a largely pediatric sample of individuals with 22q11.2 deletion syndrome, with females outperforming males (Duijff et al.,
2012; Niklasson & Gillberg,
2010). In Williams syndrome, published images of VMI response items (Bertrand et al.,
1997; Wang et al.,
1995) bear striking similarity to what we observe in 3q29del. However, we were unable to statistically compare the performance of individuals with Williams syndrome to our data on 3q29del because these studies report raw scores, rather than age-normed standardized scores.
In addition to the qualitative similarities in VMI performance between 3q29del and 22q11.2 deletion syndrome and Williams syndrome, there are also cognitive similarities between these syndromes. 22q11.2 deletion syndrome and Williams syndrome have both been associated with nonverbal learning disabilities (NLDs) (Don et al.,
1999; Goldberg et al.,
1993; Lepach & Petermann,
2011; MacDonald & Roy,
1988; Niklasson et al.,
2001; Rourke,
1995; Rourke et al.,
2002; Shprintzen et al.,
1978; Swillen et al.,
1997,
1999). NLDs are characterized by deficits in visual-spatial-organizational ability, perceptual deficits, psychomotor impairments, and a difference between verbal and nonverbal IQ (Harnadek & Rourke,
1994; Mammarella & Cornoldi,
2014; Myklebust,
1975; Rourke,
1988; Rourke et al.,
2002; Tsur et al.,
1995). Previous work by our team identified a significant performance gap between verbal and nonverbal IQ, with verbal IQ substantially higher than nonverbal (Klaiman et al.,
2022). Taken together with the substantial visual-motor integration deficits we observed in our 3q29del study sample, these data suggest that 3q29del may also qualify as an NLD. NLDs are a unique class of learning disabilities because they are not typically associated with extremely poor school performance (Mammarella & Cornoldi,
2014), and can be difficult to identify given the well-preserved verbal ability in affected individuals. Approaching 3q29 deletion syndrome as an NLD may further help to shape management strategies that directly address the most fundamental deficits of the syndrome in a cohesive way, thus maximizing long-term benefits for this community.
There are notable qualitative similarities in graphomotor function between 3q29del and other genomic syndromes, but these deficits may arise from distinct mechanisms. A study of individuals with Williams syndrome hypothesized that poor VMI performance is due to a failure to plan how to execute a given drawing, particularly with respect to the placement of components from complex geometric forms in relation to one another (Bertrand et al.,
1997). However, we did not find any meaningful association between VMI performance and executive function in our 3q29del study subjects. We also specifically examined the plan/organize from the BRIEF and did not find an association with VMI performance. Mechanisms contributing to poor visual-motor integration in other genomic disorders are not well-described, but it is possible that syndrome-specific cognitive insults are driving the ultimate graphomotor weakness phenotype observed across syndromes.
While this is the first study to assess detailed visual-motor integration phenotypes in 3q29del, it is not without limitations. First, due to our small sample size we were underpowered for some group comparisons. Because of our relatively small sample size, we reported unadjusted p values, which can be susceptible to false positives. However, we note that the magnitude of the main effect of the 3q29 deletion on graphomotor ability is quite large and is unlikely to be a false positive result. Additionally, we were unable to assess the effect of race and ethnicity because our sample is overwhelmingly white and non-Hispanic. Efforts are underway to expand recruitment in historically underserved communities, which will enable us to increase our sample size and diversity in future studies. Second, the standardized score range for the VMI has a lower limit of 45. Several of our study subjects scored a 45 on the VMI, the Visual Perception subtest, and/or the Motor Coordination subtest. Thus, we may not be capturing the true lower extent of the range for all three assessments used in the present study. Studies using the VMI in Williams syndrome typically use raw scores rather than standardized scores for this reason; future analyses of this population may be served by examining raw scores in addition to standardized scores, especially for individuals falling in the lower tail of the distribution.
In summary, this is the first study to examine nuances of visual-motor integration ability in individuals with 3q29del. We find substantial deficits in overall visual-motor integration, as well as in visual perception and motor coordination skills, with the greatest deficits in motor coordination. Notably, while these deficits have some correlation with IQ, they appear to be a unique set of phenotypes rather than simply a derivative measure of cognitive disability. Additionally, poor performance on the VMI is independent from ASD, anxiety, and executive function deficits in this sample. The data presented here, in combination with our previous work, suggest that 3q29del may qualify as an NLD. Deficits in nonverbal skills, including nonverbal IQ and visual-motor integration, have not previously been fully appreciated as fundamental deficits associated with 3q29 deletion syndrome, and may have substantial functional impacts on affected individuals and their families. Additionally, the pervasive nature of visual-motor integration deficits may offer clues about the biological mechanism underlying the pathology in 3q29 deletion syndrome. Future studies should assess whether tailored interventions including occupational therapy to address graphomotor skills may improve outcomes for individuals with 3q29 deletion syndrome.