Scientists Can Do Better Including Those With Intellectual Disability in Clinical Trials

In this special issue of American Journal on Intellectual and Developmental Disabilities, all five studies leveraged the Developmental Synaptopathies Consortium (DSC), which follows individuals with three rare genetic conditions associated with neurodevelopmental disorders (GCAND). Although three of the five articles focus exclusively on one condition, Phelan-McDermid Syndrome, (Farmer et al. investigates regression, or loss of skills; Srivastava et al. examines the longitudinal trajectory of adaptive behavior; and Levy et al. examines longitudinal trajectories based on genotypes), the other two focus on aspects of intellectual disability (ID) associated with three conditions—Phelan-McDermid Syndrome, Tuberous Sclerosis Complex, and PTEN hamartoma tumor syndrome to better understand measurement issues in those with a wide range of intellectual function.Rare genetic variants lead to a higher rate of ID and autism (Satterstrom et al., 2020), with both conditions separately and often together presenting along a spectrum. Soorya et al. presents an overarching cognitive and behavioral characterization of all three conditions. The levels of ID range from mild to severe/profound, because recruitment was agnostic to ID and some rare genetic disorders represent more of the severe/profound level of ID versus others. One may argue those with ID are the most vulnerable and the most needing of medications to address challenging behaviors and communication, which families report are among the highest priority (Haneveld et al., 2025). When planning for a clinical trial, among many issues and questions is what should be specifically considered when choosing an outcome measure for individuals with intellectual and developmental disability (IDD) with and without autism spectrum disorder (ASD). What specific approaches should be used in these groups?In the article by Soorya et al., the absolute need to properly collect and account for intellectual ability is pointed out. Presence of ID can influence the expression of other features of a condition, and in many cases, the measures were not developed for use in those with ID. These include some measures of verbal ability, and challenging and repetitive behaviors using standardized measures, and therefore may not be appropriate in studies examining outcomes. They may also not be appropriate at all for understanding impairments in those with autism and ID. The authors did show, of all the traditional measures used, that the Vineland Adaptive Behavior Scales shows the most promise to address multiple domains even in those with severe/profound ID, and should be further considered for clinical trials, acknowledging its limitations.Using a range of ID, researchers were also able to better understand the role of ID in ASD symptom presentation using standard instruments, including those directly addressing methods to better include those with IDD in clinical trials. Individuals with ID are traditionally excluded from clinical trials (Stedman et al., 2019). The data from the DSC which includes those with ID both with and without ASD provides the perfect opportunity to address questions of metrics of IQ scores as well as the impact of ID on traditional measures in clinical trials of essential treatments for ASD and ID.A somewhat unrecognized challenge in the area of outcome assessment is the presence of floor effects in ID, such that it is not possible to gather an accurate quantitative measurement of ID, making understanding of the impacts of ID difficult. Most standardized IQ measures are either age appropriate or utilize age in the denominator to give a standardized score. Some tests like the Mullen Scales of Early Learning are more developmental and may be more appropriate than adult level IQ measures. Mixing and matching and combining scores in trials creates problems with accuracy, and combining the results of different measurements is not ideal because they are not comparable. Farmer et al. discusses why and examines a range of metrics and statistical transformations to address the issues of floor effects. The floor effects include age effects, as mentioned earlier—as someone ages, the denominator of the quotient, age, increases so even small increases in score appear as loss of abilities. As age goes up, IQ goes down, ignoring the increases in raw score. There are also standard score floors. There have been efforts to adjust procedures for intellectual testing, including significantly slowing down the process and providing supports, however, floor effects still exist. Using DSC data, which is longitudinal across years, the article examined the validity of different statistical adjustments to Stanford-Binet 5 scores in those who could complete it. Based on these data, they determined that the approach for longitudinal data that was most resilient to floor effects was the person ability score. Although Dr. Farmer has published on the statistical properties of the person ability score previously (Farmer et al., 2020; Farmer et al., 2023), this study provides empirical longitudinal data that demonstrates the advantage of this approach over other adjustment procedures.Measuring ID is essential, but can be difficult in some individuals, and given the expanded focus on genetic conditions associated with neurodevelopmental disorders, more and more research studies have excluded those with ID over time. This deprives the full spectrum of individuals the opportunity to participate in clinical trials as well as benefit from their findings, but not being included to begin with. The studies in this Special Issue provide important data to better choose the appropriate measures in clinical trials as well as which metrics address floor effects. However, there is much more that needs to be done on multiple fronts: (1) Clinicians need more accessible procedures for measuring ID in community-based settings for those across the ID spectrum; (2) Measures need to be validated in those with ID; and (3) Measures in clinical trials need to be incorporated that address some of the real-world challenges of those with ID, acknowledging that these individuals may have very different experiences than those with average or higher IQs. The Special Issue clearly demonstrates how clinical trials should incorporate ID measures to better understand treatment outcomes.