Over 200,000 people with Down syndrome (DS) live in the United States [1]. Adults with DS have an increased risk for developing dementia, with a lifetime dementia risk in excess of 95%, and median age of onset of 55 years [2]. The increased risk for co-occurring Alzheimer’s disease (AD) in adults with DS [3] is thought to result from the triplication of the amyloid precursor protein (APP) gene located on chromosome 21 [4], leading to an overabundance of amyloid plaque deposition in the brain, a main characteristic of AD, eventually leading to associated morbidity and mortality [5].

The diagnosis of AD in the general population is often made based on a combination of clinical history, symptoms, cognitive assessment, imaging, biomarker testing, and post-mortem brain tissue analysis. In people with DS, many of these approach have limitations. While biomarkers show utility, use of standardized rapid cognitive tools typically used in primary care for reliable in-office cognitive assessment are not appropriate and have limited value. Symptom profiles can vary from those in the general population. In certain population subsets of individuals with DS including those with severe intellectual disability, autism spectrum disorder, and sensitivity challenges, brain imaging can be challenging to obtain, and abnormal findings on imaging are often present prior to the onset of clinical symptomatology [2, 3]. Comprehensive cognitive evaluation has been proposed as an alternative or complementary method to evaluate for co-occurring AD with tests such as a modified version of the Selective Reminding Test, the Modified Mini Mental Status Evaluation—Down Syndrome (MMMSE-DS), the Down Syndrome Mental Status Examination (DSMSE), or The Test for Severe Impairment (TSI), among others [6]. To triage those patients who should proceed to cognitive evaluation, imaging, and biomarker evaluation, a useful approach would be to screen adults with DS for signs and symptoms of memory changes associated with AD. While real-time, rapid screening for AD in DS would therefore offer important benefits, identifying a practical and useful screening tool in DS has many challenges [7, 8]. Few instruments have been validated in DS, questions should focus on functional changes, and questionnaires should be caregiver-reported as individuals with DS require lifelong caregivers who may have unique exposure to recollection bias not present in caregivers of elderly adults with AD in the general population who may be more recent in a caregiver role.

Acknowledging the importance of AD screening, the most recent evidence-based clinical practice guidelines for DS made a statement of good practice: a strong recommendation to screen for Alzheimer-type dementia annually starting at age 40 years [9]. The evidence-based clinical practice guidelines for DS specifically state that “Decline in 6 domains specified by the National Task Group–Early Detection Screen for Dementia (NTG-EDSD) should be used to identify early-stage age-related Alzheimer-type dementia.” [9].

To address this new recommendation on universal screening for AD in DS, our Down syndrome clinic began a quality improvement initiative to study the use of a newly instituted dementia screening protocol in patients with DS, 40 years and older. Although the evidence-based clinical practice guideline recommends utilizing the NTG-EDSD, we factored several additional considerations into the instrument selection process, given our experience working with caregivers of and individuals with DS, the clinical visit process and flow, and our team’s collective prior experience in diagnosing and caring for DS patients with AD. Considerations included, questionnaire length and ease for families, caregiver accessibility, instrument validation and validation population, ability to repeat screening over time to assess for change, the ability to produce a score to guide clinical care by indicating the presence or absence of signs of dementia, available in English, and publicly available for use free of charge. After searching and reviewing existing instruments to screen for Alzheimer’s disease dementia based on the above criteria [7, 8, 10, 11, 12, 13, 14, 15], we ultimately identified four validated screening tools for AD in DS that generally met the above criteria, each with strengths, limitations, and varied published use in individuals in DS [16]. The four identified screens for dementia included, the Early Detection and Screen for Dementia (the NTG-EDSD), developed by The National Task Group on Intellectual Disabilities and Dementia Practices (NTG) [17], the Adaptive Behaviour Dementia Questionnaire (ABDQ) [18], the Dementia Questionnaire for People with Learning Disabilities (DLD) [19], and the American Association on Mental Deficiency Adaptive Behavior Scale, Part I [20].

With this paper, we present the results of our newly implemented clinical dementia screening protocol which included several aims: (1) to describe our experience selecting a screening instrument for AD in DS, (2) to summarize the scoring results, and (3) to describe our attempts to pilot a modified screening instrument. To our knowledge, no study has yet to assess the implementation process or implications of a universal dementia screening recommendation for adults with DS ages 40 and older. Given the screening recommendations, this report provides the first description and results of implementing a universal AD screening protocol in adults with DS utilizing a validated instrument, thereby offering useful guidance for clinicians caring for individuals with DS seeking to implement the recommended DS clinical care guideline.

Based on recent evidence-based care guidance [9], we prospectively implemented a new clinical protocol to screen for AD in adults with DS. We selected the ABDQ screening tool, comparing ABDQ scores to the contemporaneous clinical impression of AD, and found the original ABDQ to have high specificity, but low sensitivity, and with a high negative predictive value for identifying dementia in adults with DS. Based on these properties and feedback from caregivers, we piloted a modified ABDQ with minor modifications to the instructions and response options, which did not result in significant improvements in psychometric properties.

In our clinic cohort, neither the original nor the modified ABDQ survey functioned well as a screening tool. For one patient with an original ABDQ score indicative of dementia, this did not correlate with the clinical impressions of the treating physicians or the social workers. In 9 patients in whom AD was suspected clinically, the original ABDQ was negative: giving low sensitivity and low positive predictive value. Our findings are in conflict with the strong psychometric properties published in the original validation of the ABDQ [18]. Prasher et al. [18] found the ABDQ to have “good reliability and validity, with an overall accuracy of 92%” in a cohort of caregivers of individuals with DS. However, many notable differences between our study and the original validation study by Prasher et al. [18] exist. Our study used the ABDQ in clinical practice as a clinical screener while Prasher et al. [18] studied the ABDQ in research validation. Although we obtained ABDQ results on most of our eligible patients, our cohort is smaller than the original validation sample of 150 adults. Additionally, for ease of administration prior to a clinic visit, we distributed the ABDQ by e-mail link with the questionnaire administered in web-based format through REDCap while the original validation used paper-and-pencil surveys.

To assess the 9 false negatives, and better understand the reasoning behind caregiver answers, the clinical team piloted the use of a modified ABDQ, understanding that a modified instrument would not yield validated scores. In the modified ABDQ, we changed some of the wording and then shared both the original and modified ABDQ text with caregivers. After modifications, based on caregiver feedback, we found the modified questionnaire to again have high specificity, but still have low sensitivity and low positive predictive value, suggesting that our modifications did not substantially improve the ability of the questionnaire to screen for AD in adults with DS. This lack of improvement might result from our modifications not having successfully addressed the discordance between clinical and family perspectives. Additionally, many caregivers in the modified ABDQ group were the same caregivers who had previously completed the original ABDQ, resulting in potential recall bias and memory effects – where respondents learn from completing a survey twice within a short window of time.

While the ABDQ questions focuses on adaptive function, we found in the 9 false negative patients that in addition to changes in activities of daily living, many of the early signs of dementia presented as change in sleep, memory problems, irritability, and objects misplacement. As the ABQD does not include questions assessing changes in memory or behavior, it may be of limited utility for clinical dementia screening in DS. We also found that scores differed between patients with and without clinical findings of dementia, with median scores higher, though still not reaching the recommended cut-off value, when clinical dementia was present, suggesting that establishing a lower cut-off value may be of clinical benefit.

Our study has several limitations worth highlighting which may explain the study’s negative findings. The study’s small sample size and generalizability may impact external validity, while potential for bias including social desirability, recall bias, and scoring bias may impact internal validity. Our initial sample was small, in line with our goal of piloting the ABDQ as a screening tool. In the future, we could re-study the ABDQ, or other screeners for AD, in a larger, multi-site study of adults with DS based on a larger sample size determined by pre-study power requirement calculations. We utilized the ABDQ as a one-time screening tool as originally validated in patients with DS, however, the instrument may provide greater ability to detect AD with longitudinal use, such as repeated annual screenings, which could be tested as an alternate approach to AD screening in DS. Our results are from one clinic population; the patient cohort in the Down syndrome clinic team may not generalize to all adults with DS, and our clinic protocols may not generalize to all specialty clinics for DS. Caregivers completing the instrument were familiar with the clinical team treating the individual with DS with the possibility of bias in providing socially desirable answers. Conversely, clinicians were aware of the results of the ABDQ prior to the visit (not blinded) as we were using the ABDQ as a clinical screening tool; this could have impacted the clinician’s impression of dementia. All data were collected during the COVID-19 pandemic, with potential implications on behavior changes related to social isolation and on caregiver-reported change ascertainment bias related to increased time for direct observation by caregivers.

We relied on medical record review and input from our Down syndrome clinic team to determine if patients clinically were likely to have cognitive changes or not. Additionally, it is possible that the 9 false negatives (negative ABDQ score, but positive clinical suspicion) could reflect differences in clinical and caregiver perspective: for example, there may be early subtle signs that raise clinical suspicion for AD before caregivers become concerned for AD. In our clinical practice, we did not send all adults with DS for repeat neuropsychological testing if they had past baseline evaluation. As such, some of those who screened negative on the ABDQ – those with low clinical impression of dementia, were not referred for neuropsychological evaluation—a commonly used optimal approach for diagnosing dementia in DS, especially when repeated over time, nor were additional screening modalities including biomarker of imaging tests performed. It is therefore not possible to entirely exclude dementia in individuals with a negative ABDQ screen and low clinical impression as they were not referred for further testing. Without formal additional testing such as neuropsychological testing, in a process that relies primarily on screening, it is possible that dementia may have been missed in a patient who scored negative on ABDQ with low clinical suspicion, resulting in the potential for missed or delayed diagnosis, especially given possible ascertainment bias as treating clinicians were not blinded to ABDQ scores. Although we feel the likelihood of missed dementia diagnoses are low given the clinicians and caregivers knowledge of and longitudinal relationships with the patients included in the study, the quality improvement initiative highlights the perils of adopting a screening protocol where suitable screening instruments may not be available. Although there is much research on the use of AD measures in the research setting, more research is needed to understand the best practical clinical screening approach to detect AD in adults with DS in clinical setting. The NTG-EDSD, DLD and AAMD could all be piloted using a similar clinical validation approach at our clinic or other clinics caring for individuals with DS.

Despite these limitations, this study has several important findings worth noting. Most importantly, our study found that one-time use of a previously validated population specific dementia screening tool did not improve the process for diagnosing dementia in adults with DS in a clinical setting. Making caregiver-informed face value modifications to the screening process did not improve our program’s ability to identify dementia. Clinical assessment along with caregiver report remained more helpful in guiding the dementia screening process than reliance on a one-time ABDQ score.

Beyond the importance of following evidence-based care guidelines [9], it is important to screen adults with DS for AD given the high frequency of AD co-occurrence in DS and the disease burden for individuals with DS diagnosed with AD and their caregivers. If screening for AD can identify those with the highest clinical suspicion, screening could be used to triage referrals to the services necessary for obtaining a diagnosis, such as neuropsychiatric evaluation. Although limited successful treatments are available at this time, an earlier clinical diagnosis could potentially result in earlier access to treatment if recent trials showing success [26] (e.g., amyloid-lowering immunotherapies) show similar efficacy in adults with DS. Researchers may wish to conduct additional studies to compare and contrast the use of the ABDQ, NTG-EDSD, DLD, and AAMD in clinical practice, assess the benefits of using a dementia screening questionnaire over time, or directly compare the capacity of screening questionnaires and neuropsychological testing for diagnosing dementia. Additionally, we present our results for other clinicians in DS clinics who may plan to implement similar screening protocols for AD based on the published evidence-based care guideline.

Despite strong psychometrics in original validation, the ABDQ did not function well as a clinical screener for dementia in a clinical care setting. Changing terminology in the modified ABDQ did not improve screener performance. Ongoing studies to determine the clinical utility of dementia screening tools for adults with DS are essential to achieve earlier detection, earlier diagnosis, and earlier treatment.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

NMO, MP, BGS, and SLS designed the research study, NMO, AH, CC, CB, MP, BGS, AT, AGS, and SLS performed the research, NMO, AH, AGS, and SLS analyzed the data. All authors contributed to the editorial changes in the manuscript, read and approved the final manuscript, and have participated sufficiently in the work and have agreed to be accountable for all aspects of the work.

This questionnaire and retrospective study were approved by the Massachusetts General Hospital institutional review board (IRB protocol: 2020P003890). The study was carried out in accordance with the guidelines of the Declaration of Helsinki and all patients or their families/legal guardians are informed consent.

This research received no external funding.

The authors declare no conflict of interest. Dr. Skotko occasionally consults on the topic of Down syndrome through Gerson Lehrman Group. He receives remuneration from Down syndrome non-profit organizations for speaking engagements and associated travel expenses. In the past two years, Dr. Skotko received annual royalties from Woodbine House, Inc., for the publication of his book, Fasten Your Seatbelt: A Crash Course on Down Syndrome for Brothers and Sisters. Within the past two years, he has received research funding from AC Immune, and LuMind IDSC Down Syndrome Foundation to conduct clinical trials for people with Down syndrome. Dr. Skotko is occasionally asked to serve as an expert witness for legal cases where Down syndrome is discussed. Dr. Skotko serves in a non-paid capacity on the Honorary Board of Directors for the Massachusetts Down Syndrome Congress and the Professional Advisory Committee for the National Center for Prenatal and Postnatal Down Syndrome Resources. Dr. Skotko has a sister with Down syndrome. Dr. Santoro has received research funding from LuMind IDSC Down Syndrome Foundation to conduct clinical trials for people with Down syndrome (DS) within the past 2 years. She serves in a nonpaid capacity on the Board of Directors of the Massachusetts Down Syndrome Congress, the Board of Directors of the Down Syndrome Medical Interest Group (DSMIG-USA), and the Executive Committee of the American Academy of Pediatrics Council on Genetics.

Supplementary material associated with this article can be found, in the online version, at https://doi.org/10.31083/JIN25713.