Validation of a German version of the dementia screening questionnaire for individuals with intellectual disabilities (DSQIID-G) in Down's syndrome

Abstract

Background: People with Down's syndrome (DS) are at high risk of developing Alzheimer dementia (DS-AD) due to a triplication of the amyloid precursor protein gene. While several tools to diagnose and screen for DS-AD, such as the dementia screening questionnaire for individuals with intellectual disabilities (DSQIID), are available in English, validated German versions of such instruments are scarce. Methods: A German version of the DSQIID questionnaire (DSQIID-G) was completed by caregivers before attending our specialist outpatient department for DS-AD. All participants were assessed blind to DSQIID-G scoring using clinical and neuropsychological examinations, including the Cambridge Examination for Mental Disorders of Older People with Down's Syndrome and Others with Intellectual Disabilities (CAMDEX-DS). ICD-10 and amyloid/tau/neurodegeneration (A/T/N) criteria were applied to detect and categorise cognitive decline. Results: Of 86 participants, 43 (50%) showed evidence of cognitive decline. A definite diagnosis of DS-AD was reached in 17 (19.8%) and mild cognitive impairment in seven (8.3%) participants. Secondary causes of cognitive decline were determined among 13 (15.1%) participants, and in six (7%) cases, the diagnosis remained unclassifiable due to co-morbidities. Compared with cognitively stable individuals, participants with cognitive decline (n = 43) displayed higher DSQIID-G total scores [median (range): 3 (0–21) vs. 19 (0–48), P < 0.001]. A total score of >7 provided a sensitivity of 0.94 against a specificity of 0.76, to discriminate DS-AD and participants without cognitive decline according to ROC analysis. The convergent validity against the CAMDEX-DS interview score was good (r = 0.74), and split-half reliability (r = 0.96), internal consistency (Cronbach's α r = 0.96), test–retest reliability (r = 0.88) (n = 25) and interrater reliability (r = 0.81) (n = 31) were excellent. Conclusions: The DSQIID-G showed excellent psychometric properties, including concurrent and internal validity and reliability. The cut-off value for screening was lower than in the original English validation study. For a screening instrument like DSQIID-G, a lower cut-off is preferable to increase case detection.