LAUSR

In their article “Profile of mild behavioral impairment and factor structure of the Mild Behavioral Impairment Checklist in cognitively normal older adults,” Creese et al. (2019b) sought to describe the prevalence of MBI and each individual symptom in a cognitively normal community sample and explore the latent structure of the MBI-C using exploratory factor analyses (Creese Griffiths et al. 2019b). Further, the authors examined whether the prevalence and latent structure were the same when participants responded directly and when informants responded for the participants. Mild behavioral impairment (MBI) is the sustained emergence of mild or severe neuropsychiatric symptoms that are not the result of other medical or psychiatric disorders and that occur prior to onset of dementia (Cieslak et al. 2018). Presence of MBI could be a risk factor for cognitive decline and dementia or could be the first clinical expression of dementia potentially occurring prior to any cognitive difficulties (Cieslak et al. 2018) and MBI may be an even stronger predictor of future dementia than Mild Cognitive Impairment (MCI) (Taragano et al. 2009). In 2012, a working group was formed to establish diagnostic criteria for MBI, which they define as neuropsychiatric symptoms that emerge in later life (≥ 50 years), with or without MCI, and are sustained for at least six months (Creese et al. 2019a; Ismail et al. 2016). The use of the Neuropsychiatric Inventory (NPI) and similar instruments to measureMBI is common (Andrews et al. 2018; Sheikh et al. 2018; Wise et al. 2019) and far from ideal as these scales were developed for use in populations who were already experiencing dementia and were often tested in institutionalized patients and, as such, were not completely relevant to cognitively normal or mild cognitively impaired community dwelling adults (Ismail et al. 2017). The 34-item mild behavioral impairment – checklist (MBI-C), which is free to use and in the public domain (www.MBItest.org), addresses many of these limitations and can be completed by patients, informants or clinicians and was designed to measure the five domains of MBI (i.e. apathy, emotional dysregulation, impulse dyscontrol, social inappropriateness and abnormal thoughts/perceptions) in fully functioning community dwelling adults. The study by Creese et al. (2019b) was one of the first to provide prevalence estimates for a cognitively normal community sample and 45% and 54% of participants had a least oneMBI-C symptom in selfreport and informant-report, respectively. These prevalence rates are substantially lower than a clinical sample of MCI and subjective cognitive decline in which 81.5% met the criteria for MBI (Sheikh et al. 2018). Given the clinical rather than community sample, the higher prevalence is to be expected. However, this study used the NPIQuestionnaire, which has a shorter references period and may overestimate the prevalence of MBI. In Creese et al.’s (2019b) sample the most common MBI symptoms reported by both participants and informants was those related to mood and anxiety, followed by impulse dyscontrol, apathy, social in appropriateness and abnormal thoughts and perceptions. Although Creese et al.’s (2019b) study provides one of the first estimates ofMBI prevalence in a noncognitively impaired community sample, the results may not be fully representative of most community samples. Their sample was highly educated, 57% had an undergraduate degree or higher, and women comprised 73% of the sample. Whether these prevalence rates would generalize to a more realistic sample that included more participants with lower education and amore balanced representation of sex and gender is unknown. More research is needed to understand if people with lower levels of education express symptoms of MBI in a similar manner to those with more education. Also important for future research is an exploration of potential sex and gender differences in MBI. The factor analyses for the informant and selfreport data were similar with both factor analyses resulting in five factors that at least broadly reflect the five domains that the MBI-C was designed to measure. In both factor analyses, a large proportion of items did not load onto any of the five factors: 44% and 41% in the self-report and informant factor International Psychogeriatrics (2020), 32:6, 673–675 © International Psychogeriatric Association 2020