. Author manuscript; available in PMC: 2015 Apr 7.

Published in final edited form as: Dement Geriatr Cogn Disord. 2014 Apr 7;38(1-2):89–146. doi: 10.1159/000357839

Table 6.

Differences between the BEHAVE-AD [1] and the BEHAVE-AD-FW [2] and the NPI [29, 30, 133] in the context of the assessment of treatment of BPSD-related disturbances in AD^*

Methodology	BEHAVE-AD [1] and BEHAVE-AD-FW [2]	NPI [29, 30, 133]
Information ascertainment methodology	Twenty-five, generally commonly occurring behavioral symptoms in AD, falling within 7 symptomatic categories are assessed. Each symptom is individually assessed on a 4-point severity scale for the BEHAVE-AD or on a 4-point severity scale and a 4-point frequency scale (with the exception of a 3-point frequency scale for 1 of the 25 symptomatic items) for the BEHAVE-AD-FW.	Surveys symptoms in 12 symptomatic categories. These include categories which are relevant for behavioral disturbances in AD and other categories which are not relevant for AD but applicable for mania or hypomania, affective disorder, or other conditions. Symptoms in the 12 categories are surveyed. Each category is scored on a 4-point frequency scale and a 3-point severity scale. Individual symptoms are not scored.
Nature of information obtained: relevance to AD	All 7 symptomatic categories contain symptoms which are commonly occurring at one or more dementia stages of AD. These range from a prevalence of >20% for Hallucinatory Disturbances in mild dementia and moderately severe dementia (GDS stages 4 and 6 [7]) to >60% prevalence at one or more dementia stages for the Paranoid and Delusional Ideation, Activity Disturbances, Aggressiveness, Affective Disturbances, and Anxieties and Phobias symptomatic categories [22]. Therefore, 5 of the 7 BEHAVE-AD symptomatic categories have a prevalence of >60% at one or more of the 4 GDS dementia stages.	Surveys 12 symptomatic categories, some of which have no relevance for AD or have more relevance for medication side effects than for the symptomatology of AD. For example, the NPI symptomatic category ‘Elation/Euphoria’, a symptom of mania and related conditions, was found to occur in only 1 of 214 AD persons at all levels of severity in a community survey [31]. This was similar to the prevalence in nondemented persons in the community. Therefore, Elation/Euphoria is not a symptom of AD. Another example is the NPI symptomatic category ‘Appetite/ Eating Disorders’, which includes symptoms such as increase in appetite, gaining weight, and poor appetite. These symptoms are not part of the behavioral disturbances of AD and are generally not part of AD in any context. Some of these symptoms occur in depressive disorder and they can occur as a result of side effects of putative AD therapies or be confused with the musculoskeletal wasting in the final GDS stage 7 of AD. Mixing symptoms which are not part of the behavioral pathologic syndrome of AD but which can occur as a result of medication side effects with symptoms which are behavioral disturbances can cloud the assessment of risk/benefit ratios of putative therapies for AD behavioral disturbances.
Separation of cognitive and functional changes in AD from the behavioral disturbances of AD versus mixing of cognitive and functional symptoms of AD with behavioral disturbances of AD	The purpose of a rating scale for the assessment of behavioral disturbances in AD is to assess those symptoms for treatment which may be applicable for these disturbances. These treatments include both pharmacological and nonpharmacological treatments. The BEHAVE-AD and the BEHAVE-AD-FW contain 7 symptomatic categories. All these symptomatic categories have demonstrated significant (p < 0.01) positive responses to the nonpharmacological effects of entry into the placebo arm of a clinical trial, and these effects appear to have been independent of effects on cognition [e.g., 25]. Furthermore, total BEHAVE-AD scores showed a significant additional response above that of the placebo condition to antipsychotic medication pharmacotherapy in this trial (p = 0.001). Positive responses were seen for 6 of the 7 BEHAVE-AD symptomatic categories. These effects were independent of cognition [25]. The one category of the BEHAVE-AD which did not show an additional positive response above that of the placebo condition to the pharmacotherapy in the antipsychotic medication trial of Katz et al. [25], i.e., Anxieties and Phobias, has shown a significant positive response to pharmacotherapy, above that of placebo, in an antianxiety medication trial [24]. In this study, once again, total BEHAVE-AD scores, BEHAVE-AD global scores, and each of the BEHAVE-AD symptomatic categories showed improvements on the nonspecific placebo intervention. An anxiolytic medication produced additional improvements above the placebo condition on 6 of the 7 BEHAVE-AD symptomatic categories. The one symptomatic category which showed no change in the Katz et al. [25] antipsychotic medication trial showed a significant added benefit, over and above that of the placebo condition, with the anxiolytic medication (buspirone) [24]. These effects in the Levy et al. [24] trial appear to have been independent of effects on cognition (i.e., no significant changes on the MMSE were seen). Therefore, all BEHAVE-AD symptomatic categories have been demonstrated to be potentially responsive to pharmacotherapies independent of the effects on cognition. These observations are compatible with the design of the BEHAVE-AD, which contains cognition-independent symptoms and symptomatic categories which, although related to AD, do not correlate with cognition or functional changes in AD. Another way of explaining the above statement is that all BEHAVE-AD and BEHAVE-AD-FW symptoms and symptomatic categories peak at some stage prior to the final GDS stage 7 of AD [7, 22]. Another qualitative difference of the BEHAVE-AD and BEHAVE-AD-FW symptomatic categories and the individual BEHAVE-AD and BEHAVE-AD-FW symptoms is that all these symptoms, and symptomatic categories, although generally commonly occurring at various stages, do not occur universally at any stage of AD. This prevalence is fundamentally different from cognitive and functional changes in AD, which become increasingly universal as the disease progresses.	The NPI contains a symptomatic category termed ‘Apathy/Indifference’, which, as defined, tracks very closely with the progressive cognitive and functional changes in AD. For example, as defined in the NPI, this category includes symptoms such as ‘is he/she more difficult to engage in… doing chores?’ Also, ‘does the patient contribute less to household chores?’ and ‘is it more difficult to keep a conversation going?’ These changes track very closely with the cognitive and functional progression of AD. For example, it has been demonstrated that each GDS stage of AD from MCI to mild dementia, to moderate dementia, is associated with progressive changes (losses) in self-care, household activities, general activities, food preparation, etc. [136]. Additionally, these progressive AD cognition and functional stage-related changes have been demonstrated for conversation and social functioning [136]. The problem with including a subscale which is primarily cognitively and functionally based in a behavioral inventory is that when applied to medications with positive effects on cognition and functioning, these positive effects can be misinterpreted as positive effects on behavioral disturbances.