Abstract
Objectives
To examine the association of neuropsychiatric symptom (NPS) severity with risk of transition to all-cause dementia, Alzheimer's disease (AD), and vascular dementia (VaD).
Design
Survival analysis of time to dementia, AD, or VaD onset.
Setting
Population-based study.
Participants
230 participants diagnosed with cognitive impairment, no dementia (CIND) from the Cache County Study of Memory Health and Aging were followed for a mean of 3.3 years.
Measurements
The Neuropsychiatric Inventory (NPI) was used to quantify the presence, frequency, and severity of NPS. Chi-square statistics, t-tests, and Cox proportional hazard ratios were used to assess associations.
Results
The conversion rate from CIND to all-cause dementia was 12% per year, with risk factors including an APOE ε4 allele, lower MMSE, lower 3MS, and higher CDR sum-of-boxes. The presence of at least one NPS was a risk factor for all-cause dementia, as was the presence of NPS with mild severity. Nighttime behaviors were a risk factor for all-cause dementia and of AD, while hallucinations were a risk factor for VaD.
Conclusions
These data confirm that NPS are risk factors for conversion from CIND to dementia. Of special interest is that even NPS of mild severity are a risk for all-cause dementia or AD.
Keywords: CIND, MCI, NPS, NPI, Cache County, dementia, depression, agitation, anxiety
Objective
There has been a notable increase in dementia prevalence driven by an aging population and an increased public awareness of its signs and symptoms1. This increase has been accompanied by the appreciation that there are individuals with cognitive impairment at a severity level insufficient for a diagnosis of dementia. The terms “Cognitive Impairment, No Dementia” (CIND) and “Mild Cognitive Impairment” (MCI)2 have been used to describe these individuals, many of whom are in the prodromal stages of dementia. Specifically, the amnestic form of MCI (aMCI) has been documented as a prodrome of Alzheimer's dementia (AD)3.
In community samples, transition from CIND/MCI to dementia occurs (on average) in 12-15% per year, although these rates vary4. Proposed risk factors for transition to dementia include cardiovascular variables5, stroke6, and APOE ε4 status7. There is a growing interest in identifying indicators that predict a greater likelihood of transition to dementia from CIND or MCI. The hope is to define objective ways to identify patients most likely to transition to dementia as well as clues on how to target specific treatments earlier.
Neuropsychiatric symptoms (NPS) are nearly universal in dementia through the course of illness8. As a result, and because NPS are associated with a worse prognosis in dementia9, the occurrence of NPS in MCI/CIND has become of interest to clinicians and researchers. In a recent population-based study using data from the Cache County Study of Memory Health and Aging (CCSMHA), we estimated the prevalence of NPS in individuals with CIND and MCI10. Thirty-one percent of participants with CIND had at least one NPS, a rate halfway between that in the cognitively normal (15%) and those with dementia (60%). These rates are comparable to those reported in the Cardiovascular Health Study11 and by Geda et al12. A subgroup analysis comparing participants with aMCI to those with other types of CIND showed no significant differences. Using data from the National Alzheimer Coordinating Center (NACC), the largest MCI cohort to date, Rosenberg et al. (2010) similarly found relatively few differences in NPS between aMCI and non-amnestic MCI (non-aMCI) but executive dysfunction was associated with depression, anxiety, agitation, disinhibition, irritability, and sleep problems.
In addition to being highly prevalent in MCI/CIND, several studies have linked NPS in CIND/MCI to dementia conversion7, 13-15. Modrego et al16 reported that major depression doubled the risk of transition from MCI to dementia, while a report from the Kungsholmen project suggested a doubling of transition risk with increasing levels of anxiety17.
In this follow-up to our original study of NPS in CIND/MCI using the CCSMHA, we estimate hazard of transition to all-cause dementia by overall total NPS severity, for individual NPS, and for NPS in the mood or psychosis domains. We repeat the above for AD or vascular dementia (VaD), and again for the subgroup of patients diagnosed with aMCI.
Methods
Sampling, Screening, and Procedure
The present study utilized data from the third and forth triennial waves of dementia ascertainment in Cache County, Utah (Figure 1). The methods of the CCSMHA have been detailed extensively elsewhere18, 19. Briefly, all permanent residents of the county who were 65 years or older in January 1995 (n=5677) were invited into the study. We enrolled 5092 (90%) in Wave 1, all of whom were screened for dementia in a multi-staged assessment protocol. The first stage of screening involved either a revised version of the Modified Mini-Mental State Exam (3MS)20, 21 or, for those unable to complete testing, the Informant Questionnaire of Cognitive Status in the Elderly22. Those scoring below the pre-determined education and sensory-adjusted score were selected for an informant-based interview which assessed clinical symptoms consistent with dementia23. Informed consent was obtained at each interview, directly from the participant for cognitively capable individuals (as determined by score on 3MS and clinical judgment), or from a knowledgeable informant otherwise, typically the next-of-kin. Institutional review boards at Utah State University, Duke University, and The Johns Hopkins University approved and oversaw all study procedures.
Figure 1. Cache County and Current Study Design.
1. CIND: cognitive impairment, no dementia; 2. aMCI: amnestic cognitive impairment.
Participants whose informant interviews were suggestive of dementia or a prodrome, were selected to undergo a clinical assessment with a trained research nurse and a neuropsychological technician. A designated subsample was created at Wave 1 to match each identified case of AD at a 2:1 ratio, matched on age, gender and APOE ε4 genotype group, except for participants aged 65-74, whom were matched at a 4:1 ratio18. Due to attrition, additional individuals were selected in Wave 3 to “replenish” the designated subsample. Given the expanded focus of the study in Waves 3 and 4 on mild cognitive impairment as well as dementia, additional designated subsample members were randomly selected to yield a 2:1 ratio of controls to cases, defined as all-cause dementia, “mild/ambiguous/prodromal AD,” or MCI-Peterson diagnoses, with 100% sampling of all persons aged 85 or older.
The clinical assessment consisted of neuropsychological assessment24, brief physical exam, and informant interview of the development and course of any clinical symptoms. A geropsychiatrist and neuropsychologist, with members of the examination team, reviewed results and assigned preliminary diagnoses of dementia or CIND. The diagnostic panel members were sensitive to the overlapping nature of dementia/CIND and depression, differentiating the two based on the different syndromes found in each.
Dating of age at dementia onset was based on review of the chronology of symptoms and dementia severity using the Clinical Dementia Rating25 (CDR) as a guide. Participants who met DSM-III-R criteria for dementia26 or prodromal AD were selected to undergo an MRI scan and laboratory studies to rule out non-degenerative pathology that could explain the present cognitive symptoms. Furthermore, those carrying a dementia diagnosis were invited to complete an exam by a geriatric psychiatrist. Diagnoses of AD, VaD, CIND, aMCI, and other forms of dementia were made by an interdisciplinary geropsychiatric team as previously described 27. Diagnosis of AD was made according to NINCDS-ADRDA criteria28. Vascular dementia diagnoses followed the NINDS-AIREN criteria 29. Other dementia diagnoses included: diffuse Lewy body dementia, Parkinson's Disease, frontotemporal dementia, traumatic brain injury, alcoholic dementia, and dementia of undetermined etiology. These were rarely the primary diagnosis and are hence not assessed as a subgroup in the current analysis. Participants with dementia at baseline were excluded prior to application of the diagnostic algorithm for CIND/MCI.
CIND diagnosis was assigned according to the criteria used in the Canadian Study of Health and Aging30. Diagnosis required either self report or an informant report of cognitive decline from a previous level of function and objective documentation of decline in a minimum of one of the following domains: memory, executive function, language, or visuospatial function. Operationally, decline was classified as 1.5 standard deviations below normative mean values corrected for age and education. Level of impairment could not rise to the level of dementia, operationalized as significant functional change exceeding a CDR level of 0.5. aMCI participants met the published criteria for at least one of: mild ambiguous/prodromal AD7, MCI by Petersen31, and/or MCI by CDR criteria13. The categories were not mutually exclusive. These subtypes were grouped together due to their use as the CIND types most predictive of conversion to AD. We did not have adequate sample sizes to evaluate each CIND subtype separately. By definition, participants that did not qualify for a diagnosis of CIND or dementia were assigned a CDR rating of 0.
Individuals with a designation of dementia or mild ambiguous prodromal AD were invited to complete an 18-month follow-up clinical assessment and expert panel review to confirm diagnoses (“conversion” from CIND to dementia could thus have occurred between the Wave 3 and 18-month follow-up clinical assessment, or between the latter and the Wave 4 clinical assessment). Participants without dementia at the end of each wave were eligible for participation in the subsequent wave. Mean time of follow-up was 3.3 years. In both Waves 3 and 4, the 3MS cut score was raised (to increase the sensitivity for identification of mild cognitive syndromes), the second screening stage of informant interview was eliminated (as this information was gathered at the third stage, or clinical assessment), and the designated subsample was replenished as discussed above. (Wave 2 3MS cut score: 86/87 for subjects aged <80 and 83/84 for subjects aged 80+; Wave 3 and Wave 4 3MS cut score: 90/91 for all ages).
Because the detailed neuropsychiatric Inventory (NPI)32 and detailed diagnoses of CIND were not rendered until the third study wave, the final sample for the present investigation included only those study participants who completed the NPI at the clinical assessment in wave 3 (considered “baseline” here) and wave 4 (see Figure 1 flow chart).
Assessment of Neuropsychiatric Symptoms
Neuropsychiatric symptoms were evaluated at the clinical assessment using the NPI. No other measures of NPS were used as the NPI is a systematic way to measure multiple NPS domains. The research nurse interviewed a knowledgeable informant about the presence of 12 specific domains: delusions, hallucinations, agitation/aggression, depression/dysphoria, anxiety, elation/euphoria, apathy/indifference, disinhibition, irritability/lability, aberrant motor behavior, disturbances of sleep, and disturbances of appetite/eating. The presence of symptoms in each domain during the past 30 days was queried with a screening question. If this was endorsed, specific follow up questions were asked to clarify the nature of the symptoms, and their frequency, severity, degree of change from premorbid, and treatment. As in previous studies11, 12, the NPS outcome variable was defined as either the presence or absence of a given NPS domain. The NPI was also analyzed by dividing into absence of any NPS versus ≥1 NPS (NPI Total). NPI severity was calculated at three levels: no NPS, at least one NPS but no NPS at a severity level ≥4, at least one NPS at a severity level ≥4 (NPI Severity). Lastly, the NPI mood domains (depression, anxiety, apathy, and irritability) and NPI psychosis (delusions and hallucinations) domains were analyzed as separate groups.
Analysis
The present analyses were carried out using STATA(SE) Version 10.1. T-test and χ2 analyses, as appropriate, were used to compare baseline demographics in CIND patients from Wave 3 who had follow-up to those who did not, as described above, at Wave 4. Significance was set at a p-value <0.05. Given the multiple comparisons performed, a p-value of 0.005 was considered significant for the individual NPS contrasts. Bivariate Cox proportional hazards ratio (HR) analyses were estimated for baseline characteristics, with the time variable defined as time from CIND diagnosis to dementia onset, and the failure variable being onset of all-cause dementia, or in other analyses AD or VaD. Years at risk were calculated from baseline to last follow-up or a diagnosis of dementia, AD, or VaD. The analyses of AD and VaD excluded other forms of dementia and compared participants who developed AD or VaD with those who did not develop any dementia over the follow-up period.
Multi-variable models (adjusted for age, education, Apoe status, and 3MS) were used to analyze each of the NPS (same time variable and failures as above). Multi-variable models (same adjustments as above) were calculated for presence/absence of any single NPS, total NPI severity, and the mood/psychosis domains of the NPI.
In addition, we used t-tests and χ2 to compare the 230 cases of CIND that made up the final sample with the 249 participants who did not have complete follow-up at wave 3 and wave 4 and were excluded in the final sample. There were no baseline differences between the two groups (results not shown).
Results
Table 1 displays demographic data for those who converted to all-cause dementia from CIND, and those who did not convert. Those who converted had lower scores on the MMSE or 3MS and higher scores on the CDR sum-of-boxes (CDR-SB).
Table 1. Baseline Demographic and Population Descriptors of Participants.
| MEAN (SD) / NO. (%) | MEAN (SD) / No. (%) | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| All CIND1 | Converters | Non-Converters | T-Test or Chi2* | All CIND1 | Converters | Non-Converters | T-Test or Chi2* | ||||||||
| Characteristic | (n=230) | (n=85) | (n=145) | Statistic | P Value | DF^ | Characteristic | (n=230) | (n=85) | (n=145) | Statistic | P Value | DF^ | ||
| Age, mean, y | 81.9 (5.1) | 82.7 (5.1) | 81.4 (5.1) | -1.89 | 0.060 | 228 | MMSE3, mean | 26.3 (2.6) | 25.6 (2.9) | 26.7 (2.4) | 3.11 | 0.002 | 227 | ||
| Gender, no. | 0.47 | 0.495 | 1 | 3MS4, mean | 87.4 (6.9) | 85.4 (7.3) | 88.5 (6.4) | 3.39 | <0.001 | 228 | |||||
| Male | 115 (50.0) | 45 (52.9) | 70 (48.3) | CDR5, no. | 3.2838 | 0.194 | 3 | ||||||||
| Female | 115 (50.0) | 40 (47.1) | 75 (51.7) | 0 | 20 (8.7) | 3 (3.5) | 17 (11.7) | ||||||||
| Education, mean, y | 13.4 (3.1) | 13.6 (3.3) | 13.3 (3.1) | -0.758 | 0.4494 | 228 | 0.5 | 206 (89.6) | 80 (94.1) | 126 (86.9) | |||||
| GMHR2, no. | 1.556 | 0.459 | 2 | 1 | 4 (1.7) | 2 (2.4) | 2 (1.4) | ||||||||
| Poor/Fair | 68 (29.6) | 26 (30.6) | 42 (29.0) | 2+ | 0 (0.0) | 0 (0.0) | 0 (0.0) | ||||||||
| Good | 129 (56.1) | 50 (58.8) | 79 (54.5) | CDR-SB6, mean | 1.30 (1.3) | 1.66 (1.4) | 1.08 (1.2) | -3.2986 | 0.0006 | 226 | |||||
| Excellent | 33 (14.3) | 9 (10.6) | 24 (16.5) | ||||||||||||
| ApoE Genotype, % ε4/x | 82 (35.7) | 35 (41.2) | 47 (32.4) | 1.51 | 0.2188 | 1 | |||||||||
CIND: cognitive impairment, no dementia;
GMHR: General Medical Health Rating scale;
MMSE: mini-mental state exam;
3MS: modified mini-mental state exam;
CDR: clinical dementia rating;
CDR-SB: CDR sum-of-boxes;
T-Test if mean and Chi2 if proportion;
:degrees of freedom.
Table 2 displays the past 30-day prevalence of NPS by NPI domain in individuals with CIND who did not convert to dementia, compared to those who converted to all-cause dementia, AD, and/or VaD. Statistical comparisons of converters to non-converters were non-significant unless noted below. Among those converting to all-cause dementia, only appetite/eating disturbances were significantly more common at baseline than in non-converters (χ2=5.86 p=0.016, df=1). In participants who converted to VaD, hallucinations were significantly more common at baseline (χ2=4.68, p=0.031, df=1).
Table 2. Neuropsychiatric Symptoms by NPI2 Domain at Baseline.
| No. (%) | |||||
|---|---|---|---|---|---|
| All CIND1 (n= 228) | Non –Converters (n=145) | Patients w/ CIND1 who converted to: | |||
| NPI2 Domain | All Dementia (n=83) | Alzheimer's Dementia (n=58) | Vascular Dementia (n=13) | ||
| Delusions | 1 (0.4) | 0 (0) | 1 (1.2) | 1 (1.8) | 0 (0) |
| Hallucinations | 2 (0.8) | 1 (0.7) | 1 (1.2) | 0 (0) | 1 (7.7) |
| Agitation/Aggression | 7 (3.1) | 3 (2.1) | 4 (4.8) | 3 (5.3) | 1 (7.7) |
| Depression/Dysphoria | 40 (17.5) | 24 (16.6) | 16 (19.3) | 8 (14.0) | 3 (23.1) |
| Anxiety | 11 (4.8) | 9 (6.2) | 2 (2.4) | 1 (1.8) | 0 (0) |
| Elation/Euphoria | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 0 (0) |
| Apathy/Indifference | 21 (9.2) | 14 (9.7) | 7 (8.4) | 3 (5.3) | 2 (15.4) |
| Disinhibition | 1 (0.4) | 1 (0.7) | 0 (0) | 0 (0) | 0 (0) |
| Irritability/Lability | 23 (10.1) | 14 (9.7) | 9 (10.8) | 3 (5.3) | 2 (15.4) |
| Motor Disturbance | 1 (0.4) | 0 (0) | 1 (1.2) | 1 (1.8) | 0 (0) |
| Nighttime Behaviors | 19 (8.3) | 11 (7.6) | 8 (9.6) | 6 (10.5) | 2 (15.4) |
| Appetite/Eating | 5 (2.2) | 1 (0.7) | 4 (4.8) | 2 (3.5) | 0 (0) |
CIND: cognitive impairment, no dementia;
NPI: neuropsychiatric inventory.
Table 3 displays HRs for conversion. For all-cause dementia, significant HRs were observed for increasing age, presence of at least one APOE ε4 allele, lower MMSE or 3MS scores, and higher CDR-SB scores. For AD, significant HRs were observed for increasing age, the presence of at least one APOE ε4 allele, lower 3MS scores, and higher CDR-SB. For VaD, significant HRs were observed for lower MMSE scores and higher CDR-SB.
Table 3. Hazard Ratios (HR) for Demographic and Population Descriptors of Participants (bivariate model).
| CIND1 participants who progressed to: | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| All Dementia (n=85)* | Alzheimer's Dementia (n=58)* | Vascular Dementia (n=13)* | ||||||||||
| Characteristic | HR | 95% CI | χ2 | P Value | HR | 95% CI | χ2 | P value | HR | 95% CI | χ2 | P Value |
| Age (yrs) | 1.05 | 1.01-1.10 | 6.40 | 0.011 | 1.08 | 1.02-1.13 | 8.69 | 0.003 | 1.05 | 0.94-1.16 | 0.74 | 0.389 |
| Gender (male) | 1.08 | 0.71-1.66 | 0.13 | 0.718 | 0.73 | 0.43-1.22 | 1.46 | 0.227 | 1.49 | 0.48-4.56 | 0.49 | 0.483 |
| Education (yrs) | 1.01 | 0.95-1.09 | 0.17 | 0.676 | 1.00 | 0.92-1.09 | 0.00 | 0.981 | 0.91 | 0.75-1.10 | 1.00 | 0.317 |
| GMHR2 | 0.86 | 0.62-1.20 | 0.76 | 0.386 | 1.09 | 0.73-1.62 | 0.18 | 0.673 | 0.54 | 0.22-1.33 | 1.87 | 0.172 |
| APOE Genotype: ε4/x vs. other | 1.58 | 1.02-2.44 | 4.06 | 0.044 | 2.09 | 1.24-3.52 | 7.47 | 0.006 | 1.09 | 0.34-3.55 | 0.02 | 0.885 |
| MMSE3 | 0.88 | 0.82-0.95 | 9.65 | 0.002 | 0.93 | 0.84-1.02 | 2.46 | 0.117 | 0.79 | 0.66-0.95 | 5.74 | 0.017 |
| 3MS4 | 0.96 | 0.94-0.98 | 11.0 | <0.001 | 0.97 | 0.94-0.99 | 4.28 | 0.039 | 0.98 | 0.92-1.05 | 0.20 | 0.658 |
| CDR5 | 1.74 | 0.92-3.28 | 2.20 | 0.138 | 1.69 | 0.77-3.72 | 1.32 | 0.251 | 2.54 | 0.73-8.86 | 1.31 | 0.253 |
| CDR-SB6 | 1.17 | 1.07-1.28 | 8.54 | <0.001 | 1.14 | 1.01-1.29 | 3.54 | 0.029 | 1.24 | 1.03-1.51 | 2.99 | 0.027 |
For all of the above analyses, degrees of freedom is one.
CIND: cognitive impairment, no dementia;
GMHR: General Medical Health Rating scale;
MMSE: mini-mental state exam;
3MS: modified mini-mental state exam;
CDR: clinical dementia rating;
CDR-SB: CDR sum-of-boxes.
Table 4 displays HRs for individual NPS. For all-cause dementia and AD, a significant HR was observed only for nighttime behaviors. The nighttime behaviors included difficulty falling asleep, early morning awakening, and wandering during the night. For VaD, a significant HR was observed only for hallucinations. The hallucinations assessed by NPI include experiencing sensations in the auditory, visual, olfactory and tactile modalities or other unusual sensory experiences from things that are not present or experienced by others. After adjustment, nighttime behaviors were significant predictors (p<0.005), whereas hallucinations were not.
Table 4.
Hazard Ratios (HR) for NPI Domains of CIND Participants who Progressed to Dementia (multivariate model controlled for age, education, APOE status, and 3MS).
| CIND1 participants who progressed to: | ||||||||||||
| All Dementia (n=85)* | Alzheimer's Dementia (n=58)* | Vascular Dementia (n=13)* | ||||||||||
| NPI2 Domain | HR | 95% CI | Z | P Value | HR | 95% CI | Z | P valve | HR | 95% CI | Z | P Value |
| Delusions | 0.82 | 0.22-2.99 | -0.30 | 0.761 | 0.95 | 0.35-2.63 | -0.09 | 0.928 | -- | -- | -- | -- |
| Hallucinations | 1.75 | 0.24-12.8 | 0.55 | 0.583 | -- | -- | -- | -- | 10.1 | 1.1-91.1 | 2.05 | 0.04 |
| Agitation/Aggression | 1.44 | 0.53-3.95 | 0.71 | 0.479 | 1.53 | 0.47-4.93 | 0.71 | 0.476 | 2.54 | 0.32-19.9 | 0.89 | 0.374 |
| Depression/Dysphoria | 1.10 | 0.63-1.92 | 0.33 | 0.741 | 0.77 | 0.36-1.64 | -0.68 | 0.497 | 1.76 | 0.47-6.56 | 0.85 | 0.398 |
| Anxiety | 0.55 | 0.13-2.25 | -0.83 | 0.404 | 0.43 | 0.06-3.11 | -0.84 | 0.401 | -- | -- | -- | -- |
| Elation/Euphoria | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- |
| Apathy/Indifference | 0.93 | 0.43-2.02 | -0.19 | 0.850 | 0.58 | 0.18-1.87 | -0.91 | 0.365 | 1.76 | 0.39-8.08 | 0.73 | 0.464 |
| Disinhibition | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- |
| Irritability/Lability | 0.99 | 0.49-1.98 | -0.04 | 0.971 | 0.47 | 0.15-1.51 | -1.27 | 0.206 | 1.53 | 0.34-6.96 | 0.55 | 0.581 |
| Motor Disturbance | 3.51 | 0.45-27.4 | 1.20 | 0.231 | 5.33 | 0.65-43.7 | 1.56 | 0.119 | -- | -- | -- | -- |
| Nighttime Behaviors | 1.28 | 1.08-1.52 | 2.78 | 0.005 | 1.34 | 1.12-1.60 | 3.19 | 0.001 | 1.17 | 0.59-2.29 | 0.45 | 0.656 |
| Appetite/Eating | 1.25 | 0.99-1.59 | 1.89 | 0.058 | 1.26 | 0.98-1.64 | 1.79 | 0.074 | -- | -- | -- | -- |
For all of the above analyses, degrees of freedom is five.
CIND: cognitive impairment, no dementia;
NPI: neuropsychiatric inventory;
“--” signifies a HR that could not be calculated due to inadequate sample size.
Table 5 displays HRs for the presence/absence of at least one NPS, NPI severity, and the subgroup of NPS related to mood. The sum of psychosis domains of the NPI were also analyzed, but prevalence was rare and comparisons were non-significant (data not shown). Significant HRs for all-cause dementia were seen for the presence of at least one NPS and for the presence of overall NPI domain severities. No significant HRs were observed for AD or VaD only, although this was likely due to low power.
Table 5.
Hazard Ratios (HR) for NPI subsets in CIND Participants who Progressed to Dementia (multivariate model controlled for age, education, APOE status, and 3MS).
| Mean (SD) / No. (% row) | Hazard Ratios (controlled for age, APOE status, 3MS) | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Non –Converters (NC) (n=145) | Patients w/ CIND1 who converted to: | Patients w/ CIND1 who converted to: | |||||||||||||||
| Characteristic | All Dementia (n=85) | Alzheimer's Dementia (n=58) | Vascular Dementia (n=13) | Dementia – All (n=85)* | Alzheimer's Dementia (n=58)* | Vascular Dementia (n=13)* | |||||||||||
| HR | 95% CI | Z | P Value | HR | 95% CI | Z | P Value | HR | 95% CI | Z | P Value | ||||||
| NPI2 Total, no. | 1.65 | 1.06-2.56 | 2.21 | 0.027 | 1.39 | 0.80-2.40 | 1.16 | 0.244 | 1.72 | 0.56-5.34 | 0.94 | 0.347 | |||||
| 0 | 107 (68.2) | 50 (31.8) | 37 (23.5) | 8 (8.3) | |||||||||||||
| ≥1 Domain | 38 (52.1) | 35 (47.9) | 21 (28.8) | 5 (6.8) | |||||||||||||
| NPI Severity | **Analyses for NPI, severity: minor vs. none & major vs. none** | ||||||||||||||||
| None = 0 | 107 (68.2) | 50 (31.8) | 37 (23.5) | 8 (8.3) | |||||||||||||
| Minor (no domain ≥4) | 26 (51.0) | 25 (49.0) | 14 (27.5) | 3 (5.9) | 1.65 | 1.01-2.69 | 1.99 | 0.047 | 1.31 | 0.70-2.46 | 0.84 | 0.404 | 1.41 | 0.37-5.39 | 0.50 | 0.616 | |
| Major (≥1 domain ≥4) | 12 (60.0) | 8 (40.0) | 6 (30.0) | 2 (10.0) | 1.24 | 0.58-2.66 | 0.56 | 0.576 | 1.27 | 0.52-3.08 | 0.53 | 0.596 | 3.19 | 0.63-16.2 | 1.40 | 0.162 | |
| NPI, mood domain3, mean | 0.42 (0.82) | 0.41 (0.64) | 0.26 (0.48) | 0.54 (0.88) | 0.98 | 0.75-1.29 | -0.14 | 0.890 | 0.75 | 0.49-1.14 | -1.35 | 0.176 | 1.22 | 0.68-2.21 | 0.67 | 0.505 | |
For all of the above analyses, degrees of freedom is five.
CIND: cognitive impairment, no dementia;
NPI: neuropsychiatric inventory;
Based on sum of 4 mood domains of the NPI (depression, anxiety, apathy, irritability).
Restricting the analyses to participants with aMCI (n=104), significantly lower MMSE (t- test stat=3.21, p≤0.001, df=102) and 3MS (t-test stat=2.04, p=0.02, df=102) scores were seen in converters. χ2 for the individual NPS were limited by the small number of individuals with aMCI who displayed each NPS. Significant χ2 were seen for increased prevalence of: nighttime behavior (χ2=3.86, p<0.05, df=1) in those converting to all-cause dementia, nighttime behaviors in those converting to AD (χ2=4.45, p=0.035, df=1), and for hallucinations in those converting to VaD (χ2=5.82, p=0.016, df=1).
The baseline demographic comparisons for the aMCI subgroup showed similar results to those for all CIND. Significant predictors of all-cause dementia were age (HR=1.06, 95% CI 1.01-1.12, χ2=5.76, p=0.016, df=1), MMSE (HR=0.86, 95% CI 0.79-0.95, χ2=9.05, p=0.003, df=1), 3MS (HR=0.96, 95% CI 0.94-0.99, χ2=5.30, p=0.02, df=1), and CDR sum-of-boxes (CDR-SB) (HR=1.14, 95% CI 1.00-1.29, χ2=2.94, p=0.049, df=1). Significant predictors for AD were age (HR=1.08, 95% CI 1.02-1.15, χ2=7.19, p=0.007, df=1) and the presence of at least one APOE ε4 allele (HR=2.09, 95% CI 1.12-3.88, χ2=5.21, p=0.02, df=1). Significant HRs were seen in those converting to VaD for MMSE (HR=0.78, 95% CI 0.62-0.98, χ2=3.94, p=0.047, df=1).
Among aMCI participants, individual NPS, presence/absence of single NPS, NPI severity, or the sum of the NPI mood or NPI psychosis domains were not risk factors for dementia or AD.
Discussion
We report that NPS even at low severity increase risk of transition from CIND to dementia. We observed a conversion rate from CIND to dementia of 12% per year, similar to the rates observed in prior studies14. We did not observe similar associations for conversion from aMCI to dementia. We recently reported that the prevalence of NPS in aMCI is comparable to CIND10, so the lack of significant results for the aMCI subgroup analysis is likely a result of inadequate sample size.
The characteristics associated with an increased risk of transition from CIND to all-cause dementia are not new, including known clinical risk factors for dementia incidence (older age, presence of an APOE ε4 allele) and evidence of greater cognitive and functional impairment (lower MMSE, lower 3MS, and higher CDR-SB). The presence of at least one APOE ε4 allele was also a risk factor for conversion to AD, but not to VaD, supporting its specificity to AD. The higher conversion risk associated with a higher CDR-SB was significant for all three dementia incidences analyzed: all-cause, AD, and VaD. It was the only of the baseline clinical measures to have this predictive power (i.e. MMSE, 3MS, and raw CDR score did not).
It is noteworthy that mild NPS severity had a HR for transition from MCI to dementia that was larger than for greater severity. The predictive value of individual NPS was limited to the novel finding that nighttime behaviors were a risk factor for all-cause dementia. Hallucinations were a risk factor for VaD, but given the small number of individuals experiencing hallucinations, this finding warrants replication. Given the prominence of NPS in both CIND and dementia, we were surprised that the predictive nature of individual NPS was poor.
Strengths / Limitations
The strengths of this study include its sample size, state of the art (for population studies) measurement of neuropsychiatric symptoms, and the carefully determined diagnoses of CIND/MCI. In addition, this is a study with a high participation rate among those with CIND. In terms of limitations, is the fact that NPI utilizes data provided by the caregiver, not the patient. In addition, the NPI has not been validated for use in MCI patients. Also, the Cache County population is predominantly Caucasian, possibly limiting generalizability to other populations. One limitation of the analyses is that although our assignment of age of onset for CIND and dementia was based on a review of the chronology of symptoms, it is still approximately +/- 1year.
Conclusions/Implications
Previous studies report that NPS are prevalent in all types of CIND, including aMCI, and predict dementia conversion. We confirm this in another population study. Specific individual NPS were risk factors for progression to dementia, in particular nighttime behaviors. As with other studies of CIND/MCI patients, the ultimate hope is that identification of those individuals most at risk for dementia conversion can improve prognosis and may suggest new therapeutic avenues.
Acknowledgments
Grants: Cache County Memory Study, Dementia Progression Study, and the Joseph and Kathleen Bryan ADRC: R01AG11380, R01AG21136, and AG028377.
Abbreviations
- 3MS
modified mini-mental state exam
- AD
Alzheimer's dementia
- aMCI
amnestic mild cognitive impairment
- CCSMHA
Cache County Study of Memory Health and Aging
- CIND
cognitive impairment, no dementia
- CDR
clinical dementia rating
- CDR-SB
clinical dementia rating sum-of-boxes
- CI
confidence interval
- HR
hazard ratio
- MCI
mild cognitive impairment
- MMSE
mini-mental state exam
- NACC
National Alzheimer's Coordinating Center
- Non-aMCI
non-amnestic mild cognitive impairment
- NPI
neuropsychiatric inventory
- NPS
neuropsychiatric symptoms
- NS
non-significant
- VaD
vascular dementia
Footnotes
Previous Presentation: Presented at the 2011 AAGP annual meeting in San Antonio, TX
Authors have no conflicts of interest unless listed: Dr. Rosenberg receives research support from Pfizer, Merck, Elan, Lilly, NIA, and the American Foundation for Aging research. Dr. Welsh-Bohmer is a consultant to Medivation and Zynfandel Pharmaceuticals and a paid speaker for Wyeth & Elan Pharmaceuticals. Dr. Lyketsos has received grant support (research or CME) from NIMH, NIA, Associated Jewish Federation of Baltimore, Weinberg Foundation, Forest, Glaxo-Smith-Kline, Eisai, Pfizer, Astra-Zeneca, Lilly, Ortho-McNeil, Bristol-Myers, Novartis, National Football League, Elan; has been consultant or advisor to Astra-Zeneca, Glaxo-Smith Kline, Eisai, Novartis, Forest, Supernus, Adlyfe, Takeda, Wyeth, Lundbeck, Merz, Lilly, Pfizer, Genentech, Elan, NFL Players Association, NFL Benefits Office; and has received honorarium or travel support from Pfizer, Forest, Glaxo-Smith Kline, Health Monitor.
References
- 1.Rabins PV, Lyketsos CG, Steele CD. Practical Dementia Care. 2nd. New York: Oxford University Press; 2006. [Google Scholar]
- 2.Petersen RC, Stevens JC, Ganguli M, Tangalos EG, Cummings JL, DeKosky ST. Practice parameter: early detection of dementia: mild cognitive impairment (an evidence-based review). Report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology. 2001 May 8;56(9):1133–1142. doi: 10.1212/wnl.56.9.1133. [DOI] [PubMed] [Google Scholar]
- 3.Jagust W. Is amnestic mild cognitive impairment always AD? Neurology. 2008;70(7):502. doi: 10.1212/01.wnl.0000299190.17488.b3. [DOI] [PubMed] [Google Scholar]
- 4.Tuokko H, Frerichs RJ. Cognitive impairment with no dementia (CIND): longitudinal studies, the findings, and the issues. Clin Neuropsychol. 2000 Nov;14(4):504–525. doi: 10.1076/clin.14.4.504.7200. [DOI] [PubMed] [Google Scholar]
- 5.Mielke MM, Rosenberg PB, Tschanz J, et al. Vascular factors predict rate of progression in Alzheimer disease. Neurology. 2007 Nov 6;69(19):1850–1858. doi: 10.1212/01.wnl.0000279520.59792.fe. [DOI] [PubMed] [Google Scholar]
- 6.De Ronchi D, Palmer K, Pioggiosi P, et al. The combined effect of age, education, and stroke on dementia and cognitive impairment no dementia in the elderly. Dement Geriatr Cogn Disord. 2007;24(4):266–273. doi: 10.1159/000107102. [DOI] [PubMed] [Google Scholar]
- 7.Tschanz JT, Welsh-Bohmer KA, Lyketsos CG, et al. Conversion to dementia from mild cognitive disorder: the Cache County Study. Neurology. 2006 Jul 25;67(2):229–234. doi: 10.1212/01.wnl.0000224748.48011.84. [DOI] [PubMed] [Google Scholar]
- 8.Steinberg M, Shao H, Zandi P, et al. Point and 5-year period prevalence of neuropsychiatric symptoms in dementia: the Cache County Study. Int J Geriatr Psychiatry. 2008 Feb;23(2):170–177. doi: 10.1002/gps.1858. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Lyketsos CG, Colenda CC, Beck C, et al. Position statement of the American Association for Geriatric Psychiatry regarding principles of care for patients with dementia resulting from Alzheimer disease. Am J Geriatr Psychiatry. 2006 Jul;14(7):561–572. doi: 10.1097/01.JGP.0000221334.65330.55. [DOI] [PubMed] [Google Scholar]
- 10.Peters ME, Rosenberg P, Steinberg M, et al. Prevalence of neuropsychiatric symptoms in CIND and its subtypes: the Cache County study. Am J Geriatr Psychiatry. 2011 doi: 10.1097/JGP.0b013e318211057d. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Lyketsos CG, Lopez O, Jones B, Fitzpatrick AL, Breitner J, DeKosky S. Prevalence of neuropsychiatric symptoms in dementia and mild cognitive impairment: results from the cardiovascular health study. Jama. 2002 Sep 25;288(12):1475–1483. doi: 10.1001/jama.288.12.1475. [DOI] [PubMed] [Google Scholar]
- 12.Geda YE, Roberts RO, Knopman DS, et al. Prevalence of neuropsychiatric symptoms in mild cognitive impairment and normal cognitive aging: population-based study. Arch Gen Psychiatry. 2008 Oct;65(10):1193–1198. doi: 10.1001/archpsyc.65.10.1193. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Morris JC, Storandt M, Miller JP, et al. Mild cognitive impairment represents early-stage Alzheimer disease. Arch Neurol. 2001 Mar;58(3):397–405. doi: 10.1001/archneur.58.3.397. [DOI] [PubMed] [Google Scholar]
- 14.Tuokko H, Frerichs R, Graham J, et al. Five-year follow-up of cognitive impairment with no dementia. Arch Neurol. 2003 Apr;60(4):577–582. doi: 10.1001/archneur.60.4.577. [DOI] [PubMed] [Google Scholar]
- 15.Bennett DA, Wilson RS, Schneider JA, et al. Natural history of mild cognitive impairment in older persons. Neurology. 2002 Jul 23;59(2):198–205. doi: 10.1212/wnl.59.2.198. [DOI] [PubMed] [Google Scholar]
- 16.Modrego PJ, Ferrandez J. Depression in patients with mild cognitive impairment increases the risk of developing dementia of Alzheimer type: a prospective cohort study. Arch Neurol. 2004 Aug;61(8):1290–1293. doi: 10.1001/archneur.61.8.1290. [DOI] [PubMed] [Google Scholar]
- 17.Palmer K, Berger AK, Monastero R, Winblad B, Backman L, Fratiglioni L. Predictors of progression from mild cognitive impairment to Alzheimer disease. Neurology. 2007 May 8;68(19):1596–1602. doi: 10.1212/01.wnl.0000260968.92345.3f. [DOI] [PubMed] [Google Scholar]
- 18.Breitner JC, Wyse BW, Anthony JC, et al. APOE-epsilon4 count predicts age when prevalence of AD increases, then declines: the Cache County Study. Neurology. 1999 Jul 22;53(2):321–331. doi: 10.1212/wnl.53.2.321. [DOI] [PubMed] [Google Scholar]
- 19.Lyketsos CG, Steinberg M, Tschanz JT, Norton MC, Steffens DC, Breitner JC. Mental and behavioral disturbances in dementia: findings from the Cache County Study on Memory in Aging. Am J Psychiatry. 2000 May;157(5):708–714. doi: 10.1176/appi.ajp.157.5.708. [DOI] [PubMed] [Google Scholar]
- 20.Teng EL, Chui HC. The Modified Mini-Mental State (3MS) examination. J Clin Psychiatry. 1987 Aug;48(8):314–318. [PubMed] [Google Scholar]
- 21.Tschanz JT, Welsh-Bohmer KA, Plassman BL, Norton MC, Wyse BW, Breitner JC. An adaptation of the modified mini-mental state examination: analysis of demographic influences and normative data: the cache county study. Neuropsychiatry Neuropsychol Behav Neurol. 2002 Mar;15(1):28–38. [PubMed] [Google Scholar]
- 22.Jorm AF, Jacomb PA. The Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE): socio-demographic correlates, reliability, validity and some norms. Psychol Med. 1989 Nov;19(4):1015–1022. doi: 10.1017/s0033291700005742. [DOI] [PubMed] [Google Scholar]
- 23.Kawas C, Segal J, Stewart WF, Corrada M, Thal LJ. A validation study of the Dementia Questionnaire. Arch Neurol. 1994 Sep;51(9):901–906. doi: 10.1001/archneur.1994.00540210073015. [DOI] [PubMed] [Google Scholar]
- 24.Tschanz JT, Welsh-Bohmer KA, Skoog I, et al. Dementia diagnoses from clinical and neuropsychological data compared: the Cache County study. Neurology. 2000 Mar 28;54(6):1290–1296. doi: 10.1212/wnl.54.6.1290. [DOI] [PubMed] [Google Scholar]
- 25.Morris JC. The Clinical Dementia Rating (CDR): current version and scoring rules. Neurology. 1993 Nov;43(11):2412–2414. doi: 10.1212/wnl.43.11.2412-a. [DOI] [PubMed] [Google Scholar]
- 26.Diagnostic and statistical manual of mental disorders: DSM-III-R. Washington, DC: American Psychiatric Association; 1987. [Google Scholar]
- 27.Chuang YF, Hayden KM, Norton MC, et al. Association between APOE epsilon4 allele and vascular dementia: The Cache County study. Dement Geriatr Cogn Disord. 29(3):248–253. doi: 10.1159/000285166. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan EM. Clinical diagnosis of Alzheimer's disease: report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer's Disease. Neurology. 1984 Jul;34(7):939–944. doi: 10.1212/wnl.34.7.939. [DOI] [PubMed] [Google Scholar]
- 29.Román GC, Tatemichi TK, Erkinjuntti T, et al. Vascular dementia. Neurology. 1993;43(2):250–250. doi: 10.1212/wnl.43.2.250. [DOI] [PubMed] [Google Scholar]
- 30.Ebly EM, Hogan DB, Parhad IM. Cognitive impairment in the nondemented elderly. Results from the Canadian Study of Health and Aging. Arch Neurol. 1995 Jun;52(6):612–619. doi: 10.1001/archneur.1995.00540300086018. [DOI] [PubMed] [Google Scholar]
- 31.Petersen RC, Smith GE, Waring SC, Ivnik RJ, Tangalos EG, Kokmen E. Mild cognitive impairment: clinical characterization and outcome. Arch Neurol. 1999 Mar;56(3):303–308. doi: 10.1001/archneur.56.3.303. [DOI] [PubMed] [Google Scholar]
- 32.Cummings JL. The Neuropsychiatric Inventory: assessing psychopathology in dementia patients. Neurology. 1997 May;48(5 Suppl 6):S10–16. doi: 10.1212/wnl.48.5_suppl_6.10s. [DOI] [PubMed] [Google Scholar]