Association Between Dementia Severity and Recommended Lifestyle Changes: A Retrospective Cohort Study

Abstract

Objective:

Early diagnosis of dementia leads to early treatment which is beneficial to patients and the community. We reviewed initial evaluations from the Spectrum Health Medical Group Neurocognitive Clinic (SHMGNC) to evaluate dementia stage at the time of diagnosis.

Methods:

We retrospectively reviewed 110 randomly chosen initial evaluations from September 2008 to December 2015 at the SHMGNC. Patients underwent a neurological examination, Montreal Cognitive Assessment, and a battery of neuropsychological testing.

Results:

Of all, 78.9% had moderate or severe dementia at diagnosis. The SHMGNC recommended lifestyle changes (medication assistance, financial assistance, driving restrictions, and institutional care) in 75.8% of patients with dementia. The severity of dementia was associated with the number of lifestyle changes recommended. Cohabitation with a caregiver did not lead to an early diagnosis of dementia.

Conclusions:

Patients are not undergoing evaluation at the onset of the dementia process. Diagnosis is delayed. Home-based, patient-centered care may improve early screening and detection of dementia.

Background

Alzheimer’s disease (AD) is the most common form of dementia. In 2015, an estimated 46.8 million people worldwide were living with dementia—a number that is projected to increase to 74.7 million by 2030 and to 131 million by 2050. ¹ Costs of managing the disease are astonishing. According to the World Alzheimer Report, dementia is estimated at US$818 billion for 2015, an increase from $604 billion spent in 2010. ² Informal care accounted for US$251.9 billion in 2010. ²

The medical community is striving to diagnose AD before dementia becomes debilitating. An asymptomatic stage of AD has been described in people with no cognitive deficits but in whom amyloid depositions in the brain have occurred. A prodromal stage of AD, mild cognitive impairment (MCI), has also been well described. Imaging studies can detect people in the asymptomatic stage of AD by recording deposition of extracellular Aβ plaques before symptoms appear up to 20 years before. ³ In the current state, such imaging is only available in the research world and requires a primary care provider (PCP) referral to a memory disorder clinic for a comprehensive evaluation and diagnosis. We retrospectively reviewed a random sample of initial patient evaluations to determine whether patients are presenting early in the course of the disease.

Methods

This is a retrospective review of 110 randomly chosen from 1,450 initial dementia evaluations in the Spectrum Health Medical Group Neurocognitive Clinic (SHMGNC) from September 2008 through December 2015. One thousand four hundred fifty patients were run through a simple random sample number generator that outputted 40 patients per each year from 2008 to 2015 for a total of 320 patients. After those patients were randomly generated, 110 patients were determined as having complete charts. Patients were referred to SHMGNC for a cognitive evaluation by their PCP. Data were collected on age, gender, education level, living situation, duration of symptoms, neurological examination, and Montreal Cognitive Assessment (MoCA). This study included diagnoses of no cognitive deficits, MCI, vascular dementia (VaD), AD, or a mix of VaD and AD but excluded frontotemporal dementia, Parkinson disease with dementia, Lewy body dementia, or corticobasal degeneration.

Description of the Initial Dementia Evaluation

During the initial dementia evaluation, all patients were evaluated in a multidisciplinary neurocognitive clinic by a clinical pharmacist, neurologist, social worker, and neuropsychologist. Alzheimer’s disease was diagnosed according to the National Institute of Neurological Disorders and Stroke (NINDS) diagnostic criteria for AD. ⁴ Mild cognitive impairment was diagnosed based upon MCI criteria. ⁵ Vascular dementia was diagnosed based upon NINDS criteria for VaD. ⁶ An MoCA score of 20 to 26 was considered mild dementia, 13 to 19 moderate, and <12 severe. All members of the multidisciplinary clinic reached consensus on diagnosis. The pharmacist reviewed the patient’s medications and the manner in which the medications were administered. If it was apparent that a patient was not taking their medications correctly, a recommendation involving the administration of medications was made. The social worker discussed driving habits and financial affairs with the patient and family members to determine whether the patient had a history of driving safety issues or financial irregularities. Successful completion of Trails B and processing speed have been found to correlate well with safe driving skills. ^7,8 Recommended changes to driving habits and financial affairs were made based on driving aptitude, processing skills, and stated financial irregularities. If it was apparent that the patient’s cognitive impairment was so broad reaching, a recommendation was made that the patient could not live independently. All participants were at least 65 years and on Medicare. All participants received a mailed documentation prior to evaluation concerning the process and requesting participants to have someone attend clinic with them.

Statistical Method

Descriptive statistics were used to summarize the characteristics of the population. Continuous variables are reported as mean ± standard deviation and categorical variables are shown as percentages and counts. Categorical variables were analyzed using Fisher exact test except when looking at the individual lifestyle changes and dementia severity. To determine whether the severity of dementia groupings is unequal, a χ² goodness-of-fit test was utilized. For that analysis, a χ² test was used with a Bonferroni correction applied to the P values. The study was powered before to assess whether the severity was equal among the patients with dementia using a medium effect size (w = 0.30), power of 0.80, α of 0.05, and df = 2, and the goodness-of-fit test yielded a sample size of 108 total. Post hoc power analysis was performed from the goodness-of-fit test and determined from the effect size (w = 0.34), α (0.05), sample size (95 [dementia only]) and df = 2 the power yielded 0.85. Power analysis was performed using G*Power version 3.1.5. Statistical analyses for this paper were performed using SAS Enterprise Guide software, version 7.1 of the SAS System for Windows, copyright © 2014 SAS Institute Inc. Study data were collected and managed using REDCap electronic data capture tools hosted at Spectrum Health. ⁹ The institutional review board of Spectrum Health (IRB#: 2016-099) reviewed the study.

Results

Risk factors for developing AD include advanced age, low education level, and female gender. ^{10 –12} Our cohort is mostly consistent with these published risk factors. The majority of those in the cohort diagnosed with dementia were female (61.1%). The mean age of the cohort was 80 years, with a range from 65 to 92 years. Dementia incidence increased with age except for those 90+ years old. In our study population, 8.4% (n = 8) were in their 60s, 37.9% (n = 36) were in their 70s, 48.4% (n = 46) were in their 80s, and 5.3% (n = 5) were in their 90s. Interestingly, the level of education in our cohort did not show a trend for developing dementia; only 16.9% (n = 16) of patients had less than a high school education. Most of the cohort were at least high school graduates (57.9%) and 25.3% had an advanced degree. These demographic characteristics along with additional breakdowns can be found in Table 1.

Table 1.

Study Population Characteristics.^a

	Total (N = 110)	Dementia (n = 95)	Not Dementia (n = 15)
Age (years)	79.2 ± 6.6	80.0 ± 6.3	74.5 ± 6.5
Gender
Female	72 (65.4)	58 (61.1)	14 (93.3)
Male	38 (34.6)	37 (38.9)	1 (6.7)
Age group
60s	12 (10.9)	8 (8.4)	4 (26.7)
70s	43 (39.1)	36 (37.9)	7 (46.7)
80s	50 (45.4)	46 (48.4)	4 (26.7)
90s	5 (4.6)	5 (5.3)	0 (0.0)
Education	7 (6.4)	7 (7.4)	0 (0.0)
Less than high school
Some high school	11 (10.0)	9 (9.5)	2 (13.3)
High school graduate	64 (58.2)	55 (57.9)	9 (60.0)
Advanced degree	28 (25.4)	24 (25.3)	4 (26.7)
Duration of memory loss (years)	2.7 ±2.7	2.8 ±2.8	1.9 ± 1.9

^aContinuous variables are mean ± standard deviation. All other values are number (% frequency).

Of the 110 charts screened, we identified that 13.6% (n = 15) of referred patients were diagnosed with an impairment other than dementia and 86.4% (n = 95) of patients had dementia. Of those not diagnosed with dementia, the primary diagnosis was anxiety/depression (53.3%) followed by MCI (26.7%). Of those diagnosed with dementia, the primary diagnosis was AD (89.5%) followed by a mixture of AD + VaD (9.5%; Supplement Table 1).

We next examined how severe the dementia was at referral. Severity of dementia was based upon MoCA score, a battery of neuropsychological tests, and reported functional impairments in daily life. Of the 95 with dementia, 21.1% (n = 20) had mild dementia, 48.4% (n = 46) had moderate dementia, and 30.5% (n = 29) had severe dementia (Table 2). Unfortunately, a majority of patients with dementia had progressed to moderate to severe dementia before the PCP referred for evaluation (P = .0029).

Table 2.

Dementia Severity by Living Situation.^a

Severity	Mild, 20 (21.1)	Moderate, 46 (48.4)	Severe, 29 (30.5)	Total 95
Living Situation
Alone	11 (55.0)	18 (39.1)	12 (41.4)	41 (43.2)
Married	9 (45.0)	25 (54.3)	10 (34.5)	44 (46.3)
Other	0 (0.0)	3 (6.5)	7 (24.1)	10 (10.5)

^aValues are number (% frequency).

We compared living arrangements with dementia severity to evaluate why patients are not diagnosed early in the disease process. We hypothesized that caregivers would recognize early cognitive changes resulting in a diagnosis of less severe dementia. Among those with early dementia, a nonsignificant majority lived alone (55%; n = 11) compared with patients living with a spouse (45%; n = 9). For moderate dementia, a majority lived with their spouse (54.3%; n = 25); only 6.5% (n = 3) lived with a caregiver other than a spouse and 39.1% (n = 18) lived alone. All 3 living arrangements consisted of patients with severe dementia: 41.4% (n = 12) lived alone, 34.5% (n = 10) lived with their spouse, and 24.1% (n = 7) lived with a nonspouse caregiver (Table 2). Using a Fisher exact test, we determined there was no statistically significant association between severity of disease and living arrangements (P = .0521). Interestingly, patients living alone are not disproportionately diagnosed early in the disease process. Conversely, dementia is not diagnosed early when the patient lives with a spouse or other caregiver.

As an additional measure of dementia severity, we looked at the frequency of lifestyle change recommendations for each patient: medication assistance, financial assistance, driving restrictions, and institutional care. The SHMGNC recommended at least one lifestyle change in 75.8% (n = 72) of the patients diagnosed with dementia. Patients with dementia were further stratified by the number of recommended lifestyle changes: 1 lifestyle change, 17.9% (n = 17); 2 changes, 24.2% (n = 23); 3 changes, 23.2% (n = 22); and 4 changes, 10.5% (n = 10). The SHMGNC recommended 33.7% (n = 32) of patients with dementia with broadly impaired cognitive function for institutional care. Individually, the recommendations that showed a statistically significant association with the severity of dementia were institutional care and driving (P = .0036 and .0116, respectively). The number of SHMGNC lifestyle change recommendations significantly associated with the severity of dementia (P = .001; Table 3).

Table 3.

Recommended Lifestyle Changes by Dementia Severity.^a

Severity	Total Dementia	Mild	Moderate	Severe
Number of recommended lifestyle changes
None	23	10 (43.5)	9 (39.1)	4 (17.4)
One	17	5 (29.4)	7 (41.2)	5 (29.4)
Two	23	3 (13.0)	16 (69.6)	4 (17.4)
Three	22	1 (4.5)	13 (59.1)	8 (36.4)
Four	10	1 (10.0)	1 (10.0)	8 (80.0)
Number of recommended lifestyle changes for each category
Medication	51	6 (11.8)	28 (54.9)	17 (33.3)
Financial	46	6 (13.0)	22 (47.8)	18 (39.1)
Driving	40	4 (10.0)	18 (45.0)	18 (45.0)
Institutional care	32	2 (6.3)	14 (43.8)	16 (50.0)

^aValues are number (% frequency).

Discussion

There are few published articles which discuss the severity of dementia at the time of initial diagnosis and none to our knowledge that have noted a correlation between the severity of dementia and the number of lifestyle change recommendations. ^11,13,14 This study identified that a majority of patients referred for dementia were diagnosed with moderate to severe dementia. At the time of referral, SHMGNC recommended 32 patients (33.7% of patients with dementia) for institutional care (Table 3). Further, institutional care was significantly associated with severity of dementia. These data are consistent with patients referred to neurospecialists late in disease progression irrespective of their living arrangements.

Cohabitation

Intuitively, one would expect that once a patient who cohabitates develops cognitive deficits, the caregiver would recognize the deficits and ask the PCP for a cognitive evaluation sooner in the disease process. However, in this study, severity of disease did not show an association with whether a patient lived alone or with a caregiver. Conversely, persons living alone were not diagnosed with mild dementia at significantly higher percentages compared to patients living with a caregiver (Table 2). The relationship between cohabitation and dementia severity is controversial. ^10,14,15 Counterintuitively, Sibley et al published that patients who live alone and develop cognitive deficits are diagnosed earlier compared with patients living with a caregiver. ¹⁵ In contrast, Swanwick et al showed no difference in dementia severity between living alone or living with a spouse, while patients living with a son or daughter were diagnosed with dementia earlier compared with those patients living alone or with a spouse. ¹⁰

Severity

A majority of patients with dementia were moderate or severe (n = 75, 78.9%) at the time of initial diagnosis (Table 2). The number of moderate to severe patients with dementia at presentation is higher in our cohort than previously published results. ^11,13,14 This high percentage of advanced disease is most likely attributed to the long delay between symptom onset and diagnosis. According to families and caregivers, symptoms of memory loss were present for nearly 3 years before a dementia referral. The driving factor underlying this delay is unknown. Previous publications proposed many reasons for the delay. ^16,17

Diagnosis

Despite the many published guidelines on diagnosis and management of dementia, many PCPs are not recognizing or referring patients early in the course of cognitive decline. ^{18 –20} Primary care providers are generally effective in diagnosing AD, although variability exists. ^19,21,22 It is estimated that cognitive impairment is unrecognized in 27% to 81% of affected patients in primary care practices. ^11,19,21 It may be difficult for PCPs to determine whether a patient is providing accurate information regarding their daily activities, lifestyle, or cognitive impairments especially within a brief office visit. Even when caregivers are present, symptoms might be overlooked and not communicated to the PCP. Education to patients, caregivers, and PCPs is vital to early diagnosis and impeding disease progression.

The National Institue of Health (NIH) clinical research site lists 479 ongoing, enrolling, or completed trials regarding the treatment of mild dementia between 1999 and 2015. A further 150 trials have been completed or are ongoing dealing with treatment of severe dementia. ²³ If new therapeutic interventions for the very early stages of AD or MCI are identified, it will be essential to test those novel treatments to see whether we can prevent patients in the prodromal stages of AD from progressing to definitive AD or patients with mild AD from progressing further. Economic benefits to the community are most apparent when a diagnosis of dementia is made early in the course of the disease process. ^12,24,25 The prodromal stage of AD and early AD will need to be identified with a far better efficiency to carry out those clinical trials.

Limitations

This study has several limitations. As a retrospective study, some data were not available. The delay between symptoms and diagnosis was represented in all 110 patients, not subdivided to just patients with dementia. The cause of delay from first identified symptoms and initial diagnosis after PCP referral was not discernable. It is unknown whether the PCP delayed referral by associating mild dementia symptoms as merely aging or symptoms were not mentioned to the PCP during clinic visits. Likewise, the average delay in diagnosis for each living arrangement was not determined. This is a single-center study and reflects this distinct center population. Although all patients are on Medicare, we did not assess whether copay was financial hardship creating referral delays.

The experience of our center is that neurospecialists are not seeing patients early in the disease. This is even true for patients cohabitating with their caregiver. Earlier recognition of cognitive deficits may aid in earlier referral and diagnosis. An increase in home-based, patient-centered medical care, regardless of the patient’s living status, may be one way to improve recognition of cognitive deficits and increase the frequency of important and necessary early cognitive evaluations. Educating the caregiver population on the advantages of an early diagnosis of dementia may lead to earlier requests to PCP asking for cognitive evaluation referrals.