Alzheimer’s & Dementia 4 (2008) 22–29

Prognosis of Alzheimer’s disease today: A two-year prospective study

in 686 patients from the REAL-FR Study
Frédéric Cortesa,b,*, Fati Nourhashémia,b, Olivier Guérinc, Christelle Canteta,b,
Sophie Gillette-Guyonneta,b, Sandrine Andrieub,d, Pierre-Jean Ousseta,
Bruno Vellasa,b, REAL-FR Group
a
Department of Internal Medicine and Clinical Gerontology, Centre Hospitalier Universitaire Purpan-Casselardit, Toulouse, France
b
Inserm U558, Toulouse, France
c
Department of Gerontology, University Hospital, Nice, France
d
Department of Epidemiology and Public Health, Toulouse, France

Abstract Background: The aim of the present study was to describe the long-term evolution of Alzheimer’s
disease (AD) in a prospective cohort of patients under treatment with a close follow-up.
Methods: Six hundred eighty-six AD patients from the French Network on AD (REAL-FR) were
followed up and assessed every 6 months for 2 years. Cognitive, functional, behavioral, nutritional, and
global status were evaluated by using Mini-Mental State Examination (MMSE), cognitive subscale of
AD Assessment Scale (ADAS-cog), Activities of Daily Living scale (ADL), Neuropsychiatric Inventory
(NPI), Mini-Nutritional Assessment (MNA), and Clinical Dementia Rating (CDR).
Results: There were 85.13% of patients who were specifically treated for AD during their partic-
ipation in the study. We observed significant changes (P ⬍ .0001) on MMSE, – 4.57 ⫾ 0.23;
ADAS-cog, 7.11 ⫾ 0.41; ADL, –1.32 ⫾ 0.07; NPI, 2.94 ⫾ 0.77; MNA, – 0.81 ⫾ 0.17; and sum of
boxes of the CDR (CDR-SB), 4.17 ⫾ 0.17. After 2 years, 10.79% (95% confidence interval [CI],
8.47 to 13.11) of the patients evolved twice as rapidly as the mean of the whole cohort on MMSE
(loss, ⱖ9 points), 65.89% (95% CI, 62.34 to 69.44) reported a loss of 3 to 9 points, and 23.32%
(95% CI, 20.16 to 26.46) were stable or improved (loss of –2 points maximum). Annual incidences
for institutionalization, hospitalization, and death were 11.84% (95% CI, 9.76 to 13.92), 26.13%
(95% CI, 22.52 to 29.74), and 5.95% (95% CI, 4.56 to 7.34), respectively.
Conclusions: In a recent large AD cohort mostly under treatment, AD evolution appeared to be
variable, with high incidences for death or institutionalization and with 11.84% of the patients
exhibiting a rapid cognitive decline, whereas one fourth of the cohort appeared in relatively stable
condition, and two thirds had a moderate but significant evolution of the disease. More studies are
needed to better understand these variations in patients’ evolution.
© 2008 The Alzheimer’s Association. All rights reserved.
Keywords: Alzheimer’s disease; Evolution; Prospective study; Specific management; Specific treatments

1. Introduction disease, and adapted management can be set up. Moreover,

therapies have been developed such as cholinesterase inhibi-
During the last several years, research about Alzheimer’s
tors (ChEIs) and nonspecific N-methyl D-aspartate (NMDA)
disease (AD) has intensified and allowed considerable progress
receptor antagonist that aim to relieve the symptoms and to
in the biologic and clinical knowledge of this pathologic con-
preserve the quality of life of the patients. These treatments are
dition. Diagnosis of AD can be made in the first stage of the
symptomatic, and their real efficiency and benefits are still
discussed [1]. To date, few studies have focused on the de-
*Corresponding author. Tel.: 33-5-61-77-76-06; Fax: 33-5-61-77-25-93. scription of the evolution of AD in patients treated and fol-
E-mail address: cortes.fr@chu-toulouse.fr lowed up. French Network on Alzheimer’s Disease (REAL-
1552-5260/08/$ – see front matter © 2008 The Alzheimer’s Association. All rights reserved.
doi:10.1016/j.jalz.2007.10.018
 F. Cortes et al. / Alzheimer’s & Dementia 4 (2008) 22–29
FR) is a French prospective multicentric study financed by the
French ministry of health, with a specific follow-up of AD
patients for 4 years. The objective of the present study was to
describe the long-term (2 years) evolution of AD in a large
prospective cohort of patients mostly under symptomatic treat-
ment with a close follow-up.
2. Methods
2.1. Subjects
The REAL-FR study is a prospective multicentric study
in which 686 patients with AD according to the National
Institute of Neurological and Communicative Diseases and
Stroke/Alzheimer’s Disease and Related Disorders Associ-
ation (NINCDS-ADRDA) [2] and Diagnostic and Statisti-
cal Manual of Mental Disorders (4th edition) (DSM-IV) [3]
criteria were recruited throughout France between 2000 and
2002. The methodology of the study has been described in
detail elsewhere [4]. Patients with mild to moderate disease,
Mini-Mental State Examination (MMSE) score between 10
and 26 [5], living in the community, and having a clearly
identified informal caregiver were included. At inclusion,
the patients underwent a full medical examination (includ-
ing computed tomography scan and thyroid tests). We ex-
cluded from the study patients with severe AD, those who
were institutionalized at baseline, and those with a concom-
itant disorder that could affect the short-term prognosis.
2.2. Measures
Follow-up included prospective data collection at
6-month visits for 2 years in 16 participating centers (ger-
ontology, neurology, or psychiatry). At each visit a stan-
dardized case report form was completed for each patient by
a trained, multidisciplinary medical team.
Evolution of the disease was assessed on various levels:
cognitive, functional, behavioral, nutritional, and global.
We used MMSE [5] and the cognitive subscale of the
Alzheimer’s Disease Assessment Scale (ADAS-cog) [6] for
cognitive evaluation, the Activities of Daily Living (ADL)
scale [7] to measure the capacity to carry out the activities
of daily living, the Neuropsychiatric Inventory (NPI) [8] for
the behavioral assessment, the Mini-Nutritional Assessment
(MNA) [9] for the measure of nutritional status, and the sum
of boxes of the Clinical Dementia Rating (CDR-SB) [10]
for the global status of the patients.
During follow-up, all events occurring between two vis-
its, in particular hospital admissions or nursing home place-
ments, were carefully recorded. Dropouts were also ana-
lyzed, and we distinguished deaths, entry to an institution
where follow-up was not possible, withdrawal of consent,
medical problems of patient or caregiver, or loss to follow-
up. To minimize the impact of attrition on our results, we set
up a procedure that allowed us to collect data concerning
vital status and institutionalizations of the patients who had
23
prematurely stopped the follow-up by establishing a contact
with the caregiver or general practitioner.
2.3. Statistical analysis
We first described baseline parameters of AD patients from
the REAL-FR cohort; for each of the modalities of the quali-
tative variables, the number and frequency are given; contin-
uous variables are expressed as means and standard deviations.
Longitudinal data were analyzed by using the mixed model
with random effects on intercept and time to take into account
the heterogeneity of the individual trajectories. Student t test
was used to determine the significance of the changes during
follow-up. This statistical method models individual trajecto-
ries and makes available all longitudinal data by accommodat-
ing unequal number of observations per subjects and unequal
intervals between follow-up assessments.
Concerning patients’ outcome at 2 years, incidences
were calculated for different events that occurred during the
course of the disease: institutionalization, hospitalization;
death; evolution toward a severe dementia (MMSE ⱕ10),
loss of at least 1 point on the ADL scale and loss of
independence for each item of the ADL, significant wors-
ening on the NPI total score or for each item (increase of 4
points or more, [18]), and change to a poorer nutritional
status assessed by MNA (three categories: normal nutri-
tional status [score, ⬎23.5], at risk of malnutrition [score,
17 to 23.5], or malnutrition [score, ⬍17]).
Incidences were also calculated to determine rates of
stopping or changing AD treatment (donepezil, rivastig-
mine, galantamine, and memantine).
Incidences were considered by using the time in days
from inclusion date to either event date (when the events
occurred for the first time, such as first hospitalization, first
change of AD treatment) or censoring date (which could
occur by lost to follow-up or termination of the observation)
except for stopping or changing AD treatment, where inci-
dences were calculated from the date of the prescription.
Aggravations on ADL, NPI, or MNA were evaluated ac-
cording to their baseline values.
3. Results
3.1. Baseline characteristics
Baseline characteristics of the 686 AD patients included
are summarized in Table 1. According to the inclusion
criteria, all patients presented a mild to moderate form of
AD and lived at home. The great majority of the patients
were specifically treated for AD. Figure 1 shows the evo-
lution of the cohort in detail during the 2 years of follow-up.
A total of 59 patients (8.60%) died, 58 (8.45%) withdrew
consent, 49 (7.14%) were institutionalized, 54 (7.87%) were
lost to follow-up, and 61 (8.90%) stopped the follow-up for
other reasons (removal, caregiver problem). Finally, 405
24 F. Cortes et al. / Alzheimer’s & Dementia 4 (2008) 22–29

Table 1 sponded to respective mean annual change of –2.3 points and

Baseline characteristics of the AD patients included in REAL-FR study ⫹3.5 points. We observed different profiles of evolution, with
Patients Included patients reporting very little impairment and others much
Age (y, mean ⫾ SD) 77.86 ⫾ 6.82 more; 452 patients (65.89%; 95% confidence interval [CI],
Sex (M/F, n, %) 198/488 (29.86/71.13) 62.34 to 69.44) reported a loss of 3 to 9 points on MMSE score
Duration of AD (y) 1.10 ⫾ 1.23 at 2 years, 74 (10.79%; 95% CI, 8.47 to 13.11) evolved twice
Living arrangement (n, %) as rapidly as the mean of the whole cohort and lost 9 points or
With relatives 503 (73.32)
more (maximum, –17 points), whereas 160 (23.32%; 95% CI,
Alone 183 (26.68)
Specific AD treatment (n, %) 611 (89.06) 20.16 to 26.49) were stable (evolution of MMSE score com-
MMSE score (/30, mean ⫾ SD) 20.01 ⫾ 4.23 prised between ⫹2 and –2 points) or improved (maximum,
ADAS-cog score (/70, mean ⫾ SD) 17.85 ⫾ 8.17 increase of 3 points). Regarding Table 2, we observed that
ADL score (/6, mean ⫾ SD) 5.43 ⫾ 0.90 change in MMSE score is not constant during the 2 years. This
NPI score (/144, mean ⫾ SD) 15.31 ⫾ 15.29
is confirmed by Figure 2A, which shows that the evolution of
MNA score (/30, mean ⫾ SD) 23.92 ⫾ 3.14
CDR-SB (/18, mean ⫾ SD) 6.42 ⫾ 3.30 MMSE score is not linear; the loss was about 1.9 points during
the first year and about 2.6 during the second. ADAS-cog
Abbreviation: SD, standard deviation.
evolution (Figure 2B) did not show any difference between the
first and second years. Moreover, the incidence for progression
patients (59%) completed the 2 years of follow-up, and data to severe dementia according to the MMSE score (ⱕ10) was
were available for 349 (50.9%) (Figure 1). 10.08% (95% CI, 7.98 to 12.18) of patients per year. The mean
duration of follow-up before progression to severe dementia
3.2. Cognitive evolution was 13.40 ⫾ 5.71 months.
Table 2 shows the MMSE and ADAS-cog evolution during 3.3. Functional evolution
the 2 years of follow-up. Mean changes at 2 years were sig-
nificant for these cognitive scales (P ⬍ .0001), – 4.6 points on ADL score diminished significantly (Table 2). Items
MMSE score and ⫹7.1 on ADAS-cog score, which corre- from the ADL scale with the highest incidence of decline to

Fig 1. Follow-up at each biannual visit during the 24 months of follow-up. *Medical problem of the caregiver, removal, inclusion in a therapeutic protocol.
Patients not evaluated at a visit were reassessed at one of the following visits.
 F. Cortes et al. / Alzheimer’s & Dementia 4 (2008) 22–29 25

Table 2
Mean scores at baseline, 1 year, 2 years, and global evolution at 2 years for MMSE, ADAS-cog, ADL, NPI, MNA, and CDR-SB scores by using results
from the mixed model
Parameters Baseline, Mean ⫾ SE 1 Year, Mean ⫾ SE 2 Years, Mean ⫾ SE Global Evolution at 2 Years, Mean ⫾ SE P Value*

MMSE (/30) 20.02 ⫾ 0.16 18.09 ⫾ 0.22 15.45 ⫾ 0.28 ⫺4.57 ⫾ 0.23 ⬍.0001
ADAS-cog (/70) 17.68 ⫾ 0.31 21.24 ⫾ 0.39 24.79 ⫾ 0.54 7.11 ⫾ 0.41 ⬍.0001
ADL (/6) 5.46 ⫾ 0.03 4.80 ⫾ 0.05 4.14 ⫾ 0.08 ⫺1.32 ⫾ 0.07 ⬍.0001
NPI (/144) 15.11 ⫾ 0.55 16.58 ⫾ 0.51 18.05 ⫾ 0.72 2.94 ⫾ 0.77 .0002
MNA (/30) 23.89 ⫾ 0.12 23.48 ⫾ 0.11 23.08 ⫾ 0.15 ⫺0.81 ⫾ 0.17 ⬍.0001
CDR-SB (/18) 6.40 ⫾ 0.13 8.15 ⫾ 0.16 10.57 ⫾ 0.20 4.17 ⫾ 0.17 ⬍.0001

Abbreviation: SE, standard error.

* Student test is used to test the significance of the evolution.

dependence were dressing and bathing (Table 3). The an-
nual incidence for the loss of one point or more on this scale
was equal to 51.64% (95% CI, 46.31 to 56.97), with a mean
duration of follow-up of 11.22 ⫾ 5.77 months before this loss.
Functional decline appeared to be constant (Figure 2C).
3.4. Behavioral evolution
NPI score was statistically impaired from its value at
baseline after 2 years of follow-up (Table 2). Apathy was
the behavioral disorder with the highest incidence (Table 3).

Fig 2. Evolution of the MMSE (A), ADAS-cog (B), ADL (C), NPI (D), MNA (E), and CDR-SB (F) scores (solid lines) ⫾ standard deviation (dotted lines)
in the REAL-FR cohort during the 24 months of follow-up.
26 F. Cortes et al. / Alzheimer’s & Dementia 4 (2008) 22–29

Table 3 and death were 26.13% (95% CI, 22.52 to 29.74), 11.84%
Incidence of decline of each ADL and NPI item (95% CI, 9.76 to 13.92), and 5.95% (95% CI, 4.56 to 7.34),
Items Incidence (% of 95% CI respectively.
Patients per Year)

ADL 3.8. AD progression according to baseline severity

Bathing 31.62 27.17 to 36.07
Dressing 34.35 29.51 to 39.19 We performed analysis regarding AD progression at 1
Toileting 15.11 12.50 to 17.72 and 2 years after stratifying patients by AD severity at
Transfer 11.22 8.97 to 13.47 baseline according to MMSE score. Three groups of pa-
Continence 22.92 19.37 to 26.47
Feeding 14.33 11.78 to 16.88
tients were obtained and were characterized as mild when
NPI MMSE score was 21 to 26 (n ⫽ 344), moderate when
Delusions 7.85 5.93 to 9.77 MMSE score was 16 to 20 (n ⫽ 223), and moderately
Hallucinations 3.96 2.67 to 5.25 severe when MMSE score was 10 to 15 (n ⫽ 117). As
Agitation/aggression 19.04 15.71 to 22.37 shown in Table 4, changes were significant for MMSE,
Depression/dysphoria 19.35 16.00 to 22.70
Anxiety 22.55 18.81 to 26.29
ADAS-cog, ADL, NPI, MNA, and CDR-SB scores for all
Relation/euphoria 3.95 2.64 to 5.26 groups of patients at 1 and 2 years except for MNA in
Apathy/indifference 41.21 35.02 to 47.40 moderate groups and NPI in moderately severe group.
Disinhibition 7.12 5.36 to 8.88 Moreover, MMSE, NPI, MNA, and CDR-SB scores
Irritability/lability 22.74 19.07 to 26.41
changes were not different between the three groups,
Aberrant motor behavior 20.85 17.40 to 24.30
Sleep 12.85 10.32 to 15.38 whereas ADL and ADAS-cog scores showed statistically
Appetite and eating disorders 23.08 19.38 to 26.78 significant differences. ADAS-cog evolution appeared to be
more important in the moderate group and the functional
NOTE. Decline ADL: dependency for the item of the scale. Decline
NPI: increase of 4 points or more. deterioration measured by ADL augmented with AD sever-
ity. Annual incidences for ADL aggravation (loss of 1
point or more on this scale) were 39.93% (95% CI, 33.68
Moreover, we found an incidence for a significant change to 46.18) in the mild group, 61.14% (95% CI, 50.46 to
(ⱖ4 points) in NPI total score of 54.44% (95% CI, 48.70 to 71.82) in the moderate group, and 80.10% (95% CI,
60.02), with a mean duration of follow-up before this in- 62.09 to 98.11) in the moderately severe group (P ⬍
crease of 10.33 ⫾ 5.30 months. NPI total score showed a .0001). Risk of institutionalization significantly increased
linear increase during the 2 years of follow-up (Figure 2D). with the severity of the disease (P ⫽ .0003); annual
3.5. Nutritional evolution incidences for placement were 8.29% (95% CI, 5.92 to
10.66) for the mild group, 14.43% (95% CI, 10.35 to
Mean MNA score decreased significantly during the 2 18.51) for the moderate group, and 19.69% (95% CI,
years of follow-up (Table 2), and the annual incidence for 12.65 to 26.73) for the moderately severe group. We did
the aggravation of the nutritional status was 27.86% (95% not find any statistically significant differences between
CI, 24.12 to 31.60). MNA evolution was linear (Figure 2E). the three groups for death, but it was near to the threshold
3.6. Global evolution of significance (P ⫽ .0607), with annual incidences for
mild, moderate, and moderately severe patients that were
The global evolution assessed by CDR-SB was statisti- 4.61% (95% CI, 2.90 to 6.32), 6.55% (95% CI, 3.98 to
cally significant after 2 years (Table 2). As shown in Figure 9.12), and 9.21% (95% CI, 4.84 to 13.58), respectively.
2F, curve of CDR-SB score evolution was not linear, with No statistical differences were reported concerning an-
an augmentation of 1.75 points during the first year and 2.4 nual incidences for hospitalization according to AD se-
during the second. verity at baseline: 24.43% (95% CI, 19.65 to 29.21) for
the mild group, 27.48% (95% CI, 20.89 to 34.07) for the
3.7. Treatment, hospitalizations,
moderate group, and 29.00% (95% CI, 19.26 to 38.74)
institutionalizations, death
for the moderately severe group (P ⫽ .6127). There were
Concerning AD-specific treatment, 584 (85.13%) pa- also no significant differences between the three groups
tients were treated continuously during the 2 years of with regard to annual incidences for stopping (P ⫽ .7821)
follow-up or before their withdrawal. Incidences for chang- or changing (P ⫽ .2050) AD treatments. It concerned
ing or stopping AD-specific treatment were 6.46% (95% CI, 1.83% (95% CI, 0.79 to 2.87) and 6.41% (95% CI, 4.43
5.05 to 7.87) and 2.14% (95% CI, 1.34 to 2.94) of patients, to 8.39) in the mild group, 2.46% (95% CI, 0.93 to 3.99)
respectively, per year. The duration of treatment before and 7.91% (95% CI, 5.09 to 10.73) in the moderate
these events was 19.63 ⫾ 13.51 months for the change of group, and 2.52% (95% CI, 0.50 to 4.54) and 4.35%
AD treatment and 19 ⫾ 16.3 months for its stopping. An- (95% CI, 1.65 to 7.05) in the moderately severe group,
nual incidences for hospitalizations, institutionalizations, respectively.
 F. Cortes et al. / Alzheimer’s & Dementia 4 (2008) 22–29 27

Table 4
Mean score at baseline and changes at 1 and 2 years for MMSE, ADAS-cog, ADL, NPI, MNA, and CDR-SB scores according to AD severity at
baseline by using results from the mixed model
Parameters Baseline Score Change at 1 Year Change at 2 Years P Value†

Mean ⫾ SE Mean ⫾ SE P Value* Mean ⫾ SE P Value*

MMSE (/30) .8284

Mild 23.47 ⫾ 0.10 ⫺1.83 ⫾ 0.21 ⬍.0001 ⫺4.56 ⫾ 0.32 ⬍.0001
Moderate 18.36 ⫾ 0.12 ⫺2.14 ⫾ 0.27 ⬍.0001 ⫺4.59 ⫾ 0.27 ⬍.0001
Moderately severe 13.03 ⫾ 0.17 ⫺1.87 ⫾ 0.38 ⬍.0001 ⫺4.65 ⫾ 0.61 ⬍.0001
ADAS-cog (/70) .0037
Mild 13.08 ⫾ 0.30 ⫹2.23 ⫾ 0.35 ⬍.0001 ⫹5.90 ⫾ 0.54 ⬍.0001
Moderate 19.52 ⫾ 0.38 ⫹3.56 ⫾ 0.46 ⬍.0001 ⫹8.86 ⫾ 0.74 ⬍.0001
Moderately severe 29.39 ⫾ 0.53 ⫹4.27 ⫾ 0.72 ⬍.0001 ⫹7.31 ⫾ 1.29 ⬍.0001
ADL (/8) ⬍.0001
Mild 5.68 ⫾ 0.05 ⫺0.44 ⫾ 0.04 ⬍.0001 ⫺0.88 ⫾ 0.09 ⬍.0001
Moderate 5.36 ⫾ 0.06 ⫺0.74 ⫾ 0.06 ⬍.0001 ⫺1.48 ⫾ 0.12 ⬍.0001
Moderately severe 4.99 ⫾ 0.08 ⫺1.16 ⫾ 0.08 ⬍.0001 ⫺2.33 ⫾ 0.17 ⬍.0001
NPI (/144) .5196
Mild 12.38 ⫾ 0.76 1.62 ⫾ 0.53 .0024 3.24 ⫾ 1.06 .0024
Moderate 16.99 ⫾ 0.95 1.83 ⫾ 0.69 .0081 3.65 ⫾ 1.37 .0081
Moderately severe 19.62 ⫾ 1.32 0.48 ⫾ 1.00 .6318 0.96 ⫾ 2.00 .6318
MNA (/30) .3519
Mild 24.45 ⫾ 0.16 ⫺0.39 ⫾ 0.12 .0012 ⫺0.78 ⫾ 0.24 .0012
Moderate 23.47 ⫾ 0.21 ⫺0.30 ⫾ 0.16 .0550 ⫺0.61 ⫾ 0.31 .0550
Moderately severe 22.99 ⫾ 0.29 ⫺0.69 ⫾ 0.22 .0021 ⫺1.39 ⫾ 0.45 .0021
CDR-SB (/18) .1803
Mild 4.54 ⫾ 0.14 ⫹1.53 ⫾ 0.15 ⬍.0001 ⫹3.93 ⫾ 0.23 ⬍.0001
Moderate 7.24 ⫾ 0.17 ⫹1.86 ⫾ 0.19 ⬍.0001 ⫹4.42 ⫾ 0.30 ⬍.0001
Moderately severe 10.26 ⫾ 0.23 ⫹2.21 ⫾ 0.27 ⬍.0001 ⫹4.39 ⫾ 0.43 ⬍.0001

Abbreviation: SE, standard error.

NOTE. Mild MMSE score, 21 to 26; moderate MMSE score, 16 to 20, and moderately severe MMSE score, 10 to 15.
* Student test is used to test the significance of the evolution at 1 and 2 years in each group.
†
Fisher test is used to compare evolutions between the three groups.

4. Discussion severity of the disease at entry for MMSE because we did

AD is a progressive insidious disorder that poses many
problems in epidemiologic studies because of many events
such as attrition, nonlinearity of longitudinal changes, and
variable length of follow-up that complicate statistical anal-
ysis. That is why we used a mixed model because this
overcomes problems in a longitudinal data collection of this
type as outlined by Mohs et al [11]. All data contribute to
the longitudinal analysis, and even incomplete data can be
used (patients who prematurely stop the follow-up) [12].
Global changes on MMSE and ADAS-cog are statisti-
cally significant. The 2-year evolution is largely variable,
with almost two thirds of the cohort (65.89%) with a mod-
erate but significant change on MMSE score (loss between
three and nine points), 10.79% who report a loss of nine
points or more corresponding to a significant worsening in
cognitive status, and 23.32% who are stable or improve. We
have already observed the existence of different profiles of
evolution according to MMSE score after the first year of
follow-up in the REAL-FR cohort, with almost 30% of the
patients reporting a more rapid cognitive decline by using
the threshold of a loss of at least three points during the year
on MMSE [13]. This decline seemed to be not related to the
not find any difference in AD progression when stratifying
by AD severity at baseline and as it has already been shown
in this cohort [14]. Annual rate of loss is –2.3 points on
MMSE score, but it appears that this decline is nonlinear,
with a loss of 1.9 points during the first year and 2.6 points
during the second for the whole cohort. This result supports
the nonlinearity of MMSE score evolution [15] and conse-
quently the nonlinearity of the cognitive function deterio-
ration [16]. In their study, Stern et al [16] used ADAS-cog
in 111 AD subjects followed every 6 months for up to 90
months and concluded that cognitive decline was slow
during mild and severe phases of AD and more rapid
during the moderate stage. In REAL-FR, ADAS-cog evo-
lution is linear; shorter length of follow-up might explain
this difference with the results of Stern et al. But when
regarding precisely ADAS-cog evolution according to
baseline severity, it appears that cognitive decline as-
sessed by ADAS-cog is significantly increased in the
moderate patients (MMSE score, 16 to 20) in accordance
with the results from Stern et al. These data underline
the better sensitivity of the ADAS-cog compared with
MMSE.
28 F. Cortes et al. / Alzheimer’s & Dementia 4 (2008) 22–29
As we observe for cognition, ability to perform ADL also
declines continuously during the course of AD. The loss on
ADL score is statistically significant, and about 50% of
patients lose one basic activity of daily living. We also
observed that patients with more severe AD at baseline are
more likely to report functional loss because the decline on
ADL increases with AD severity according to MMSE score
at baseline. It has already been shown that basic ADL are
more sensitive to change in patients presenting a more
severe stage of AD [17]. Because patients present a well-
preserved autonomy at baseline, this work from Green et al
[17] could also explain the relatively long duration of
follow-up before functional loss in the whole cohort (ap-
proximately 11 months) and that, according to the final
score of the ADL, patients are not very much impaired at 2
years. Dressing and bathing were the two items of the ADL
scale with the highest incidence of decline. Global change in
behavioral disorders was statistically significant after 2
years in the whole cohort. There was no significant differ-
ence in the NPI global evolution according to AD severity
at baseline. Annual incidence for a significant worsening
(ⱖ4-point total NPI score increase) [18] was almost 55%,
indicating that there is a significant worsening in these
behavioral disorders for many patients. The symptom for
which the incidence of aggravation was the most important
was apathy, which is a major problem because it arises early
and is persistent during the duration of AD [19]. The rela-
tionship between apathy and cognitive impairment contin-
ues to be discussed, with some studies that found a rela-
tionship between low MMSE score and apathy, whereas
other works did not report this relation [20]. Nutritional
status also showed a significant worsening at 2 years and did
not seem to be influenced by AD severity at baseline. The
impact of nutritional status on cognitive decline has already
been demonstrated in this cohort [12,21]. All these results
on cognitive, functional, behavioral, and nutritional status
reported a progression in the AD symptoms, and finally, the
evolution of the CDR-SB confirmed that AD worsened
significantly during the 2 years of follow-up in the whole
cohort and in each group obtained according to baseline
severity. Morris et al [13] in the CERAD study, with pa-
tients assessed every year and not specifically treated for
AD, reported that the rate of progression of dementia was
slower in patients with mild dementia than in those with
more advanced stage. In the same way, in the REAL-FR
cohort, AD evolution tended to be faster during the second
year in spite of treatments and specific management. Use of
AD-specific treatments seemed to slow AD progression but
did not modify the profile of evolution of the disease.
Looking at the incidences for stopping and changing these
drugs, it appears that they were very low, indicating that
these treatments were well-tolerated, and the long period of
treatment of roughly 19 months before stopping or changing
demonstrated that this tolerance was of long duration. It also
appears that the risk for changing or stopping these treat-
ments was not influenced by baseline severity. The specific
and regular follow-up setup for this cohort might explain
these low incidences. Almost one fourth of the participants
at the survey were hospitalized each year whatever their
baseline severity, more than one patient in 10 was institu-
tionalized, and one in 15 died. These results showed that in
spite of the maintenance of a good global status, rates for
hospitalization, institutionalization, and death were high. It
also appears that the rate of institutionalization was influ-
enced by AD severity at baseline, with an increased risk of
being institutionalized for more severe patients, whereas it
was not the case for death and hospitalizations.
In a previous study [22] we described that the evolution
of AD during the first year of follow-up appeared to be
slower after the marketing of ChEIs than before. Our data
seem to confirm this slowing, with the relatively long du-
ration of follow-up before evolution to severe dementia, loss
of one ability on ADL, and significant increase in behavioral
disorders (more than 1 year). The explanation for this phe-
nomenon could be a better follow-up as well as a potential
impact of the drugs.
Thanks to the inclusion criteria of the REAL-FR study
that are larger than those of clinical trials, results obtained in
this cohort are more representative of the general popula-
tion. However, some methodologic skews must be pointed
out. Patients have been included and followed in specialized
centers and could present different characteristics compared
with AD patients managed in nonspecific centers [23].
Moreover, the use of specific treatment in 85% of the cohort
might not be representative of the general use of these
treatments in France. This work allowed us to describe the
evolution of AD patients who benefit from specific man-
agement and AD treatments.
In conclusion, in the present study on a large population
sample we were able to show, after 2 years of follow-up, a
significant decline of AD patients’ conditions in terms of
cognitive, functional, behavioral, nutritional, and global sta-
tus. Results obtained show that the heterogeneity of the
evolution was large, with almost 25% of the patients cog-
nitively stable at 2 years and 11% who evolved twice as
rapidly as the mean of the whole cohort. Further analysis
will be important to compare data between slow and rapid
decliners to determine the characteristics or events that can
explain the occurrence of an episode of rapid decline. Ad-
justment by age of patients, duration of illness, duration of
specific AD treatment, educational level, living arrange-
ment, or other parameters and life events such as repeated
hospitalization and institutionalization would allow us to
know what characteristics can influence the rate of decline.
Such results would be very important to adapt the manage-
ment of patients at risk of rapid decline. Moreover, because
the ChEI treatments seem to provide benefits and because
we observe patients changing or stopping these treatments,
it would also be interesting to characterize individual pa-
rameters or life events that can be associated with these
 F. Cortes et al. / Alzheimer’s & Dementia 4 (2008) 22–29
events. Data obtained in the REAL-FR study will allow us
to work on these subjects.
Acknowledgements
This work was supported by a grant from the Clinical
Research Program Hospital from the French Ministry of
Health (PHRC No 98-47N, PHRC No 0101001).
REAL-FR group: principal investigator: Prof B. Vellas
(Toulouse); associate investigators: Pr M. Rainfray, Dr S.
Richard-Harston (Bordeaux); Pr P. Couturier, Pr A.
Franco (Grenoble); Pr F. Pasquier, Dr M. A. Mackowiak-
Cordoliani (Lille); Dr B. Frigard, Dr H. Idiri, Dr K. Gallouj
(Wasquehal); Dr B. Michel (Marseille); Pr Cl. Jeandel
(Montpellier); Pr J. Touchon, Dr F. Portet, Dr S. Lerouge
(Montpellier); Pr Ph. Robert, Dr P. Brocker, C. Bertogliati
(Nice); Pr B. Forette, Dr L. Teillet, Dr L. Lechowski (Paris);
Pr J. Belmin, D. S. Pariel-Madjlessi (Paris); Pr M. Verny,
Dr M. A. Artaz (Paris); Pr F. Forette, Dr A. S. Rigaud, Dr
F. Latour (Paris); Pr P. Jouanny, Dr S. Belliard, Dr O.
Michel (Rennes); Dr C. Girtanner, Dr Thomas- Anterion
(Saint Etienne); study coordinators: F. Cortes, S. Gillette-
Guyonnet, Prof. F. Nourhashemi, Dr P.J. Ousset (Tou-
louse); epidemiologist: Pr S. Andrieu; data processing: C.
Cantet.
References
[1] Courtney C, Farrell D, Gray R, Hills R, Lynch L, Sellwood E, et al.
AD 2000 Collaborative Group. long term donepezil treatment in 565
patients with Alzheimer’s disease (AD 2000)—randomised double
blind trial. Lancet 2004;363:2105–15.
[2] 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;34:939 – 44.
[3] American Psychiatric Association. Diagnostic and statistical manual
of mental disorders. 4th ed. Washington, DC: APA; 1994.
[4] Gillette-Guyonnet S, Nourhashemi F, Andrieu S, Cantet C, Micas M,
Ousset PJ, et al. REAL.FR group: the REAL.FR research program on
Alzheimer’s disease and its management: methods and preliminary
results. J Nutr Health Aging 2003;7:91– 6.
[5] Folstein MF, Folstein SE, McHugh PR. Mini-Mental State: a practical
method for grading the cognitive state of patients for the clinician.
J Psychiatr Res 1975;12:189 –98.
[6] Rosen WG, Mohs RC, Davis KL. A new rating scale for Alzheimer’s
disease. Am J Psychiatry 1984;141:1356 – 64.
29
[7] Katz S, Ford AB, Moskowitz RW, Jackson BA, Jaffe MW. The index
of ADL: A standardized measure of biological and psychosocial
function. JAMA 1963;185:914 –9.
[8] Cummings JL, Mega M, Gray K, Rosenberg-Thompson S, Carusi
DA, Gornbein J. The Neuropsychiatric Inventory: comprehensive
assessment of psychopathology in dementia. Neurology 1994;44:
2308 –14.
[9] Vellas B, Guigoz Y, Garry PJ, Nourhashemi F, Bennahum D, Lauque
S, et al. The Mini Nutritional Assessment (MNA) and its use in
grading the nutritional state of elderly patients. Nutrition 1999;2:
116 –22.
[10] Hughes CP, Berg L, Danziger WL, Coben LA, Martin RL. A new
clinical scale for the staging of dementia. Br J Psychiatry 1982;140:
566 –72.
[11] Mohs RC, Schmeidler J, Aryan M. Longitudinal studies of cognitive,
functional and behavioural changes in patients with Alzheimer’s
disease. Stat Med 2000;19:1401–9.
[12] Miliken JK, Edland SD. Mixed effect models of longitudinal Alzhei-
mer’s data: a cautionary note. Stat Med 2000;19:1617–29.
[13] Cortes F, Gillette-Guyonnet S, Nourhashemi F, Andrieu S, Cantet C,
Vellas B. Recent data on the natural history of Alzheimer’s disease:
results from the REAL.FR Study. J Nutr Health Aging 2005;9:86 –93.
[14] Dumont C, Voisin T, Nourhashemi F, Andrieu S, Koning M, Vellas
B. Predictive factors for rapid loss on the mini-mental state exami-
nation in Alzheimer’s disease. J Nutr Health Aging 2005;9:163–7.
[15] Morris JC, Edland S, Clark C, Galasko D, Koss E, Mohs R, et al. The
Consortium to Establish a Registry for Alzheimer’s Disease
(CERAD): part IV—rates of cognitive change in the longitudinal
assessment of probable Alzheimer’s disease. Neurology 1993;43:
2457– 65.
[16] Stern RG, Mohs RC, Davidson M, Schmeidler J, Silverman J,
Kramer-Ginsberg E, et al. A longitudinal study of Alzheimer’s dis-
ease: measurement, rate and predictors of cognitive deterioration.
Am J Psychiatry 1994;151:390 – 6.
[17] Green CR, Mohs RC, Schmeidler J, Aryan M, Davis KL. Functional
decline in Alzheimer’s disease: a longitudinal study. J Am Geriatr
Soc 1993;41:654 – 61.
[18] Mega MS, Masterman DM, O’Connor SM, Barclay TR, Cummings
JL. The spectrum of behavioral responses to cholinesterase inhibitor
therapy in Alzheimer’s disease. Arch Neurol 1999;56:1388 –93.
[19] Mega MS, Cummings JL, Fiorello T, Gornbein J. The spectrum of
behavioral changes in Alzheimer’s disease. Neurology 1996;46:
130 –5.
[20] Landes AM, Sperry SD, Strauss ME, Geldmacher DS. Apathy in
Alzheimer’s disease. J Am Geriatr Soc 2001;49:1700 –7.
[21] Vellas B, Lauque S, Gillette-Guyonnet S, Andrieu S, Cortes F,
Nourhashemi F, et al. Impact of nutritional status on the evolution of
Alzheimer’s disease and on response to acetylcholinesterase inhibitor
treatment. J Nutr Health Aging 2005;9:75– 80.
[22] Gillette-Guyonnet S, Andrieu S, Cortes F, Nourhashemi F, Cantet C,
Ousset PJ, et al. Outcome of Alzheimer’s disease: potential impact of
cholinesterase inhibitors. J Gerontol A Biol Sci Med Sci 2006;61:
516 –20.
[23] Massoud F, Devi G, Stern Y, Lawton A, Goldman JE, Liu Y, et al. A
clinicopathological comparison of community-based and clinic-based
cohorts of patients with dementia. Arch Neurol 1999;56:1368 –73.