It is estimated that more than half of elderly patients with dementia will develop some degree of behavioral and psychological symptoms of dementia (BPSD). BPSD is associated with significant morbidity, rapid functional decline, and psychiatric hospitalization, and there are currently no medications approved by the Food and Drug Administration (FDA) for its treatment. One treatment option that has been evaluated is the antiepileptic drug valproate. This retrospective medical record review evaluated the tolerability of valproate in elderly patients with dementia. A total of 62 patients met inclusion criteria for this review, which sought to determine whether there was an association either between the initial valproate dose and discontinuation or dose reduction (DCDR) or the maximum valproate dose achieved and DCDR. Both the total daily dose and total daily dose in mg/kg were assessed. For both the maximum daily dose (OR 1.0; 95% CI 0.99, 1.01) and maximum daily mg/kg dose (OR =1.15; 95% CI 0.92, 1.49) there was also no association with DCDR. There was no associated risk of DCDR when evaluating initial daily doses (OR: 1.0; 95% CI: 0.99 – 1.01). However, an association between DCDR and initial daily mg/kg dose was found (OR: 1.92; 95% CI: 1.11–4.02). For both the initial dose and initial mg/kg dose model, African Americans were associated with the need for DCDR (OR 20.75; 95% CI: 1.77–660). Results from this study suggest that higher initial starting doses based on weight may lead to higher rates of side effects necessitating a discontinuation or dose reduction; however, large trials are needed to confirm these results.

It is estimated that greater than 50% of elderly patients with dementia will develop some degree of behavioral and psychological symptoms of dementia (BPSD), which is associated with significant morbidity, rapid functional decline, and psychiatric hospitalization.1,2 There are no Food and Drug Administration (FDA) approved medications for the BPSD. A common treatment option includes the use of low-dose antipsychotics. These should be used judiciously as the FDA has issued a boxed warning due to findings of increased mortality in elderly with dementia.3 Antiepileptic drugs (AEDs), including valproate, have also been evaluated in the treatment of BPSD. The literature to support valproate use is minimal and hypotheses surrounding why valproate may be beneficial in BPSD are related to the GABA agonism, stimulation of the serotonergic system, mediation limbic dysfunction, and the inhibition of corticotropin-releasing hormone.1,2,4 However, due to the lack of well conducted studies, a Cochrane review concluded that there is no evidence showing efficacy of valproate preparations for treatment of BPSD.5 Also, a recent retrospective cohort study evaluating elderly patients with dementia found the associated mortality risk with valproate to be similar to risperidone and higher than quetiapine. Therefore, valproate was used as a non-antipsychotic comparator in this study.5 

Dosing recommendations of valproate for BPSD may be extrapolated from small trials. However, the mean dosage in these trials vary greatly (134 mg to 1000 mg).6–22 If risk outweighs benefit, it may be advisable to initiate valproate in the elderly at 250 mg daily and titrate to effect or to a serum concentration of 50–90 mcg/mL as some literature suggests. However, serum concentrations often do not correlate to therapeutic effect in the elderly.17,19 

Initiation of any medication in the elderly should be done with caution due to increased susceptibility to side effects. For valproate, these include sedation, gait disturbances, and tremor.8 Other risks include thrombocytopenia, hyperammonemia, pancreatitis, and liver injury. To assess the dosing and tolerability of valproate in elderly patients with dementia, its use in 2 psychiatric units of a large urban teaching hospital in Richmond, Virginia was evaluated.

This was an Institutional Review Board (IRB) approved retrospective medical record review. Information was obtained via electronic medical records for patients meeting the inclusion and exclusion criteria with a psychiatric admission between January 1, 2005 and December 31, 2008. Inclusion criteria included an age greater than 60, a diagnosis of any dementia type, and having received at least one dose of valproate during a psychiatric admission. Exclusion criteria included a diagnosis of bipolar disorder, primary psychotic disorder, impulse control disorder, seizure disorder, concomitant AED use, valproate use upon admission, and concomitant medications known to cause clinically significant drug-drug interactions with valproate.

To detect tolerability of valproate in patients with dementia, medical records were evaluated for discontinuations or dose reductions and the reason for the medication change. Other data assessed included age, sex, race, weight, serum valproate levels, and concomitant psychotropic use.

Data obtained were entered into a Microsoft™ Access database. Using JMP 10.0, a logistic regression model adjusted for age, race, gender, and weight was used to predict whether there was an association either between the initial valproate dose and discontinuation or dose reduction (DCDR) or the maximum valproate dose achieved and DCDR. The same was conducted for the initial and maximum dose per kilogram of body weight (mg/kg). All doses are reported as total daily doses. Outputs from the model were reported as odds ratios (OR) with 95% confidence intervals (CI).

Two hundred-fourteen patients were identified for any dementia type by ICD-9 codes. Of these patients, 151 patients did not meet the criteria for inclusion and 1 patient was excluded because of incomplete data. The most common reason for exclusion was a comorbid Axis I diagnosis. The mean age of all patients included (n=62) was 78 + 8.1 years. Males accounted for 53% of the population and Caucasians made up 70% of the cohort. Patients had a mean weight, starting dose, and initial mg/kg dose of 71 + 12.6 kg, 409 mg + 150.4 mg, and 5.96 mg/kg + 2.1 mg/kg, respectively. The mean maximum dose and maximum mg/kg dose achieved were 554.4 mg + 255.7 mg and 8.0 mg/kg + 3.8 mg/kg, respectively. Concomitant medication use was similar between those that had a DCDR and those that did not. These medications included oral anxiolytics, antipsychotics, antidepressants, and cholinesterase-inhibitors/memantine. The mean valproate concentration was 37.9 + 15.8 mcg/mL in the 42 patients in which it was obtained. There were 2 discontinuations due to sedation and gait disturbance. One patient required a dose reduction for sedation and another for gait disturbance. One patient had a dose reduction but the reason could not be elucidated from the medical record.

For both the maximum daily dose (OR 1.0; 95% CI 0.99, 1.01) and maximum daily mg/kg dose (OR =1.15; 95% CI 0.92, 1.49) there was also no association with DCDR. There was no associated risk of DCDR when evaluating initial daily doses (OR: 1.0; 95% CI: 0.99 – 1.01). However, an association between DCDR and initial daily mg/kg dose was found (OR: 1.92; 95% CI: 1.11–4.02). For both the initial dose and initial mg/kg dose model, African Americans were associated with the need for DCDR (OR 20.75; 95% CI: 1.77–660).

The primary objective of this study aimed to evaluate an elderly population with dementia admitted to a psychiatric unit and assess their ability to tolerate valproate therapy. Current literature suggests a “start low and go slow” approach without a clear target dose or serum concentration that should be achieved. Also no trials discussed dosing in a weight-based manner which may have an important impact on many patients. Patients who had a DCDR on average weighed more than those that did not have a DCDR. These patients may have been perceived as being able to tolerate more medication, resulting in doses that were 1–3 standard deviations from the mean.

An association of African American race and DCDR was also detected by models that evaluated initial dosing and may not be of clinical significance. Interestingly it has recently been discovered that a partial metabolic pathway for valproate polymorphism (CYP2A6*20 allele) results in decreased enzymatic activity and is specific to African Americans. Although allele frequency is only predicted to be 1.6% (95% CI: 0.6–4.5), other CYP polymorphisms that are either specific to African Americans or occur at higher frequencies may allow for increased medication accumulation.23 Yet, elderly have increasingly decreased CYP activity over time and because phase I metabolism does not account the majority of valproate's metabolism, future studies would need to determine if CYP polymorphisms in relation to valproate are clinically significant.

A small sample size may have resulted in the non-significant findings and further well conducted studies are needed to evaluate if other dosing strategies are not related to discontinuation or reductions in valproate dosing. Additionally, a true assessment of tolerability may be limited in this study as maximum doses and associated serum concentrations achieved in this cohort of patients may be considered as low. No patients developed thrombocytopenia, detected hyperammonemia, acute liver injury, or pancreatitis during their admission. However, the average length of stay was only 17.8 days, limiting the ability to detect these types of events. Lastly, discontinuation and dose reduction were used as markers of tolerability. It is possible that patients experienced a side effect with no DCDR which could not be identified retrospectively.

Results from this study suggest that higher initial starting doses based on weight may lead to higher rates of side effects necessitating a discontinuation or dose reduction. Large trials are needed to confirm these results. There were no discontinuations or reductions of valproate due to severe or life threatening side effects in this study.

1.
Pratt
CE
,
Davis
SM.
Divalproex sodium therapy in elderly with dementia-related agitation
.
Ann Pharmacother
.
2002
;
36
(
10
):
1625
8
.
PubMed PMID: 12243614.
2.
Lott
AD
,
McElroy
SL
,
Keys
MA.
Valproate in the treatment of behavioral agitation in elderly patients with dementia
.
J Neuropsychiatry Clin Neurosci
.
1995
;
7
(
3
):
314
9
.
PubMed PMID: 7580190.
3.
Antipsychotics
.
FDA Alert. Available http://www.fda.gov/Drugs/DrugSafety. Accessed August 15, 2009
.
4.
Lonergan
E
,
Luxenberg
J.
Valproate preparations for agitation in dementia
.
Cochrane Database Syst Rev. Chichester, UK
;
2009
;(
3
):
CD003945
. .
5.
Kales
HC
,
Kim
HM
,
Zivin
K
,
Valenstein
M
,
Seyfried
LS
,
Chiang
C
et al.
Risk of mortality among individual antipsychotics in patients with dementia
.
Am J Psychiatry
.
2012
;
169
(
1
):
71
9
.
6.
Mizukami
K
,
Hatanaka
K
,
Ishii
T
,
Iwakiri
M
,
Sodeyama
N
,
Tanaka
Y
et al.
Effects of sodium valproate on behavioral disturbances in elderly outpatients with dementia
.
Geriatr Gerontol Int
.
2010
;
10
(
4
):
324
6
.
7.
Amann
B
,
Pantel
J
,
Grunze
H
,
Vieta
E
,
Colom
F
,
Gonzalez-Pinto
A
et al.
Anticonvulsants in the treatment of aggression in the demented elderly: an update
.
Clin Pract Epidemiol Ment HealthClinical Practice and Epidemiology in Mental Health
.
2009
;
5
:
14
.
DOI: 10.1186/1745-0179-5-14. PubMed PMID: 19531220; PubMed Central PMCID: PMC2704187
.
8.
Divalproex sodium
.
Lexi-Comp Online. Available http://www.lexionline.com. Accessed August 15, 2009
.
9.
Porsteinsson
AP
,
Tariot
PN
,
Erb
R
,
Cox
C
,
Smith
E
et al.
Placebo-controlled study of divalproex sodium for agitation in dementia
.
Am J Geriatr Psychiatry
.
2001
;
9
(
1
):
58
66
.
10.
Tariot
PN
,
Raman
R
,
Jakimovich
L
,
Schneider
L
,
Porsteinsson
A
,
Thomas
R
et al.
Divalproex sodium in nursing home residents with possible or probable Alzheimer disease complicated by agitation
.
Am J Geriatr Psychiatry
.
2005
;
13
(
11
):
942
49
.
11.
Porsteinsson
AP
,
Tariot
PN
,
Erb
R
,
Gaile
S.
An open trial of valproate for agitation in geriatric neuropsychiatric disorders
.
Am J Geriatr Psychiatry
.
1997
;
5
(
4
):
344
351
.
12.
Forester
B
,
Vanelli
M
,
Hyde
J
,
Perez
R
,
Ahokpossi
C
,
Sribney
W
et al.
Report on an open-label prospective study of divalproex sodium for the behavioral and psychological symptoms of dementia as monotherapy and in combination with second-generation antipsychotic medication
.
Am J Geriatr Pharmacother
.
2007
;
5
(
3
):
209
17
.
13.
Felix
S
,
Sproule
BA
,
Hardy
BG
,
Naranjo
CA.
Dose-related pharmacokinetics and pharmacodynamics of valproate in the elderly
.
J Clin Psychopharmacol
.
2003
;
23
(
5
):
471
8
.
14.
Birnbaum
AK
,
Ahn
JE
,
Brundage
RC
,
Hardie
NA
,
Conway
JM
,
Leppik
IE.
Population pharmacokinetics of valproic acid concentrations in elderly nursing home residents
.
Therapeutic Drug Monitoring
.
2007
;
29
(
5
):
571
5
.
15.
Narayan
M
,
Nelson
JC.
Treatment of dementia with behavioral disturbance using divalproex or a combination of divalproex and a neuroleptic
.
J Clin Psychiatry
.
1997
;
58
(
8
):
351
4
.
PubMed PMID: 9515973.
16.
Kasckow
JW
,
McElroy
SL
,
Cameron
RL
,
Mahler
LL
,
Fudala
SJ.
A pilot study on the use of divalproex sodium in the treatment of behavioral agitation in elderly patients with dementia: assessment with the behave-ad and CGI rating scales
.
Current Therapeutic Research
.
1997
;
58
(
12
):
981
989
.
17.
Herrmann
N.
Valproic acid treatment of agitation in dementia
.
Can J Psychiatry
.
1998
;
43
(
1
):
69
72
.
PubMed PMID: 9494750.
18.
Tariot
PN.
Treatment of agitation in dementia
.
J Clin Psychiatry
.
1999
;
60
Suppl 8
:
11
20
.
PubMed PMID: 10335667
.
19.
Zaraa
AS.
Dementia update. Pharmacologic management of agitation and psychosis in older demented patients
.
Geriatrics
.
2003
;
58
(
10
):
48
50
,
52
3
.
PubMed PMID: 14569642
.
20.
Lesser
JM
,
Hughes
S.
Psychosis-related disturbances. Psychosis, agitation, and disinhibition in Alzheimer's disease: definitions and treatment options
.
Geriatrics
.
2006
;
61
(
12
):
14
20
.
PubMed PMID: 17184138.
21.
Tariot
PN.
Valproate use in neuropsychiatric disorders in the elderly
.
Psychopharmacol Bull
.
2003
;
37
Suppl 2
:
116
28
.
PubMed PMID: 15021866
.
22.
Kunik
ME
,
Puryear
L
,
Orengo
CA
,
Molinari
V
,
Workman
RH
Jr.
The efficacy and tolerability of divalproex sodium in elderly demented patients with behavioral disturbances
.
Int J Geriatr Psychiatry
.
1998
;
13
(
1
):
29
34
.
23.
Fukami
T
,
Nakajima
M
,
Higashi
E
,
Yamanaka
H
,
McLeod
HL
,
Yokoi
T.
A novel CYP2A6*20 allele found in African-American population produces a truncated protein lacking enzymatic activity
.
Biochem Pharmacol
.
2005
;
70
(
5
):
801
8
.