Factors influencing diazepam kinetics were assessed in 4 equal groups (n = 11) ... diazepam (intrinsic clearance) tended to be higher in the females than in the ...
Diazepam disposition determinants Factors influencing diazepam kinetics were assessed in 4 equal groups (n = 11) of young male and female (aged 21 to 37 yr) and elderly male and female (aged 61 to 84 yr) subjects, all of whom were healthy. In all 44. plasma diazepam concentrations were determined by electron-capture gas-liquid chromatography in multiple samples drawn for as long as 9 days after a single 5- to JO-mg intravenous dose. Based upon total (bound + free) diazepam concentrations, volume of distribution (V d) ranged from 0.7 to 4.7 llkg, and became larger both with increasing age and with female sex. Clearances of total (bound + free) diazepam in young and elderly females were nearly identical (0.51 and 0.48 ml/min/kg), but clearance was higher in young than elderly males (0.39 and 0.24 mllminlkg, p < 0.01). The unbound fraction of diazepam in plasma (range, 0.9% to 2.7%) did not depend on sex. but was greater in the elderly than in the young. In part this related to lower plasma albumin concentrations in the elderly. After correction of kinetic data for individual differences in free fraction. V d was larger in the females than in the males. but the effect of age was small. Clearance of unbound diazepam (intrinsic clearance) tended to be higher in the females than in the males of both age groups. and was higher in the young than in the elderly of both sexes (male: 29.9 and 14.9 mllminlkg, p < 0.005; female: 43.6 and 28.0 ml/min/kg, p < 0.05). Smoking was associated with higher clearance values. particularly among young subjects.
David J. Greenblatt, M.D., Marcia Divoll Allen, R.N., Jerold S. Harmatz, B.A., and Richard I. Shader, M.D. Boston, Mass. Clinical Pharmacology Unit. Massachusetts General Hospital; Psychopharmacology Research
Laboratory. Massachusetts Mental Health Center; and the Division of Clinical Pharmacology. Departments of Psychiatry and Medicine, Tufts-New England Medical Center
The effect of age on drug disposition depends on the particular compound in question and the characteristics of the study population. 10, 12.25,32 Diazepam, a benzodiazepine, is used extensive-
Supported in part by Grant MH-12279 from the United States Public Health Service and Grant 77-611 from the Foundations' Fund for Research in Psychiatry. Presented in part at the Eightieth Annual Meeting of the American Society for Clinical Pharmacology and Therapeutics. Kansas City. Mo.• March 23. 1979. Received for publication June 8. 1979, Accepted for publication June 14. 1979, Reprint requests to: Dr, David J, Greenblatt. Division of Clinical Pharmacology, Box 1007, New England Medical Center. 171 Harrison Ave., Boston. MA 02111.
0009-9236/80/030301 + 12$01.20/0
ly by the elderly, and clinical reports suggest increased sensitivity of older persons to its sedative actions. 2, 6. 27 Studies have assessed whether age-related changes in diazepam kinetics could explain this phenomenon": 21 but the use of small sample sizes or failure to consider such factors as sex, protein binding, or cigarette smoking have left interpretation of results uncertain. Our study was designed to assess this problem more thoroughly. Methods
Subjects. Forty-four healthy subjects participated in the study (Table I). They were di-
© 1980 The C. V. Mosby Co.
301
302
Greenblatt et al.
Clin. Pharmacol. Ther. March 1980
Table I. Subject characteristics and kinetic variables for total and unbound diazepam Mean (with range) values for:
Subject characteristics Age (yr) Weight (kg) Plasma albumin concentration (gm/ I00 ml) Smoking habits (cigarettes/day) No. of subjects taking other drugs (yes/no) Kinetics variables for total (free Volume of distribution (l/kg) Elimination half-life (hr) Clearance (mllmin/kg)
Young male
Young female
Elderly male
Elderly female
27.7 (23-37) 7\.7 (52.3-79.5) 4.65 (4.4-5.1)
28.3 (21-32) 56.8 (49.1-70.0) 4.36 (3.2-5.0)
68.9 (63-76) 87.9 (64.5-118.2) 4.18 (3.6-4.7)
68.9 (61-84 ) 6\.1 (45.4-83.6) 4.26 (3.7-4.7)
13.3* (0-50) 0/11
5.1 (0-30) 0/11
\.8 (0-20) 3/8
2.7 (0-15) 2/9
1.73 (1.34-2.14) 42.4 (26.6-63.0) 0.51 (0.35-0.93)
1.83 (0.92-2.88) 98.5 (50.2-173.3) 0.24 (0.11-0.34)
2.64 (1.20-4.66) 7\.8 (31.9-198) 0.48 (0.27-0.81)
1.23 (0.90-1.55) 145.2 (91-188) 43.6 (22.4-76.7)
1.66 (\.1-2.65) 114.9 (73.0-184) 14.9 (7.0-25.4)
\.72 (1.5-2.0) 154.6 (75-255) 28.0 (13.9-45.1)
+ bound)
diazepam \.11 (0.70-1.78) 36.0 (20.4-66.7) 0.39 (0.28-0.77)
Kinetic variables for unbound diazepam 1.34 Unbound (%) ( 1.0-1.75) Volume of distribution 87.3 (44-178) (l/kg) Clearance (mllmin/kg) 29.9 (18.3-56.9)
'Smoking data were not included for one young male subject who smoked a nontobacco preparation.
vided into 4 equal groups of 11 representing young male, young female, elderly male, and elderly female. All young subjects were free of medical disease and were taking no medications. The elderly subjects were ambulatory, active, and in good general health; 5 were taking other medications. One elderly man had mild essential hypertension controlled with a diuretic, one had a prior episode of atrial fibrillation and was on digoxin, and one had osteoarthritis for which he took aspirin. Among the elderly females, one with atrial fibrillation took digoxin, and another with mild maturity-onset diabetes took chlorpropamide. The mean age of both elderly groups was 69 yr; for the young it was 28 yr (Table I). Elderly subjects tended to be heavier and to smoke fewer cigarettes than the young. Consistent with
a previous report, the elderly had lower plasma albumin concentrations than the young." Procedure. All subjects received 5 to 10 mg diazepam by intravenous infusion over 15 to 30 sec. The dose was approximately adjusted for body weight. Venous blood samples were drawn into heparinized tubes before the dose, and thereafter at 5, 15, 30, 45 min, 1 hr, 1.5, 2.0,2.5,3,4,6,8, 12, and 24 hr. Thereafter, samples were taken every 1 to 2 days for 5 to 9 days. An indwelling Butterfly cannula, kept patent by flushing with a dilute solution of heparin in 5% dextrose injection (50 Vlml), was used to obtain blood samples during the first 8 hr after dosing. All other samples were drawn by separate venipuncture. Blood specimens were centrifuged, and the plasma was separated and stored at -20 until assayed. 0
Diazepam disposition determinants
Volume 27 Number 3
Value of Student's t for: Young male
Young female
vs
vs
elderly male
elderly female
trations after intravenous infusion were analyzed by weighted iterative nonlinear leastsquares regression techniques. 13, 22, 31 Data points were fitted to the following 2 functions: C = Ae- at + Be-Ilt C = Ae- al
17.21
23.74 (p
< 0.001)
(p
1.16
2.88 (p
< 0.01)
(p
< 0.005)
(NS)
0.598
3.73
(NS)
0.707
1.99
(NS)
(NS)
2.91
3.77 (p
< 0.005)
(p
< 0.001)
(p
< 0.01)
(p
< 0.01)
(p
< 0.05)
2.13
4.91
0.458
3.15
(NS)
5.45
1.93 (NS)
(p
(NS)
2.35
3.81
< 0.005)
< 0.001) 0.475
1.78 (NS) (p
< 0.001)
(p
< 0.05)
Analysis of samples. Concentrations of diazepam and its major metabolite, desmethyldiazepam, were determined by electron-capture gasliquid chromatography. 7 The extent of diazepam binding to plasma protein for each subject was determined in a single sample drawn in the nonfasting state at least 72 hr after the dose, such that the patient had received no heparin systemically for at least 60 hr. Binding was determined by an equilibrium dialysis technique" using duplicate 2-ml plasma samples spiked to contain 500 ng diazepam/ml. Mean recovery of diazepam from the dialysis system was 100%, and the mean coefficient of variation for 118 replicate samples was 5.4%. Binding was independent of total plasma diazepam concentration up to levels of at least 5 j.tg/ml. Analysis of data. Plasma diazepam concen-
303
+ Pe ? " + Be-Ill
(I)
(2)
In both of these equations, C is the plasma diazepam concentration at time t after the dose. A, P, and B are hybrid coefficients, and a, 7T, and {3 are hybrid exponents, which can be related to "micro" rate constants for specific 2- or 3-compartment kinetic models, 5, 11, 34 The choice between equations 1 and 2 as function of best fit was determined by comparison of sum of squares of weighted residual errors, and by the scatter of actual data points about the fitted function." Coefficients and exponents from the fitted function were used to calculate the following kinetic variables: total apparent volume of distribution using the area method (V d), apparent elimination half-life (tss), and total clearance.": 11,34 Assuming that binding of diazepam to plasma protein for any given subject was constant over time and independent of total concentration, values of Vd and clearance for each subject were divided by that subject's unbound fraction yielding values of unbound Vd and unbound (intrinsic) clearance. Differences in kinetic variables between young and elderly groups of both sexes were determined by Student's t test. In addition, the simultaneous influence of several identifiable independent factors (age, sex, plasma albumin concentration, smoking, coingestion of other drugs, and extent of protein binding) on kinetic variables for total and unbound diazepam were determined by multiple regression analysis." Results
In all subjects, elimination of diazepam from plasma after intravenous injection was satisfactorily described by either equation 1 or 2. Diazepam disappearance was always mirrored by appearance of its pharmacologically active metabolite, desmethyldiazepam (Figs. 1 and 2). Because the rate of desmethyldiazepam elimination was considerably slower than of the parent compound, the duration of sampling did not
304
Greenblatt et al.
Clin. Pharmacal. Ther. March 1980
400
----0;
-------
0;
0lA2EPAIol DESIolETHYLDIAZEPAIol
200
100
50
20
10
~al.,
z
24 yeors
o
i= -c
Q::
f-
Z
LoJ
U Z
300
o
u
«
~
(/)
«
100
...J
c,
50
20
'9--------
...... .8-
--------------------0
10
Mal., 69 years
DAYS AFTER DOSE
Fig. 1. Plasma concentrations of diazepam and its major metabolite, desmethyldiazepam, in representative young and elderly male subjects following intravenous injection of diazepam. Also shown are kinetic functions for diazepam. Note the slower disappearance of diazepam, as well as the slower rate of metabolite formation, in the elderly subject.
permit formal analysis of desmethyldiazepam kinetics. Kinetic variables for total (bound + free) diazepam were influenced by both age and sex. Values of V d were larger in females than males of both age groups. Vd was larger in the elderly than in the young of both sexes (Table I, Fig. 3). The t,/,,8 was longer in the elderly than in the young of the same sex (Fig. 4). Total clearance of diazepam in elderly males was less than in young males but mean diazepam clearance in the 2 female groups was nearly identical (Table I, Figs. I, 2, and 5). Cigarette smoking appeared to influence diazepam clearance in the young group of both sexes, inasmuch as those
with the highest clearance values were heavy cigarette smokers (Fig. 5). These observations were confirmed by multiple regression analysis (Table II). The 6 independent variables explained 47% to 49% of the overall variability in the 3 kinetic variables. Standardized regression coefficients (beta weights) indicated that sex was a more important determinant of Vd and clearance than age. Cigarette smoking had an important influence on clearance: higher values were associated with heavier cigarette smoking. In all subjects, diazepam was extensively bound to plasma protein. The free fraction ranged from 0.9% to 2.7% (Table I, Fig. 6).
Volume 27 Number 3
Diazepam disposition determinants
305
• =DIAZEPAM
500
0=
DESMETHYLDIAZEPAM
200
100
50
20
10
Female. 29 years
z o I« 0::
I-
Z W
800
U
Z
o
500
u
«
:::;;
200
(/)
«
--J
0...
100
50
20
10
t 0
4
5
0
8
DAYS AFTER DOSE
Fig. 2. Plasma concentrations of diazepam and its major metabolite, desmethyldiazepam, in representative young and elderly female subjects following intravenous injection of diazepam. Also shown are kinetic functions for diazepam.
Sex did not appear to influence binding, but unbound fraction correlated positively with age (r = 0.55) and was higher in elderly than in young subjects of both sexes. In the case of females, the age effect was significant (p < 0.001). The relation of age to diazepam binding could be partly explained by plasma albumin concentration, which correlated negatively both with age (r = -0.36) and with unbound fraction (r = -0.25) (Fig. 7). Multiple regression analysis, however, indicated that age was the major single determinant of binding (Table II), with sex, plasma albumin, and other drugs contributing much less. In any case, a relatively small proportion of over-all variability in un-
bound fraction was accounted for (multiple = 0.32). Correction of kinetic variables for individual differences in unbound fraction indicates the importance of such differences in the interpretation of kinetic data. Unbound Vd was larger in females than in males of both age groups, but the effect of age was small (Fig. 8). Multiple regression analysis indicated that sex was the most important single determinant (Table II). Clearance of unbound diazepam (intrinsic clearance) tended to be greater in females than in males of both age groups (Table I, Fig. 9). Intrinsic clearance in elderly males was 50% less than in young males (p < 0.005)-an even
R2
306
Greenblatt et al.
Clin. Pharmacal. Ther. March 1980
YOUNG
5.0
ELDERLY o
s:
~
4.0
CD
a::
V)
o u...
•
o
w
000
- -- --.~ - - - -
-I
>
~~~~~3~~:---o
•
::E ::>
o
00 0
o
00
o
- - - - - iI- - - - - ----
-.-•••
o - - - --
•
-----~------
1.0
0.0
~-----------
•
/!,
2.0
o
••
l-
••• MALE FEMALE FEMALE MALE +-------,-------+------.------------1
Fig. 3. Volume of distribution for total (free + bound) diazepam. Individual and mean (±SE) values for each of the 4 subject groups are shown. Circles represent nonsmokers or smokers of no more than 5 cigarettes/day; triangles represent smokers of 10 or more cigarettes/day.
YOUNG
ELDERLY
200.0
o
•
175.0
~
(Il
.....
150.0
••
:::J 0
..r:
'-"
w
125.0
u, _J
FEMALE
MALE
I
u, -I
100.0
I
••• -- - - - iI- - - - --
Z
0
/!,
75.0
I-
•
-c Z
~
-I
w
50.0
____4 - - --
25.0
o
.o.
-.-
....
o.
00
_
- - - --
•
o
• • •
- - - - - - - -0- - - 00 0
o
000 - - - - .QQ- - - -/!,
o o
/!,/!,
MALE
FEMALE
0.0
Fig. 4. Values of diazepam elimination half-life. See Fig. 3 for explanation of symbols.
Diazepam disposition determinants
Volume 27 Number 3
YOUNG
ELDERLY
1.0
0.8
00
o
E E
6
•
a
CD
z
::::>
1.5
f-
Z
W
-.•
a:: 1.0
o
-----
-
- - - - - It
_
•
U
w a...
• • •-
-
•••
--
0
a
- - - - - -8- - - --
o
--
-
• - -.- - - -•• •
000
o
_____ 0
_
000 000
•••
•
o
0
0.5
MALE
0.0
FEMALE
MALE
FEMALE
+-------,-------+--------.-------1
Fig. 6. Extent of diazepam binding to plasma protein. Individual and mean (±SE) values for each of the 4 groups are shown (no distinction was made between smokers and nonsmokers).
307
308
Greenblatt et al.
Clin. Pharmacal. Ther. March 1980
• = YOUNG MALE o = YOUNG FEMALE • = ELDERL Y MALE
3.0
+
= ELDERL Y
FEMALE
2.5 ~
4: o,
•
w
N
4:
Ci
2.0
•
0
z
r=-O.25
::> 0
+
•
+
CD
Z
::>
+
1.5
+
t-
•
Z
w
+
0
•
0
o,
• +
()
0::: W
+
•
• +
o
+
0
• 0 0
•
.0 0
•
1.0
+
• • • • •
• 0
•••
0
0
0.5 + - - - - - - , - - - - - - - , - - - - - - - , - - - - - - - , - - - - - - - - - , 3.0 3.5 4.0 4.5 5.0 5.5
PLASMA ALBUMIN CONCENTRATION (gm/100 ml)
Fig. 7. Relation between plasma albumin concentration and diazepam protein-binding. Solid line was determined by linear regression analysis (r = -0.25).
Table II. Multiple regression analysis Standardized regression coefficients for independent variables Dependent variables
Age
I
Sex *
Plasma albumin
I
I
Smoking t
For total (free + bound) diazepam Volume of distribution 0.33 0.50 Elimination half-life -0.28 Clearance
-0.45 0.12 -0.53
-0.13 0.1 I -0.12
0.06 -0.15 0.36
For unbound diazepam Fraction unbound Volume of distribution Clearance
0.03 0.51 -0.47
-0.06
-§
0.48 0.18 -0.42
0.02 0.31
I
Ingestion of other drugs's 0.14 0.31 -0.10 0.14
I
Fraction unbound
Multiple R2
0.12 -0.06 O. I I
0.48 0.49 0.47 0.32 0.29 0.52
-Entered as 0 for female, I for male. tEntered as average number of cigarettes/day. tEntered as 0 for no, I for yes. §Dash (-) indicates that the variable was not entered into the analysis.
greater difference than for clearance of total (bound + free) drug (Fig. 9). Clearance of unbound diazepam in elderly females was likewise lower than in young females: the age effect in women just attained significance (p < 0.05)
and was considerably less than in men. Cigarette smoking had an important influence on intrinsic clearance in the young (Fig. 9), inasmuch as those with higher clearance were heavy cigarette smokers. Multiple regression analysis
Volume 27 Number 3
Diazepam disposition determinants
YOUNG
309
ELDERLY
300.0
"""' (Jl ~
....J
0
-
- - -~ ~_ ... - - --
......•
0
>
----
•
0 00
----
•
0
0 0
-----------0 000
------------
• ••• ~
~
Cl
Z
:::> 0
•• ---8-------
100.0
0
0006
V1
0
•
6
:::>
50.0
lD
z
•
• MALE
:::>
MALE
FEMALE
FEMALE
0.0
Fig. 8. Volume of distribution for unbound diazepam. See Fig. 3 for explanation of symbols.
indicated that the effects of age and sex on intrinsic clearance were approximately equal, with higher values of clearance associated with younger age and female sex (Table II). Smoking also contributed importantly. Of the overall variability in intrinsic clearance, 52% was explained by these 3 factors. Discussion
Previous studies on the effect of age on diazepam kinetics have not been conclusive. Klotz et al. 19 reported that age was associated with larger values of V d and tV'/l. Although clearance correlated negatively with age (r = -0.39), the association did not attain significance (p = 0.089). There was no apparent effect of smoking on diazepam clearance. Furthermore, age did not appear to influence plasma protein binding. In that study, most but not all subjects were male, and the sexes were not treated separately. MacLeod et aPl in a study of 19 young and elderly subjects of both sexes, found that women had lower clearance of total diazepam than men but that age did not appear to influ-
ence clearance. The effects of cigarette smoking, serum albumin concentration, protein binding, and underlying medical disease were not assessed. Our study suggests that several factors may have influenced the outcome of the above studies, thereby contributing to apparent differences among laboratories. Assessment of kinetic variables for total (free + bound) diazepam indicates that both age and sex have an important influence, and that age effects may differ with sex. Diazepam distribution was more extensive in females than in males of all ages and in the elderly than in the young regardless of sex. Clearance of diazepam was significantly lower in elderly than in young males, whereas no age effect was observed in females. Since diazepam is extensively bound to plasma protein, findings based upon total diazepam concentrations in plasma reflect the extent of binding as well as drug distribution and clearance, and can be strongly influenced by small differences in unbound fraction between individuals. Since only the free fraction is available for diffusion out of the vascular sys-
310
Greenblatt et al.
Clin. Pharmacal. Ther. March 1980
YOUNG
ELDERLY
80.0 [:, ~
(J)
.Y
