European Journal of Clinical Pharmacology © by Springer-Verlag 1977
Europ. J. clin. Pharmacol. 11,443-448 (1977)
Effects of Halazepam and Diazepam on the Motor Coordination of Geriatric Subjects* M.-A. Gagnon, Y. Langlois, D. R. Boghen, and M. Verdy Institut National de la Recherche Scientifique (INRS-Sant6), Louis-H. Lafontaine Hospital Montreal and H6tel-Dieu de Montr6al, Montreal, Quebec, Canada
Summary. Two doses of halazepam, 20 and 40 mg, were compared to diazepam 5 mg and placebo in this double-blind, parallel experiment. The medications were administered thrice daily during the first 3 days, and twice daily during the following 11 days. Their effects were measured on the tandem walk and manual dexterity of 59 elderly female subjects. Most volunteers had some walking difficulty prior to the study and 21 suffered from some illness. With the higher dose of halazepam a statistically and clinically significant impairment of motor coordination was observed. Halazepam 40 mg, at the regimen above mentioned, should therefore not be used in elderly persons. This is substantiated by the results on all pertinent variables: tandem walk, manual dexterity, side effects and compliance to the drug regimen; indeed only 5 subjects out of 15 could tolerate halazepam 40 mg twice daily. With the manual dexterity test, significant effects of diazepam 5 mg and halazepam 20 mg were also observed. Furthermore, this test as well as the analysis of the side effects, shows that the lower dose of halazepam is well tolerated when compared to diazepam 5 mg.
Key words: Halazepam, gait, manual dexterity, motor coordination, geriatry.
Benzodiazepines are well known for producing ataxia, especially in older people [4]. In the course of early clinical trials on halazepam, a new benzodiazepine, the present study was set up in order to investigate the safety of the drug rather than its efficacy; more precisely, the purpose of the experiment was to establish a safe dose range in older subjects * Supported in part by Schering Corporation
with regard to its effects on motor coordination. These were estimated by the assessment of gait and manual dexterity. The measurement of side effects and drug regimen compliance yielded additional evidence on the hazards of halazepam high dosage.
Methods
Subjects Fifty-nine female members of a religious order, volunteered and were accepted for this experiment after giving an informed consent. Their age varied from 57 to 88 years (average: 69.5) and their weight from 40.9 to 79.5 kg (average: 60.3). Although 21 suffered from some illness and 21 were taking some medication prior to the study (Table 1), they were all judged, after a medical examination by their clinician, to be fit for the experiment. Volunteers with a renal or hepatic pathology, which could hinder the metabolism or excretion of the drugs, were excluded from the study, as well as those suspected of benzodiazepine hypersensitivity. Most subjects had some walking difficulty, related to osteoarthritis or cerebral arteriosclerosis, yet all were able to perform the tandem walk and manual dexterity tests within a 20-min training period.
Drugs Two doses of halazepam, 20 and 40 rag, were compared to diazepam 5 mg and placebo. The medications were contained in identical capsules. These were administered thrice daily, at 7.00, 15.00 and 22.00 h, during the first 3 days of the experiment; the afternoon dose was dropped afterwards. No other psychotropic drugs were taken during the ex-
444
M.-A. Gagnon et al.: Halazepam and Diazepam in Geriatric Subjects
Table 1. Clinical Condition of the Subjects
deviations (foot placed beside the marked line) was counted and the time to perform the task was meaClinical Diagnosis Medication sured. The exercise was repeated 3 or 4 times successively and the average for each variable served as Placebo # 10 Hypothyroidism Sodium Levothyroxine 0.2 mg o.d. the subject's individual datum for each session. # 11 Mild coronary insufficiency Nitroglycerin, Pentaerythritol Crawford's test estimated manual dexterity [1]. Tetranitrate In this test, the subjects had to place the maximum Diazepam 5 mg o. d. # 27 number of pins on a board with multiple perfora# 42 Hypertension 195/115 tions during 60 sec. This was done separately first Thyroid extract 2 gr/day # 47 Euthyroid goiter Chlordiazepoxide HC1 & with each hand and then with both hands together. Clidinium Bromide p. r. n. The number of pins placed were counted after 30 Chlordiazepoxide, Ethchlorvynol # 57 Mild anxiety and 60 sec for each of these three situations. Six reHalazepam-20 mg sults per session were thus recorded for each subject Acetylsalicylic Acid, # 10steoarthritis and two more were obtained by adding together the Xanthinol Niacinamide three results at 30 and 60 sec respectively; these adAcetylsalicylic Acid # 6 Arthralgia Dihydroergotoxine Mesylate, # 24 ditions yielded two total scores to be individually Maalox analyzed. Conjugated Estrogens, Chlor# 28 The side effects were noted by a nurse during Tripolon the experiment and reviewed by the clinician during Rauwolfia-Bendroflumethiazide # 32 Systolic Hypertension an interview with the subjects on the last experi210/95 Chlorpropamide 125 mg o. d. # 39 Diabetes mental day. These were volunteered symptoms; no # 48 Obesity check list was used. Compliance to the drug regimen Diazepam 5 mg o. d. # 55 Mild anxiety was assessed daily by counting the remaining capHalazepam-40 mg sules in the container. Complete hemogram, urine Digitalis, Spironolactone, # 18 Arteriosclerotic heart analysis and the SMA-12 battery of laboratory Hydrochlorothiazide, Quinidine disease Xanthinol Diacinate 150 mg b. i. d. tests* were performed prior to and after the study. # 20 # 22
Phenylbutazone 150 mg b. i. d., Methotrimeprazine, Vitamine C Magnesium-Aluminum Gel
# 31 Psoriasis, Peptic ulcer # 41 Paget's disease Nitroglycerin # 52 Angor pectoris Diazepam-5 mg # 9 Arterial hypertension a-Methyl dopa 180/100 # 12 Familial hypercholesteremia Acetylsalicylic acid # 23 Osteoarthritis # 34 Stasis dermatitis Iron, Dihydroergotoxinemesylate # 37 Past history of anemia # 38 Polycythemia, Hypertension 190/100 Hydrochlorothiazide # 49 Peripheral oedema
periment, except for one subject (in the placebo group) who continued to use ethchlorvynol 250-500 mg for insomnia; the other subjects, who were already taking some psychotropic medications, stopped at least a week prior to the study. All other non-psychotropic drugs were continued at the same dosage regimen.
Measurements Tandem walk is a classical neurological test for the evaluation of gait [2, 3]. The subjects were asked to walk on a straight line, 25 feet long, placing one foot immediately in front of the other. The number of
Design, Protocol and Analysis This study was planned to be a double-blind comparative experiment with four equal groups of subjects, each group receiving one of the four treatments to be compared: halazepam 20 and 40 mg (H20 and H40), diazepam 5 mg (D) and placebo (P). The clinical selection ended up with a sample of 59 volunteers; after randomization, group D started with 14 subjects whereas the three others had 15. Three subjects in the H40 group did not terminate the study; data were thus available for analysis on a total of 56 subjects; respectively 15 (H20), 12 (H40), 14 (D) and 15 (P) for the 4 groups. On day 0, measurements were recorded for the main variables (tandem walk and manual dexterity). Drug administration began on day 1 and the same variables were measured again on days 3 and 14. The tandem walk was assessed between 9.00 and 12.00 h and Crawford's test was performed between 10.00 and 14.00 h. The investigator was free to remove from the study any subject with significant side effect(s). This subject was not to be replaced. Although she was * These were: SGOT, Ca ++, Inorganic Phosphorus, Glucose, Urea, Uric Acid, Cholesterol, Total Protein, Albumin, Total Bilirubin, Alkaline Phosphatase, LDH.
445
M.-A. Gagnon et al.: Halazepam and Diazepam in Geriatric Subjects too ataxic to perform the tandem walk on day 3, one subject receiving halazepam 40 mg remained in the trial and was tested on day 14. Another subject in the same group dropped out on day 5; her data for day 3 (and day 0 for covariance analysis) were used in the analysis. The analysis of variance was performed on the data obtained on 'day 0 in order to assess any initial difference among the groups. These initial data served as the covariate in the separate analyses of covariance for days 3 and 14. Heterogeneity of regression per treatment was tested in order to justify the conclusions of the latter type of analysis. When the global difference between the treatments was significant (p < 0.05), individual comparisons were performed by a series of a priori tests taking into account the unequal number of subjects per group. These analyses were done on the main variables only, Fisher's exact probability test being used for sideeffect, drop-out and drug-compliance rates.
Methods of Control Both a placebo and an active standard (diazepam) were used in this comparative study. The experiment was conducted under double-blind conditions, the code of individual subjects being available only in case of emergency. The report forms on the main variables, once completed, were not available to the observers. The treatments were allocated to the subjects by the use of a random number table and the experimental conditions were otherwise controlled by the following standardization: all subjects entered the study on the same day; the same clinician supervised all subjects; one observer recorded all measurements on gait and another obtained all data on manual dexterity; uniform drug administration and fixed dosage were to be maintained. Finally the intersubject variation was minimized by weighing, through covariance analysis, the results of the main variables according to their relation to the initial values.
Results
Tandem Walk Despite the health status difference among the subjects and their walking difficulties at the beginning of the study, randomization yielded homogeneity among the groups for the two main variables, for which no significant differences (p > 0.20) were found on day 0. Global differences between treatments were
found to be significant, during drug administration, on all measurements of the tandem walk. The adjusted treatment means are presented in Figure 1, which shows the outstanding effect of halazepam 40 mg. This treatment significantly (p < 0.05 or p < 0.01) increases both the number of deviations and the time needed to walk the required distance. Halazepam 40 mg does not only differ significantly from placebo, but also from each of the two other (active) treatments, except in one occasion (the variable 'time' at day 3) when it is not found to be different from diazepam 5 mg. Placebo, halazepam 20 mg and diazepam 5 mg do not differ significantly from one another. A simple analysis of variance on the data would not have shown the differences described above. Since there was no significant heterogeneity of regression on the initial data (for an example, see Table 2) and since the correlation coefficient was high in all occasions (r = 0.6, day 3; r = 0.7, day 14 - p < 0.0001), the analysis of covariance was highly indicated and permitted us to find these conclusive results: halazepam 40 mg hinders tandem walk more than halazepam 20 mg, placebo or diazepam 5 mg.
Manual Dexterity The correlation coefficients with the initial data on days 3 and 14 are very significant, with even higher values (0.73 < r < 0.87) than for the gait; the regression is also homogeneous among treatments. With the Crawford's test more differences were found, particularly between the placebo and the active drugs (Fig. 2); furthermore, global differences between treatments are highly significant (p < 0.01 or p < 0.001), except for some results of day 14, when the differences are generally smaller than on day 3. Discrimination between treatments increases in the following order: right hand, left hand, both hands and total score. The 30 and 60 s counts yield approximately the same results with a similar discrimination. The manual dexterity is significantly and linearly decreased by the increase of the halazepam dose. The low dose is sometimes (4/16 occasions) significantly different from placebo and always produces a lower mean score. The high dose always significantly decreases the performance, compared to placebo (p < 0.01); it also significantly differs from the low dose in most situations. Diazepam 5 mg decreases manual dexterity, significantly differing from placebo most of the times. The mean results obtained with this drug are situated between those of the halazepam 20 mg and the halazepam 40 mg groups, and are significantly
446
M.-A. Gagnon et al.: Halazepam and Diazepam in Geriatric Subjects
TANDEM WALK Time
DAY3
z2zIYI4
MANUAL DEXTERITY (60see) DAYI4
Number of pins £),dYJ both hands
50
30
l
25
25
i i
20 0 Number of deviations
Total score 70-
3
-
2
-
--I
60-
50i 0 q" ~
0 I'
PLACEBO HALAZFPAM HALAZEPAM DIAZEPAM
PLACEBO i I J HALAZEPAM 20mg r///.4 HALAZEPAM 4 0 ~ g DIAZEPAM 5 mg r.;~~:;!
I I 20mg ~//~ 40n,g ~ 5mg ~P::::~
Fig. 1
Fig. 2
different from them only occasionally; once from the halazepam 20 mg and thrice from the halazepam 40 mg group.
mg fell to the ground. With this dose, three subjects left the study because of too many adverse effects on days 2, 5 and 10 respectively. The frequency of volunteered symptoms is presented in Table 4. Although they may not all be drug-related, it is felt that a complete list of plausible effects should be recorded in an early phase study. The active treatments are compared to placebo according to Fisher's exact probability test. Significantly more subjects in each of the active treatment groups complained of drowsiness than in the placebo group (p < 0.01). Only in the halazepam 40 mg group were there significantly more subjects who suffered from confusion, ataxia and unsteady gait (p < 0.005); 'poor memory' was also volunteered more often by these subjects. With diazepam 5 rag, more subjects were fatigued, were ataxic and had 'depressed motor activity'. The quality of sleep seems to be significantly improved with halazepam 20 rag.
Other Effects The unsteady gait or ataxia was so marked in some volunteers, as clinically observed in the first 3 days, that the investigator decided to reduce from three to two the number of doses per day as soon as the third day measurements were recorded. As found out afterwards, this mostly affected the subjects of the halazepam 40 mg group. Even after this dosage reduction, 8 volunteers took only one capsule per day, skipping the morning dose which could most affect their performance for the tests. This happened from days 4 to 9 and all but one of these subjects were in the halazepam 40 mg group; the other was receiving diazepam 5 mg (Table 3). The intensity of ataxia was such that four subjects receiving halazepam 40
447
M.-A. Gagnon et al.: Halazepam and Diazepam in Geriatric Subjects
Table 4. Summary of Volunteered Symptoms
Table 2. Statistical Analysis Tandem walk: no. of deviations on D a y 14 df Analysis of variance Treatments Error Analysis of covariance Treatments Regression (single) Error Test of heterogeneity Regression (multiple) Heterogeneity Error
M.S.
F
p
> 0.05
3 52
8.11 2.94
2.76
3 l 51
4.93 76.3l 1.50
3.29 50.97
0.028 < 0.001
4 3 48
20.11 1.36 1.50
13.36 < 1
< 0.001 -
H2o 1.42 1.20 1.42
H40 1.89 2.90 2.66
D 1.38 1.21 1.46
Means P* 1.91 1.60 1.34
Day 0 D a y 14 (unadjusted) D a y 14 (adjusted)
Individual Comparisons of Adjusted Means P vs H20 P vs D D vs H20
F < 1 < i < 1
p n.s. n.s. n.s.
F 7.73 6.02 6.75
H40 vs P H40 vs D H40 vs H
p < 0.01 0.017 0.012
* P: Placebo (n = 15); D: D i a z e p a m 5mg (n = 14); H20: H a l a z e p a m 20 m g (n = 15); H40: H a l a z e p a m 40 m g (n =
12)
Agitation or Restlessness: Anorexia Ataxia, unsteady gait Blurred Vision Confusion Depression Depressed motor activity Dizziness Drowsiness or Sleepiness Dry mouth Euphoria Fatigue Insomnia Itching Nausea Poor memory Speech disturbance Urge to urinate Weakness Diminution of anxiety Improvement in quality of sleep Improvement of mood Increased appetite Non compliance and drop-outs # subjects with no other "effects' # subjects with undesirable 'effects" # subjects with desirable 'effects' only: Total subjects given dosage: Drop-outs
Table 3. Compliance to Drug Regimen a
Drugs
Placebo Halazepam 20 mg Halazepam 40 mg Diazepam 5 mg
Number Compli- Compliance ance
Once/day Dropouts
Day 1
D a y 1-3 D a y 4 - 1 4 D a y 4 - 1 4
15 15 15 14
15 15 14 14
15 15 5 13
0 0 7 1
0 0 3 0
Almost every subject in the three active treatment groups had one or more 'undesirable' symptoms; this is highly significant when compared to the placebo group. The analysis also confirms the outstanding non-compliance to the drug regimen in the halazepam 40 mg group. Upon examination of the laboratory findings which could be of clinical significance in the opinion of the investigator, no direct relationship with the drugs could be established.
Discussion In this experiment, the approach was pragmatic, rather than puristic ('approches pratique et explicative' [5]). The subjects were not healthy young vol-
p < 0.05,
u p < 0.01,
P 2 0 0 2 0 0 0 1 0 0 0 1 3 0 1 0 0 1 0
H20 0 0 2 0 0 1 1 3 6b 0 0 4 0 0 0 1 0 0 0
H40 1 1 10 c 1 8c 0 2 3 13 c 0 3 5 0 0 2 5a 1 0 0
D 1 0 5a 0 3 0 4a 1 8c 2 0 7a 0 1 0 1 2 2 I
0 3 1 1
2 1@ 2 0
0 6 1 1
0 7 2 0
0 8 5 2
0 10 ° 1 2a 0c 0° 13 c 14 c 13 c 0 1 1
15
15
15
14
0
0
3
0
c p < 0.005 (versus placebo)
Fisher's exact probability test, one-tailed
unteers, as they would be in the latter approach. Ours was not a classical drug-oriented study to assess ataxia as a pharmacological effect in man. On the contrary, we wanted to determine a tolerable dose-range for this important side-effect of benzodiazepines in a population more at risk. Thus the sample did not exclude the significant proportion of elderly subjects who often suffer from one disease or another. All active treatments significantly decrease motor coordination as estimated by the manual dexterity test. This effect, however, is not completely translated into the complex psychomotor task performed during tandem walk. In the latter, compensation seems more readily achieved against small pharmacological effects; in contrast, the disturbance seems even more obvious when gait instability is produced, as with halazepam 40 mg. In other words, the effects on tandem walk correspond almost to an all-or-nothing response. Ataxia and non-compliance to the therapeutic regimen must be taken into account in the interpretation of the results on the main variables. A bias in favor of halazepam 40 mg was created both for day
448
3 and day 14 measurements, the latter being much pronounced. On day 3, one subject had already dropped out whereas another could not perform the tandem walk (she was not submitted to Crawford's test either). If they had been given the lowest possible scores, the two subjects would have widened the gap between halazepam 40 mg and all other treatments. The same applies for the other two drop-outs who were not tested on day 14. At this time, the bias is exagerated by the fact that only 5 subjects were still taking 40 mg b.i.d. The subjects' performance seems to improve from day 3 to day 14. Practice may explain part of this improvement, since it is observed also in the placebo patients. The possible r61e of drug tolerance cannot be assessed, because of the dosage change after day 3. This is particularly true for the high dose of halazepam. Clearly halazepam 40 mg, taken thrice daily for 3 days and twice afterwards, is a dose which should not be used in this type of female persons, most of whom already had some walking difficulties. This is substantiated by the results obtained on all pertinent variables which were measured: tandem walk, manual dexterity test, other 'effects' and compliance to drug regimen. Indeed only 5 subjects could tolerate halazepam 40 mg twice daily. With the manual dexterity test significant effects of diazepam 5 mg and halazepam 20 mg on motor coordination could be observed. Yet this test as well as the analysis of the other effects shows that the lower dose of halazepam is well tolerated, when compared to diazepam; this lower dose seems to be
M.-A. Gagnon et al.: Halazepam and Diazepam in Geriatric Subjects
at least as good as the dose utilized for the active drug standard. It is possible that, as with the other benzodiazepines, there is a particular sensitivity of elderly subjects to halazepam. From our data, it seems that the dosage of halazepam, for this type of geriatric subjects, should not exceed 20 mg b.i.d, or t.i.d. Acknowledgements. We thank Dr. A. Clermont, of Schering Corp. (Canada) Ltd., for his help in this study. References 1. Crawford, J.E., Crawford, D.M.: Small parts dexterity test. New York: The Psychological Corporation 1956 2. Daubs, J.G.: Vision/non-vision ataxia test battery. Amer. J. Optom. Physiol. Optics 51, 400-413 (1974) 3. Hertzberg, C.E.: The relationship of motor ability to the intelligence of kindergarten children. J. Educ. Psychol. 20, 507-519
(1929) 4. Jarvik, M.E.: Drugs used in the treatment of psychiatric disorders. In: The Pharmacological Basis of Therapeutics (eds. L. S. Goodman and A. Gilman), pp. 179-180. Toronto: Macmillan Company 1970 5. Schwartz, D., Flamant, R., Lellouch, J.: L'essai th6rapeutique chez l'homme. Paris: Flammarion 1970
Received: December 7, 1976, accepted: February 15, 1977 M.-A. Gagnon Institut National de la Recherche Scientifique (INRS-Sant6) Louis-H. Lafontaine Hospital 7401 Hochelaga Street Montr6al, Quebec, Canada H1N 3M5
Europ. J. clin. Pharmacol. 11,443-448 (1977)
Effects of Halazepam and Diazepam on the Motor Coordination of Geriatric Subjects* M.-A. Gagnon, Y. Langlois, D. R. Boghen, and M. Verdy Institut National de la Recherche Scientifique (INRS-Sant6), Louis-H. Lafontaine Hospital Montreal and H6tel-Dieu de Montr6al, Montreal, Quebec, Canada
Summary. Two doses of halazepam, 20 and 40 mg, were compared to diazepam 5 mg and placebo in this double-blind, parallel experiment. The medications were administered thrice daily during the first 3 days, and twice daily during the following 11 days. Their effects were measured on the tandem walk and manual dexterity of 59 elderly female subjects. Most volunteers had some walking difficulty prior to the study and 21 suffered from some illness. With the higher dose of halazepam a statistically and clinically significant impairment of motor coordination was observed. Halazepam 40 mg, at the regimen above mentioned, should therefore not be used in elderly persons. This is substantiated by the results on all pertinent variables: tandem walk, manual dexterity, side effects and compliance to the drug regimen; indeed only 5 subjects out of 15 could tolerate halazepam 40 mg twice daily. With the manual dexterity test, significant effects of diazepam 5 mg and halazepam 20 mg were also observed. Furthermore, this test as well as the analysis of the side effects, shows that the lower dose of halazepam is well tolerated when compared to diazepam 5 mg.
Key words: Halazepam, gait, manual dexterity, motor coordination, geriatry.
Benzodiazepines are well known for producing ataxia, especially in older people [4]. In the course of early clinical trials on halazepam, a new benzodiazepine, the present study was set up in order to investigate the safety of the drug rather than its efficacy; more precisely, the purpose of the experiment was to establish a safe dose range in older subjects * Supported in part by Schering Corporation
with regard to its effects on motor coordination. These were estimated by the assessment of gait and manual dexterity. The measurement of side effects and drug regimen compliance yielded additional evidence on the hazards of halazepam high dosage.
Methods
Subjects Fifty-nine female members of a religious order, volunteered and were accepted for this experiment after giving an informed consent. Their age varied from 57 to 88 years (average: 69.5) and their weight from 40.9 to 79.5 kg (average: 60.3). Although 21 suffered from some illness and 21 were taking some medication prior to the study (Table 1), they were all judged, after a medical examination by their clinician, to be fit for the experiment. Volunteers with a renal or hepatic pathology, which could hinder the metabolism or excretion of the drugs, were excluded from the study, as well as those suspected of benzodiazepine hypersensitivity. Most subjects had some walking difficulty, related to osteoarthritis or cerebral arteriosclerosis, yet all were able to perform the tandem walk and manual dexterity tests within a 20-min training period.
Drugs Two doses of halazepam, 20 and 40 rag, were compared to diazepam 5 mg and placebo. The medications were contained in identical capsules. These were administered thrice daily, at 7.00, 15.00 and 22.00 h, during the first 3 days of the experiment; the afternoon dose was dropped afterwards. No other psychotropic drugs were taken during the ex-
444
M.-A. Gagnon et al.: Halazepam and Diazepam in Geriatric Subjects
Table 1. Clinical Condition of the Subjects
deviations (foot placed beside the marked line) was counted and the time to perform the task was meaClinical Diagnosis Medication sured. The exercise was repeated 3 or 4 times successively and the average for each variable served as Placebo # 10 Hypothyroidism Sodium Levothyroxine 0.2 mg o.d. the subject's individual datum for each session. # 11 Mild coronary insufficiency Nitroglycerin, Pentaerythritol Crawford's test estimated manual dexterity [1]. Tetranitrate In this test, the subjects had to place the maximum Diazepam 5 mg o. d. # 27 number of pins on a board with multiple perfora# 42 Hypertension 195/115 tions during 60 sec. This was done separately first Thyroid extract 2 gr/day # 47 Euthyroid goiter Chlordiazepoxide HC1 & with each hand and then with both hands together. Clidinium Bromide p. r. n. The number of pins placed were counted after 30 Chlordiazepoxide, Ethchlorvynol # 57 Mild anxiety and 60 sec for each of these three situations. Six reHalazepam-20 mg sults per session were thus recorded for each subject Acetylsalicylic Acid, # 10steoarthritis and two more were obtained by adding together the Xanthinol Niacinamide three results at 30 and 60 sec respectively; these adAcetylsalicylic Acid # 6 Arthralgia Dihydroergotoxine Mesylate, # 24 ditions yielded two total scores to be individually Maalox analyzed. Conjugated Estrogens, Chlor# 28 The side effects were noted by a nurse during Tripolon the experiment and reviewed by the clinician during Rauwolfia-Bendroflumethiazide # 32 Systolic Hypertension an interview with the subjects on the last experi210/95 Chlorpropamide 125 mg o. d. # 39 Diabetes mental day. These were volunteered symptoms; no # 48 Obesity check list was used. Compliance to the drug regimen Diazepam 5 mg o. d. # 55 Mild anxiety was assessed daily by counting the remaining capHalazepam-40 mg sules in the container. Complete hemogram, urine Digitalis, Spironolactone, # 18 Arteriosclerotic heart analysis and the SMA-12 battery of laboratory Hydrochlorothiazide, Quinidine disease Xanthinol Diacinate 150 mg b. i. d. tests* were performed prior to and after the study. # 20 # 22
Phenylbutazone 150 mg b. i. d., Methotrimeprazine, Vitamine C Magnesium-Aluminum Gel
# 31 Psoriasis, Peptic ulcer # 41 Paget's disease Nitroglycerin # 52 Angor pectoris Diazepam-5 mg # 9 Arterial hypertension a-Methyl dopa 180/100 # 12 Familial hypercholesteremia Acetylsalicylic acid # 23 Osteoarthritis # 34 Stasis dermatitis Iron, Dihydroergotoxinemesylate # 37 Past history of anemia # 38 Polycythemia, Hypertension 190/100 Hydrochlorothiazide # 49 Peripheral oedema
periment, except for one subject (in the placebo group) who continued to use ethchlorvynol 250-500 mg for insomnia; the other subjects, who were already taking some psychotropic medications, stopped at least a week prior to the study. All other non-psychotropic drugs were continued at the same dosage regimen.
Measurements Tandem walk is a classical neurological test for the evaluation of gait [2, 3]. The subjects were asked to walk on a straight line, 25 feet long, placing one foot immediately in front of the other. The number of
Design, Protocol and Analysis This study was planned to be a double-blind comparative experiment with four equal groups of subjects, each group receiving one of the four treatments to be compared: halazepam 20 and 40 mg (H20 and H40), diazepam 5 mg (D) and placebo (P). The clinical selection ended up with a sample of 59 volunteers; after randomization, group D started with 14 subjects whereas the three others had 15. Three subjects in the H40 group did not terminate the study; data were thus available for analysis on a total of 56 subjects; respectively 15 (H20), 12 (H40), 14 (D) and 15 (P) for the 4 groups. On day 0, measurements were recorded for the main variables (tandem walk and manual dexterity). Drug administration began on day 1 and the same variables were measured again on days 3 and 14. The tandem walk was assessed between 9.00 and 12.00 h and Crawford's test was performed between 10.00 and 14.00 h. The investigator was free to remove from the study any subject with significant side effect(s). This subject was not to be replaced. Although she was * These were: SGOT, Ca ++, Inorganic Phosphorus, Glucose, Urea, Uric Acid, Cholesterol, Total Protein, Albumin, Total Bilirubin, Alkaline Phosphatase, LDH.
445
M.-A. Gagnon et al.: Halazepam and Diazepam in Geriatric Subjects too ataxic to perform the tandem walk on day 3, one subject receiving halazepam 40 mg remained in the trial and was tested on day 14. Another subject in the same group dropped out on day 5; her data for day 3 (and day 0 for covariance analysis) were used in the analysis. The analysis of variance was performed on the data obtained on 'day 0 in order to assess any initial difference among the groups. These initial data served as the covariate in the separate analyses of covariance for days 3 and 14. Heterogeneity of regression per treatment was tested in order to justify the conclusions of the latter type of analysis. When the global difference between the treatments was significant (p < 0.05), individual comparisons were performed by a series of a priori tests taking into account the unequal number of subjects per group. These analyses were done on the main variables only, Fisher's exact probability test being used for sideeffect, drop-out and drug-compliance rates.
Methods of Control Both a placebo and an active standard (diazepam) were used in this comparative study. The experiment was conducted under double-blind conditions, the code of individual subjects being available only in case of emergency. The report forms on the main variables, once completed, were not available to the observers. The treatments were allocated to the subjects by the use of a random number table and the experimental conditions were otherwise controlled by the following standardization: all subjects entered the study on the same day; the same clinician supervised all subjects; one observer recorded all measurements on gait and another obtained all data on manual dexterity; uniform drug administration and fixed dosage were to be maintained. Finally the intersubject variation was minimized by weighing, through covariance analysis, the results of the main variables according to their relation to the initial values.
Results
Tandem Walk Despite the health status difference among the subjects and their walking difficulties at the beginning of the study, randomization yielded homogeneity among the groups for the two main variables, for which no significant differences (p > 0.20) were found on day 0. Global differences between treatments were
found to be significant, during drug administration, on all measurements of the tandem walk. The adjusted treatment means are presented in Figure 1, which shows the outstanding effect of halazepam 40 mg. This treatment significantly (p < 0.05 or p < 0.01) increases both the number of deviations and the time needed to walk the required distance. Halazepam 40 mg does not only differ significantly from placebo, but also from each of the two other (active) treatments, except in one occasion (the variable 'time' at day 3) when it is not found to be different from diazepam 5 mg. Placebo, halazepam 20 mg and diazepam 5 mg do not differ significantly from one another. A simple analysis of variance on the data would not have shown the differences described above. Since there was no significant heterogeneity of regression on the initial data (for an example, see Table 2) and since the correlation coefficient was high in all occasions (r = 0.6, day 3; r = 0.7, day 14 - p < 0.0001), the analysis of covariance was highly indicated and permitted us to find these conclusive results: halazepam 40 mg hinders tandem walk more than halazepam 20 mg, placebo or diazepam 5 mg.
Manual Dexterity The correlation coefficients with the initial data on days 3 and 14 are very significant, with even higher values (0.73 < r < 0.87) than for the gait; the regression is also homogeneous among treatments. With the Crawford's test more differences were found, particularly between the placebo and the active drugs (Fig. 2); furthermore, global differences between treatments are highly significant (p < 0.01 or p < 0.001), except for some results of day 14, when the differences are generally smaller than on day 3. Discrimination between treatments increases in the following order: right hand, left hand, both hands and total score. The 30 and 60 s counts yield approximately the same results with a similar discrimination. The manual dexterity is significantly and linearly decreased by the increase of the halazepam dose. The low dose is sometimes (4/16 occasions) significantly different from placebo and always produces a lower mean score. The high dose always significantly decreases the performance, compared to placebo (p < 0.01); it also significantly differs from the low dose in most situations. Diazepam 5 mg decreases manual dexterity, significantly differing from placebo most of the times. The mean results obtained with this drug are situated between those of the halazepam 20 mg and the halazepam 40 mg groups, and are significantly
446
M.-A. Gagnon et al.: Halazepam and Diazepam in Geriatric Subjects
TANDEM WALK Time
DAY3
z2zIYI4
MANUAL DEXTERITY (60see) DAYI4
Number of pins £),dYJ both hands
50
30
l
25
25
i i
20 0 Number of deviations
Total score 70-
3
-
2
-
--I
60-
50i 0 q" ~
0 I'
PLACEBO HALAZFPAM HALAZEPAM DIAZEPAM
PLACEBO i I J HALAZEPAM 20mg r///.4 HALAZEPAM 4 0 ~ g DIAZEPAM 5 mg r.;~~:;!
I I 20mg ~//~ 40n,g ~ 5mg ~P::::~
Fig. 1
Fig. 2
different from them only occasionally; once from the halazepam 20 mg and thrice from the halazepam 40 mg group.
mg fell to the ground. With this dose, three subjects left the study because of too many adverse effects on days 2, 5 and 10 respectively. The frequency of volunteered symptoms is presented in Table 4. Although they may not all be drug-related, it is felt that a complete list of plausible effects should be recorded in an early phase study. The active treatments are compared to placebo according to Fisher's exact probability test. Significantly more subjects in each of the active treatment groups complained of drowsiness than in the placebo group (p < 0.01). Only in the halazepam 40 mg group were there significantly more subjects who suffered from confusion, ataxia and unsteady gait (p < 0.005); 'poor memory' was also volunteered more often by these subjects. With diazepam 5 rag, more subjects were fatigued, were ataxic and had 'depressed motor activity'. The quality of sleep seems to be significantly improved with halazepam 20 rag.
Other Effects The unsteady gait or ataxia was so marked in some volunteers, as clinically observed in the first 3 days, that the investigator decided to reduce from three to two the number of doses per day as soon as the third day measurements were recorded. As found out afterwards, this mostly affected the subjects of the halazepam 40 mg group. Even after this dosage reduction, 8 volunteers took only one capsule per day, skipping the morning dose which could most affect their performance for the tests. This happened from days 4 to 9 and all but one of these subjects were in the halazepam 40 mg group; the other was receiving diazepam 5 mg (Table 3). The intensity of ataxia was such that four subjects receiving halazepam 40
447
M.-A. Gagnon et al.: Halazepam and Diazepam in Geriatric Subjects
Table 4. Summary of Volunteered Symptoms
Table 2. Statistical Analysis Tandem walk: no. of deviations on D a y 14 df Analysis of variance Treatments Error Analysis of covariance Treatments Regression (single) Error Test of heterogeneity Regression (multiple) Heterogeneity Error
M.S.
F
p
> 0.05
3 52
8.11 2.94
2.76
3 l 51
4.93 76.3l 1.50
3.29 50.97
0.028 < 0.001
4 3 48
20.11 1.36 1.50
13.36 < 1
< 0.001 -
H2o 1.42 1.20 1.42
H40 1.89 2.90 2.66
D 1.38 1.21 1.46
Means P* 1.91 1.60 1.34
Day 0 D a y 14 (unadjusted) D a y 14 (adjusted)
Individual Comparisons of Adjusted Means P vs H20 P vs D D vs H20
F < 1 < i < 1
p n.s. n.s. n.s.
F 7.73 6.02 6.75
H40 vs P H40 vs D H40 vs H
p < 0.01 0.017 0.012
* P: Placebo (n = 15); D: D i a z e p a m 5mg (n = 14); H20: H a l a z e p a m 20 m g (n = 15); H40: H a l a z e p a m 40 m g (n =
12)
Agitation or Restlessness: Anorexia Ataxia, unsteady gait Blurred Vision Confusion Depression Depressed motor activity Dizziness Drowsiness or Sleepiness Dry mouth Euphoria Fatigue Insomnia Itching Nausea Poor memory Speech disturbance Urge to urinate Weakness Diminution of anxiety Improvement in quality of sleep Improvement of mood Increased appetite Non compliance and drop-outs # subjects with no other "effects' # subjects with undesirable 'effects" # subjects with desirable 'effects' only: Total subjects given dosage: Drop-outs
Table 3. Compliance to Drug Regimen a
Drugs
Placebo Halazepam 20 mg Halazepam 40 mg Diazepam 5 mg
Number Compli- Compliance ance
Once/day Dropouts
Day 1
D a y 1-3 D a y 4 - 1 4 D a y 4 - 1 4
15 15 15 14
15 15 14 14
15 15 5 13
0 0 7 1
0 0 3 0
Almost every subject in the three active treatment groups had one or more 'undesirable' symptoms; this is highly significant when compared to the placebo group. The analysis also confirms the outstanding non-compliance to the drug regimen in the halazepam 40 mg group. Upon examination of the laboratory findings which could be of clinical significance in the opinion of the investigator, no direct relationship with the drugs could be established.
Discussion In this experiment, the approach was pragmatic, rather than puristic ('approches pratique et explicative' [5]). The subjects were not healthy young vol-
p < 0.05,
u p < 0.01,
P 2 0 0 2 0 0 0 1 0 0 0 1 3 0 1 0 0 1 0
H20 0 0 2 0 0 1 1 3 6b 0 0 4 0 0 0 1 0 0 0
H40 1 1 10 c 1 8c 0 2 3 13 c 0 3 5 0 0 2 5a 1 0 0
D 1 0 5a 0 3 0 4a 1 8c 2 0 7a 0 1 0 1 2 2 I
0 3 1 1
2 1@ 2 0
0 6 1 1
0 7 2 0
0 8 5 2
0 10 ° 1 2a 0c 0° 13 c 14 c 13 c 0 1 1
15
15
15
14
0
0
3
0
c p < 0.005 (versus placebo)
Fisher's exact probability test, one-tailed
unteers, as they would be in the latter approach. Ours was not a classical drug-oriented study to assess ataxia as a pharmacological effect in man. On the contrary, we wanted to determine a tolerable dose-range for this important side-effect of benzodiazepines in a population more at risk. Thus the sample did not exclude the significant proportion of elderly subjects who often suffer from one disease or another. All active treatments significantly decrease motor coordination as estimated by the manual dexterity test. This effect, however, is not completely translated into the complex psychomotor task performed during tandem walk. In the latter, compensation seems more readily achieved against small pharmacological effects; in contrast, the disturbance seems even more obvious when gait instability is produced, as with halazepam 40 mg. In other words, the effects on tandem walk correspond almost to an all-or-nothing response. Ataxia and non-compliance to the therapeutic regimen must be taken into account in the interpretation of the results on the main variables. A bias in favor of halazepam 40 mg was created both for day
448
3 and day 14 measurements, the latter being much pronounced. On day 3, one subject had already dropped out whereas another could not perform the tandem walk (she was not submitted to Crawford's test either). If they had been given the lowest possible scores, the two subjects would have widened the gap between halazepam 40 mg and all other treatments. The same applies for the other two drop-outs who were not tested on day 14. At this time, the bias is exagerated by the fact that only 5 subjects were still taking 40 mg b.i.d. The subjects' performance seems to improve from day 3 to day 14. Practice may explain part of this improvement, since it is observed also in the placebo patients. The possible r61e of drug tolerance cannot be assessed, because of the dosage change after day 3. This is particularly true for the high dose of halazepam. Clearly halazepam 40 mg, taken thrice daily for 3 days and twice afterwards, is a dose which should not be used in this type of female persons, most of whom already had some walking difficulties. This is substantiated by the results obtained on all pertinent variables which were measured: tandem walk, manual dexterity test, other 'effects' and compliance to drug regimen. Indeed only 5 subjects could tolerate halazepam 40 mg twice daily. With the manual dexterity test significant effects of diazepam 5 mg and halazepam 20 mg on motor coordination could be observed. Yet this test as well as the analysis of the other effects shows that the lower dose of halazepam is well tolerated, when compared to diazepam; this lower dose seems to be
M.-A. Gagnon et al.: Halazepam and Diazepam in Geriatric Subjects
at least as good as the dose utilized for the active drug standard. It is possible that, as with the other benzodiazepines, there is a particular sensitivity of elderly subjects to halazepam. From our data, it seems that the dosage of halazepam, for this type of geriatric subjects, should not exceed 20 mg b.i.d, or t.i.d. Acknowledgements. We thank Dr. A. Clermont, of Schering Corp. (Canada) Ltd., for his help in this study. References 1. Crawford, J.E., Crawford, D.M.: Small parts dexterity test. New York: The Psychological Corporation 1956 2. Daubs, J.G.: Vision/non-vision ataxia test battery. Amer. J. Optom. Physiol. Optics 51, 400-413 (1974) 3. Hertzberg, C.E.: The relationship of motor ability to the intelligence of kindergarten children. J. Educ. Psychol. 20, 507-519
(1929) 4. Jarvik, M.E.: Drugs used in the treatment of psychiatric disorders. In: The Pharmacological Basis of Therapeutics (eds. L. S. Goodman and A. Gilman), pp. 179-180. Toronto: Macmillan Company 1970 5. Schwartz, D., Flamant, R., Lellouch, J.: L'essai th6rapeutique chez l'homme. Paris: Flammarion 1970
Received: December 7, 1976, accepted: February 15, 1977 M.-A. Gagnon Institut National de la Recherche Scientifique (INRS-Sant6) Louis-H. Lafontaine Hospital 7401 Hochelaga Street Montr6al, Quebec, Canada H1N 3M5
