Theme : Dreams

Sleep Disturbances and EEG Slowing in Alzheimer's Disease

Jacques Montplaisir 1,2, Dominique Petit 1,2, Serge Gauthier 3, Hélène Gaudreau 1 and Anne Décary 1
1Centre d'étude du Sommeil, Hôpital du Sacré-Coeur, Montreal, Quebec, H4J 1C5, Canada, 2 Centre de Recherche Fernand-Seguin Department of Psychiatry, Université de Montréal, Montreal, Quebec, H1N 3V2, Canada, 3McGill Center for Studies in Aging, Douglas Hospital, McGill University, Montreal, Quebec, H4H 1R3, Canada
Abstract
Changes in sleep structure, and especially REM sleep, and in EEG activation were studied in relation to the cholinergic deficit found in Alzheimer's Disease (AD). With respect to sleep architecture, only REM sleep percent was reduced in AD patients compared to controls as a result of a decrease in mean REM episode duration. Different results were obtained in patients with progressive supranuclear palsy (PSP). These results are discussed with respect to the role of brainstem and forebrain cholinergic populations in REM sleep generation in humans. More importantly, it was shown by means of spectral analyses that EEG slowing is much more prominent in REM sleep than in wakefulness in AD. Furthermore, there is a distinct topographical pattern of REM sleep EEG slowing in AD patients which is in agreement with findings from neuroradiological and neuropathological studies. Using the ratio of slow over fast frequencies from the temporal regions, a correct classification of 90.4% of subjects was obtained for the REM sleep EEG. This discrimination rate is the best marker of AD so far using a single measure. Quantitative REM sleep EEG was also used to evaluate patients' biological response to cholinergic treatments. Finally, we present here preliminary data on the progression of EEG slowing in wakefulness and in REM sleep. After six months on a placebo, there was only a decrease in alpha activity in wakefulness over all regions studied. No changes were observed for REM sleep.

Current Claim: Quantitative REM sleep EEG measures can be used as a diagnostic tool and also to evaluate patients' biological responses to cholinergic treatment.

Activate the ShortNotes by clicking on this link. Your notes will be stored in this area and automatically retrieved upon your next visit.
A great interest has emerged in the last 20 years in finding biological markers of different diseases and psychiatric disorders. The series of studies we undertook was an attempt to find such a marker for Alzheimer's Disease (AD), through the analysis of sleep architecture and sleep microstructure variables and quantitative EEG analyses. One of the long-term goals was to be able to better evaluate the effects of pharmacological treatments of AD.
Sleep variables
One aspect of our research is to study sleep architecture and microstructure in mild to moderate AD (Montplaisir et al., 1995b). Our main finding is that REM sleep percent was reduced in AD patients compared to controls and this, as a result of a decrease in mean REM episodes duration. Other REM sleep variables, such as REM atonia, eye movement density, phasic EMG, number of REM episodes and REM latency, were all unchanged for the total REM sleep periods. In other words, variables pertaining to the initiation of REM sleep and to its characteristic features were unaffected in mild AD. This is probably because these variables are under the control of the mesopontine cholinergic populations, structures which are spared in mild AD. The lower REM sleep percentage, however, could be due to degeneration of the nucleus basalis of Meynert (NB). This nucleus normally exerts an inhibitory influence on the nucleus reticularis of the thalamus (Buzsaki et al., 1988), the rhythm generator responsible for NREM sleep. If the NB degenerates to the point of being unable to fully activate the cortex during wakefulness and REM sleep, its inhibitory influence on the nucleus reticularis might be weakened, leading to the curtailment of REM sleep episodes. This model suggests that a reduced activity in the NB, as during immobility or following a lesion, would result in an increase of slow delta waves. A similar observation can be expected in case of degeneration of the NB as in AD. However, percentage of NREM sleep did not vary significantly in our studies of AD patients as compared to previous studies (Prinz et al., 1982; Vitiello et al., 1990), probably because our group of patients were in mild stages of the disease. In contrast, the analysis of sleep microstructure revealed a decreased number of sleep spindles and K-complexes in AD patients compared to normal elderly subjects.

Quantitative EEG
We reasoned that the power of the EEG in diagnosing AD might be enhanced if it were assessed using REM sleep instead of wakefulness. The rationale was that the cholinergic basal forebrain, which is implicated in cortical activation and which degenerates early in AD (Whitehouse et al., 1982), is likely to be more crucial for EEG activation of REM sleep than of wakefulness. Indeed, the EEG activation achieved in wakefulness is the result of many convergent neuronal and neurotransmitter systems, many of which (including norepinephrine and histamine) are silent during REM sleep (Hobson et al., 1975; Sakai et al., 1990). This leaves mainly the cholinergic nucleus basalis (Buzsaki et al., 1988) and glutamatergic thalamo-cortical cells (Steriade and McCarley, 1990) to ensure EEG activation during the latter state. Thus, REM sleep EEG might be more likely to serve as a biological marker of AD than the awake EEG.

This is indeed what we have found. Slowing is more prominent in the REM sleep EEG than in the waking EEG of AD patients. When only the temporal regions, the cortical areas most and first affected in AD, are considered, the REM sleep EEG correctly classified 100% of controls and AD patients at an early stage of the disease (Petit et al., 1992). This first study was, however, conducted on a modest number of AD patients (n=8) and controls (n=8). The observed EEG slowing was the result of a change in all four classical frequency bands: an increase in percent power in delta and theta bands and a decrease in the percent power in alpha and beta bands. Furthermore, there was a distinct topographical pattern of REM sleep EEG slowing in AD patients which is in agreement with findings from neuroradiological and neuropathological studies and which was not observed for the waking EEG (Petit et al., 1993a). In REM sleep, the EEG slowing was indeed the most pronounced in the temporal regions, next most pronounced in the parieto-occipital, then the frontal regions, whereas the controls showed the same cortical activity values for all three regions studied. The REM sleep EEG was indeed correlated with a global assessment of cognitive functioning (the Mini-Mental State) and with interhemispheric asymmetry of regional cerebral blood flow by Single Photon Emission Computerized Tomography (Montplaisir et al., 1996).

The premortem diagnosis of probable or possible Alzheimer Disease is obtained exclusively from clinical examination. Without neuropathological confirmations, there are ambiguities concerning the reliability of the diagnosis, especially in mild cases, where the evolution of the clinical manifestations of AD are highly variable. As mentioned previously, the discriminative power of EEG between AD patients and control subjects, especially in REM sleep, demonstrated a high sensitivity and specificity in our earlier studies. Since diagnosing AD is troublesome in the first stages of the disease, the spectral EEG analysis yields additional objective and repeatable data for an early diagnosis of this heterogenous pathology.

To better pinpoint brain regions presenting significant cortical slowing, significance probability mapping was used in conjunction with a more complete EEG recording (Hassainia et al., 1994). It was confirmed that for REM sleep, EEG slowing was greater in the temporo-parietal and frontal regions, whereas for wakefulness, EEG slowing was greater for the frontal pole. Moreover, in order to reevaluate the discriminative power of our measure, a new and larger group of mild to moderate AD patients (n=27) and of controls (n=25) were studied with a 16-channel 10-20 recording montage (Hassainia et al., 1997). Using the ratio of slow over fast frequencies from the temporal regions, a correct classification of 90.4% of subjects (sensitivity: 81.5%, specificity: 100%) was obtained for the REM sleep EEG. The best discrimination rate for the waking EEG, obtained from the frontal regions, was only 80.8% (sensitivity: 66.7%, specificity: 96%). The discrimination rate obtained with the REM sleep EEG is the best marker of AD so far using a single measure. Tonic REM sleep also has the advantage over wakefulness of being a fairly homogenous state; it does not pose problems concerning variations in the patient's level of vigilance.

Role of brainstem and forebrain cholinergic populations
The implication of brainstem and forebrain cholinergic neurons in sleep-wake cycle regulation is well known. Knowledge of the role of these structures is based mostly on pharmacological, neuroanatomical, biochemical and electrophysiological studies in animals. The interest for the cholinergic mechanisms of REM sleep control probably began in the early 1960's with Jouvet and his first pontine lesion experiment in the cat (Jouvet, 1962). Later, he showed (Jouvet, 1969) that, in a pontine cat, REM sleep could be enhanced by the anticholinesterase eserine and suppressed by the muscarinic antagonist atropine. Moreover, injection of hemicholinium which blocks uptake of choline, the biosynthetic precursor of ACh, decreases REM sleep in cats (Domino and Stawiski, 1970; Hazra, 1970).

The understanding of human REM sleep is more obscure and derives from pharmacological studies using receptor blocking agents, cholinesterase inhibitors and muscarinic antagonists. Indeed, consistent with animal studies, scopolamine and atropine decrease or suppress REM sleep in normal subjects without affecting total sleep time (Sagales et al., 1975; Toyoda et al., 1966). In contrast, cholinomimetic agents like physostigmine or arecoline are known to increase REM sleep by inducing REM sleep episodes with no influence on their duration (Sitaram and Gillin, 1980).

We have studied further the differential role of brainstem and forebrain cholinergic populations by conducting a similar study in PSP patients (Montplaisir et al., 1997). In this neurodegenerative disease, one of the structures most affected (along with substantia nigra and basal ganglia) is the cholinergic pedunculopontine nucleus (PPT) of the brainstem (Jellinger, 1988). Since PPT is centrally involved in the control of REM sleep initiation and its characteristic features, it was expected that REM sleep variables would be altered in patients with PSP. Although the nucleus basalis of Meynert is also affected to a certain degree by cell loss, the remaining intact neurons do not demonstrate a significant reduction in nuclear volume, suggesting normal neurotransmitter production (Mann, 1982). Thus, a significant degree of EEG slowing in REM sleep was not expected for PSP patients.

We found that the percentage of REM sleep was lower in PSP patients than in controls. This was due to both a lower mean REM episode duration and a tendency to have fewer REM sleep episodes. REM density was also significantly reduced in these patients. Although mean latency to the first REM episode was not significantly different from that of controls, it showed much more variability in PSP patients.

In awake PSP patients, a slowing of the EEG (as determined by a spectral ratio) was found for the 6 frontal leads, C4, P4 and T4 compared to control subjects. For the REM sleep EEG, there were no significant between-group differences in the spectral ratio for any of the 16 leads studied. The frontal EEG slowing during wakefulness is consistent with the results of numerous neuropsychological studies which show deficits to be related to frontal lobe functions (Dubois et al., 1988; Maher et al.,1985). The fact that no EEG slowing was found in REM sleep suggests that the slowing observed for wakefulness was not likely due to a cholinergic deficit. This is consistent with findings that normal neocortical and hippocampal choline acetyltransferase activity was found in some PSP patients (Kish et al., 1985). Dopamine levels are, however, severely reduced in the caudate, putamen and substantia nigra in PSP patients (Kish et al., 1985). A frontal deafferentation from the striato-pallidal complex is thought to be responsible for the impairment since there are extensive fiber connections between these nuclei and the prefrontal region. Indeed, the positive correlations between degree of impairment on frontal tasks and EEG slowing observed in our PSP patients suggest that both impairments could be the result of a dopaminergic deficiency.

Evaluation of pharmacological treatments of AD
The final aspect of our work was the evaluation of pharmacological treatments acting upon the cholinergic system in mild to moderate AD patients. Our first attempt was with tetrahydroamino-acridine (THA ou tacrine), an acetylcholinesterase inhibitor (Petit et al., 1993b). THA did not significantly alter any sleep variable. Compared with the placebo condition, THA also had no significant effect on mean EEG spectral power for either wakefulness or REM sleep. However, a subgroup of AD patients seemed to respond to the drug, as both their Mini-Mental State scores and quantitative EEG measures (especially the REM sleep EEG) improved.

Linopirdine, an enhancer of stimulus-induced acetycholine release, and Xanomeline, a muscarinic (M1) receptor agonist, both appeared to stop the progression of the waking EEG slowing over six months compared to placebo (Montplaisir et al., 1995a; Petit et al., 1994). These drugs even seemed to produce a faster REM sleep EEG than what was observed at baseline. Interestingly, spectral maps of the REM sleep EEG revealed that the regions most reactivated by Xanomeline were the temporal regions, the slowest regions at baseline in these patients.

Clinical trials of experimental drugs for AD are often based on a six-month evaluation and are aimed at stopping the progression of the disease rather than reversing the process. However, there are not a lot of studies investigating what happens to EEG spectral power in a six-month period in mild AD. Previous studies reported that slowing of the awake EEG progressed with time (Coben et al., 1983; Fenton, 1986; Helkala et al., 1991; Penttilä et al., 1985; Sloan and Fenton, 1993) but no study has ever looked at the changes of quantitative EEG in REM sleep over time.

SHORT REPORT

Activate the ShortNotes by clicking on this link. Your notes will be stored in this area and automatically retrieved upon your next visit.
We present here preliminary data from a new study aimed at exploring the modifications of the EEG spectral composition in patients with mild to moderate AD in wakefulness and in REM sleep over time.

Eight AD patients (3 men and 5 women; mean age 68.4 years) were reassessed six months following the first investigation. All patients met the NINCDS-ADRDA criteria of probable Alzheimer's disease (McKhann et al., 1984). Patients were at mild to moderate stages of AD, i.e., stages 3 and 4 of Reisberg's scale (Reisberg et al., 1982). None of the subjects had a history of drug or alcohol abuse and all had been free of psychotropic medication for at least two weeks prior to the recordings.

Subjects were recorded in the sleep laboratory for one baseline night and one night six months later. Sixteen electrodes were placed according to the international 10-20 system, using a referential montage. A 16-channel Grass polygraph (amplifier gain 7.5µV/mm, bandpass 0.3-100 Hz) was used to amplify the signals and record them on paper. The signals were also relayed to a computer where they were digitized at a sampling rate of 128 Hz and filtered with a digital filter having a cutoff frequency at 64 Hz. Amplitude spectral analyses by fast Fourier transform were performed on artifact-free sections (96 s in total) of awake and of tonic REM sleep EEGs. The awake EEG was recorded while patients were lying on the bed with eyes closed on the morning following the night recording. The samples of REM sleep were visually selected from the first, middle and last REM sleep periods. Four frequency bands were defined as: delta (0.75 to 3.75 Hz), theta (4 to 7.75 Hz), alpha (8 to 12.75 Hz) and beta (13 to 20.25 Hz).

Activate the ShortNotes by clicking on this link. Your notes will be stored in this area and automatically retrieved upon your next visit.
The expected increase in low frequency bands was not observed in this group of AD patients six months after the first recording. In fact, changes in delta and theta bands were slight and non-significant during both wakefulness and REM sleep (Wilcoxon, p>0.21; Table 1). A marked reduction in absolute alpha activity was the only factor contributing to EEG slowing. This decrease in alpha activity was significant, specifically during wakefulness, for all five cortical regions investigated: frontal, central, occipital, parietal and temporal (Wilcoxon p<0.04; Table 2). The decrease in alpha activity during wakefulness was found in all eight AD patients. No significant changes were found during REM sleep for the alpha band in any cortical region.

The decrease in the alpha activity during wakefulness seems to parallel the evolution of AD, this modification following in time the increase in slow frequency bands. Coben et al. (1985) reported indeed that a low alpha activity is a phenomenon observed in more advanced stages of AD patients. The important reduction in alpha power observed in eight AD patients after six months indicates that these changes could be related to the rate of progression of the disease, at least for a brief period of time.

Many studies reported that the decline in the power of high frequencies was correlated with advancing stages of dementia and with a decline in neuropsychological performance (Brenner et al., 1986; Duffy et al., 1984; Penttila et al., 1985; Streletz et al., 1990). A low absolute alpha power has been reported to parallel poor performance on the MMSE test (Kuskowski et al., 1993), a global measure of cognitive decline in AD patients. Our observations are in agreement with this view. This biological marker of AD might eventually have important implications for adequate follow-up and care of AD patients. It may also allow prospective studies in patients with a positive neuropsychological diagnosis and a normal EEG on visual inspection; a mild EEG slowing might be detected at one point. It would also be interesting to investigate if EEG abnormalities can be observed before cognitive decline in individuals who would eventually develop AD.

Activate the ShortNotes by clicking on this link. Your notes will be stored in this area and automatically retrieved upon your next visit.
In conclusion, the study of sleep variables in specific neurodegenerative diseases can help to identify the role of different sleep-generating structures in humans. Quantitative EEG measures, especially those derived from REM sleep, can be used not only as a diagnostic tool but also to evaluate patients' biological responses to cholinergic treatment by a reactivation or a stabilization of EEG parameters.

1. Brenner RP, Ulrich RF, Spiker DG, Sclabassi RJ, Reynolds CF 3d, Marin RS, Boller F. Computerized EEG spectral analysis in elderly normal, demented and depressed subjects. Electroencephalogr Clin Neurophysiol 1986; 64: 483-92.
2. Buzsaki G, Bickford RG, Ponomareff G, Thal LJ, Mandel R, Gage FH. Nucleus basalis and thalamic control of neocortical activity in the freely moving rat. J Neurosci 1988; 8: 4007-26.

3. Coben LA, Danziger WL, Berg L. Frequency analysis of the resting awake EEG in mild senile dementia of Alzheimer's type. Electroencephalogr Clin Neurophysiol 1983; 55: 372-80.

4. Coben LA, Danziger W, Storandt M. A longitudinal EEG study of mild senile dementia of Alzheimer type: changes at 1 year and at 2.5 years. Electroencephalogr Clin Neurophysiol 1985; 61: 101-12.

5. Domino EF, Stawiski M. Effect of cholinergic antisynthesis agent HC-3 on the wake-sleep cycle of the cat. Psychophysiol 1970; 7: 315-6.

6. Dubois B, Pillon B, Legault F, Agid Y, Lhermitte F. Slowing of cognitive processing in progressive supranuclear palsy. Arch Neurol 1988; 45: 1194-9.

7. Duffy FH, Albert MS, McAnulty G. Brain electrical activity in patients with presenile and senile dementia of the Alzheimer type. Ann Neurol 1984; 16: 439-48.

8. Fenton GW. Electrophysiology of Alzheimer's disease. Br Med Bull 1986; 42: 29-33.

9. Hassainia F, Petit D, Montplaisir J. Significance probability mapping: the final touch in t-statistic mapping. Br Topog 1994; 7: 3-8.

10. Hassainia F, Petit D, Nielsen T, Gauthier S, Montplaisir J. Quantitative EEG and statistical mapping of wakefulness and REM sleep in the evaluation of mild to moderate Alzheimer's disease. Eur Neurol 1997; 7: 219-24.

11. Hazra J. Effect of hemicholinium-3 on slow-wave and paradoxical sleep of cat. Eur J Pharmacol 1970; 11: 395-7.

12. Helkala EL, Laulumaa V, Soininen H, Partanen J, Riekkinen PJ. Different patterns of cognitive decline related to normal or deteriorating EEG in a 3-year follow-up study of patients with Alzheimer's disease. Neurology 1991; 41: 528-32.

13. Hobson JA, McCarley RW, Wyzinski PW. Sleep cycle oscillation: reciprocal discharge by two brainstem neuronal groups. Science 1975; 189: 55-8.

14. Jellinger K. The pedunculopontine nucleus in Parkinson's disease, progressive supranuclear palsy and Alzheimer's disease. J Neurol Neurosurg Psychiatr 1988; 51: 540-3.

15. Jouvet M. Recherches sur les structures nerveuses et les mécanismes responsables des différentes phases du sommeil physiologique. Arch Ital Biol 1962; 100: 125-206.

16. Jouvet M. Biogenic amines and the states of sleep. Science 1969; 163: 32-41.

17. Kish SJ, Chang LJ, Mirchandani L, Shannak K, Hornykiewicz O. Progressive supranuclear palsy: relationship between extrapyramidal disturbances, dementia, and brain neurotransmitter markers. Ann Neurol 1985; 18: 530-6.

18. Kuskowski MA, Mortimer JA, Morley GK, Malone SM, Okaya AJ. Rate of cognitive decline in Alzheimer's disease is associated with EEG alpha power. Biol Psychiatry 1993; 33: 659-62.

19. Maher ER, Smith EM, Lees AJ. Cognitive deficits in the Steele-Richardson-Olszewski syndrome (progressive supranuclear palsy). J Neurol Neurosurg Psychiatr 1985; 48: 1234-9.

20. Mann DMA. Nerve cell protein metabolism and degenerative disease. Neuropath Appl Neurobiol 1982; 8: 161-76.

21. McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan EM. Clinical diagnosis of Alzheimer's disease: Report of 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.

22. Montplaisir J, Petit D, Décary A, Masson H, Bédard M-A, Panisset M, Rémillard G, Gauthier S. Sleep and quantitative EEG in patients with progressive supranuclear palsy. Neurology 1997; 49: 999-1003.

23. Montplaisir J, Petit D, Hassainia F, Gauthier S. EEG brain mapping analyses of Xanomeline effects in Alzheimer's disease. Sleep Res 1995a; 24: 408.

24. Montplaisir J, Petit D, Lorrain D, Gauthier S, Nielsen T. Sleep in Alzheimer's disease: further considerations on the role of brainstem and forebrain cholinergic populations in sleep-wake mechanisms. Sleep 1995b; 18: 145-8.

25. Montplaisir J, Petit D, McNamara D, Gauthier S. Comparisons between SPECT and quantitative EEG measures of cortical impairment in mild to moderate Alzheimer's disease. Eur Neurol 1996; 36: 197-200.

26. Penttila M, Partanen JV, Soininen H, Riekkinen PJ. Quantitative analysis of occipital EEG in different stages of Alzheimer's disease. Electroencephalogr Clin Neurophysiol 1985; 60: 1-6.

27. Petit D, Hassainia F, Gauthier S, Montplaisir J. Brain mapping analyses of Dup996 effects on the EEG in Alzheimer's disease. Sleep Res 1994; 23: 377.

28. Petit D, Lorrain D, Gauthier S, Montplaisir J. Regional spectral analysis of the REM sleep EEG in mild to moderate Alzheimer's disease. Neurobiol Aging 1993a; 14: 141-5.

29. Petit D, Montplaisir J, Lorrain D, Gauthier S. Spectral analysis of the rapid eye movement sleep electroencephalogram in right and left temporal regions: a biological marker of Alzheimer's disease. Ann Neurol 1992; 32: 172-6.

30. Petit D, Montplaisir J, Lorrain D, Gauthier S. THA does not affect sleep or EEG spectral power in Alzheimer's disease. Biol Psychiatry 1993b; 33: 753-4.

31. Prinz PN, Vitaliano P, Vitiello MV, Bokan J, Raskind M, Peskind E, Gerber C. Sleep, EEG and mental functions changes in mild, moderate and severe senile dementia of the Alzheimer's type. Neurobiol Aging 1982; 3: 361-70.

32. Reisberg B, Ferris S, deLeon M, Crook T. Global Deterioration Scale (GDS). Psychopharm Bull 1982; 24: 661-3.

33. Sagales T, Erill S, Domino EF. Effects of repeated doses of scopolamine on the electroencephalographic stages of sleep in normal volunteers. Clin Pharmacol Ther 1975; 18: 727-32.

34. Sakai K, El Mansari M, Lin JS, Zhang JG, Vanni-Mercier G. The posterior hypothalamus in the regulation of wakefulness and paradoxical sleep. In: Mancia M, Marini G, eds. The Diencephalon and Sleep. New York: Raven Press, 1990, pp. 171-98.

35. Sitaram N, Gillin JC. Development and use of pharmacological probes of the CNS in man: evidence of cholinergic abnormalities in primary affective illness. Biol Psychiatry 1980; 15: 925-55.

36. Sloan EP, Fenton GW. EEG power spectra and cognitive change in geriatric psychiatry: a longitudinal study. Electroencephalogr Clin Neurophysiol 1993; 86: 361-7.

37. Steriade M, McCarley RW. Brainstem Control of Wakefulness and Sleep. New York: Plenum Press, 1990: 1-499.

38. Streletz LJ, Reyes PF, Zalewska M, Katz L, Fariello RG. Computer analysis of EEG activity in dementia of the Alzheimer's type and Huntington's disease. Neurobiol Aging 1990; 11: 15-20.

39. Toyoda J, Sakiu K, Kuriharo K. A polygraphic study of the effect of atropine on human nocturnal sleep. Folia Psychiat Neurol Japan 1966; 120: 275-89.

40. Vitiello MV, Prinz PN, Williams DE, Frommlet MS, Ries RK. Sleep disturbances in patients with mild-stage Alzheimer disease. J Gerontol 1990; 45(4): M131-8.

41. Whitehouse PJ, Price DL, Struble RG, Clark AW, Coyle JT, DeLong MR. Alzheimer's disease and senile dementia: loss of neurons in the basal forebrain. Science 1982; 215: 1237-9.

This research was supported by the Medical Research Council of Canada and by the "Fonds de la Recherche en Santé du Québec-Hydro Québec."

Original address of this text :

http://www.sro.org/bin/article.dll?Paper&1412&0&0

Please copy this address to the address bar of your internet browser and press the "enter" key.
(We prefer not to put actual links because often page locations change and then our log files get cluttered with error messages
- if the address does not work try to find it from the homepage of the website in question).