- Email: [email protected]
P300 from normals and adult children of alcoholics
Alcohol. Vol. 4, pp. 301-305. ©Pergamon Journals Ltd., 1987. Printed in the U.S.A.
0741-8329/87 $3.00 + .00
P300 From Normals and Adult Children of Alcoholics I J O H N P O L I C H 2 A N D F L O Y D E. B L O O M
Division o f Preclinical Neuroscience and Endocrinology Scripps Clinic and Research Institute, La Jolla, CA 92037
POLICH, J. AND F. E. BLOOM. P3OOfrom normals and adult children of alcoholics. ALCOHOL 4(4) 301-305, 1987.The P300 event-related brain potential (ERP) was obtained from 24 pairs of undergraduate male subjects. One member of each pair reported having a father who was alcoholic (FH+), the other reported no alcoholic family member (FH-). Pairs were matched on height, weight, academic performance, and personal drinking history. Three auditory task situations were employed which manipulated stimulus discrimation difficulty. All tasks employed 20% target and 80% standard tones randomly presented with the subjects required to move their index finger whenever a target tone was detected. No differences in P300 amplitude or latency were obtained between the groups. FH+ subjects tended to demonstrate decreased amplitudes with increased amounts of reported alcohol consumption but only for the most difficult task. The results of the present study suggest that the relationship between the P300 and the heritability for alcoholism is not yet clear and may be modulated by differences in task requirements, subject populations, and personal drinking history. P300
ERP
Family history
Task difficulty
A variety of evidence suggests that alcoholism is an inherited disease [23]. In particular, children of alcoholic parents raised by nonalcoholic foster parents are at a much greater risk for developing alcoholism than are the biological children of nonalcoholic parents [3, 4, 10]. Recent electrophysiological reports comparing individuals with a family history of alcoholism ( F H + ) to individuals without such a history ( F H - ) have also demonstrated apparent genetic differences in background electroencephalographic (EEG) activity between these two groups: F H + boys were found to have more high-frequency activity compared to F H - male subjects [8]. Additional evidence supporting the contribution of genetic background to neuroelectric activity comes from event-related brain potentials (ERPs) recorded from the scalp. The P300 or P3 component has been used to study possible population differences for susceptibility to alcoholism. When male F H + individuals were compared to F H control subjects (n= 15 pairs) from the general population in an auditory vigilance task with or without an alcohol challenge and after a placebo control condition, the amplitude of the P3 component was reduced for the F H + subjects with no effect observed for the F H - subjects [7]. However, a subsequent sample (n=10 pairs) obtained no overall amplitude differences between groups, although peak latency of the P3 increased with increases in the reported amounts of alcohol
typically consumed by the F H + group. The matched group of F H - subjects demonstrated no effects of alcohol comsumption on P3 latency [14]. A similar relationship between P3 latency and amount of alcohol consumed has been observed in a larger sample (n=24) of male and female F H + college students, with the F H - subjects demonstrating the same, somewhat weaker effect. Amplitude of the P3 was not reliably related to alcohol consumption for either group of this study even though a tendency for the F H - subjects to decline in amplitude with increased amounts was observed [ 19]. Furthermore, an additional study has suggested that P3 amplitude is smaller in the sons of alcoholic fathers compared to age matched controls (n=25) before either subject group begins to drink. N o group differences for P3 latency were found [1]. Thus, family background for alcoholism and personal drinking history both appear to affect the P3 ERP component, albeit with some variation from test group to test group. Because the P3 potential may originate in hippocampal and associated brain regions---neural centers important for learning and memory [ 11,15], and since it is thought to reflect memory updating operations during information processing [5, 6, 13], variations in component amplitude and latency may reflect fundamental individual differences in cognitive capability. This assertion is supported by previous findings relating P3 latency to normal [12,21] and impaired [2, 9, 22]
~The data reported here were presented at the AlcohoFERP Symposium of the Eighth International Conference on Event-Related Potentials of the Brain at Stanford University, June 1986. This research was supported by the J. M. McDonald Foundation and NIAAA grant AA03504 and is publication 4428BCR from the Research Institute of Scripps Clinic. 2Requests for reprints should be addressed to John Polich, Preclinical Neuroscience---BCRl, Scripps Clinic, 10666 N. Torrey Pines Road, La Jolla, CA.
301
302
POLICH AND BLOOM TABLE 1
FREQUENCY 11000 vs 2000 Hz)
MEAN V A L U E S OF VARIABLES USED TO M A T C H I N D I V I D U A L S
EASY INTENSITY (40 vs 60 riB)
HARD INTENSITY 141"2vs 45 dR)
WITH A POSITIVEFAMILYHISTORYFOR ALCOHOLISM(FH+) AND THOSE WITH A NEGATIVE FAMILYHISTORY (FH-) ;14
FH+ Age (years) Height (inches) Weight (pounds) Grade Point Drinks/Occasion
20.4 70.6 160.6 2.9 4.1
FH20.4 70.7 155.5 2.8 4.0
i'4: ! Pz
~
j-
~,
/
~:v
J '::i
),z ij
memory function such that shorter P3 latencies are obtained from individuals with good cognitive function, while longer latencies are associated with individuals exhibiting poor cognitive function. Because neuropsychological studies have found that the amount of alcohol typically consumed is negatively correlated with performance on abstraction tasks for normal sober social drinkers [16-18], differential slowing of P3 latency in normal drinkers [14,19] suggests that memory function is related to alcohol's residual influences and that F H + individuals may be more sensitive to these effects than F H - individuals. If substantiated, such a relationship could provide a means of assessing individuals at risk for the effects of alcohol because their individual P3 latencies would exceed the normative data obtained from F H - individuals. In order to clarify the relationship between cognitive function, alcohol consumption, and familial background for alcoholism, the P3 ERP component was measured in F H + and F H - individuals who reported comparable social drinking patterns in a large sample. In addition, discrimination difficulty was also manipulated within subjects to assess the effects of task processing variables on P3 amplitude and latency with respect to alcohol consumption since a variety of stimulus parameters have been used in previous studies. No alcohol was administered at any time during the data collection procedures. METHOD
Subjects Twenty-four pairs of male subjects were obtained with the F H + member having a father who met the DSM III criteria for alcoholism (serious personal problems caused by alcohol consumption: divorce, j o b loss, arrests for intoxicated driving, participation in an alcohol treatment program, etc.). None of the F H + subjects' mothers were alcoholic. Familial background for alcoholism, personal drinking history, screening for neurologic and psychiatric disorders, etc. were assessed by means of standard survey and individual subject interviews. F H - members were defmed as having no first or second degree alcoholic relatives. Subject pairs were matched for sex, age (18-26 years), height, weight, educational background (all subjects were university undergraduates), academic performance, and on their reported amounts and frequency of alcohol consumption. The subject matching data are summarized in Table 1. The association between F H + and F H - subjects for the reported amount of alcohol typically consumed was quite high (r= .97, p <0.001). Although the correlation between the two groups for reported frequency of alcohol consumption was also strong (r=.42, p<0.05), subjects were more closely
+ 10~
10uV~
IO.V
[) 15o I[~o 450 fioo
ii ibl) 3~)u
~;L/L
Latency (msec)
FIG. i. Superaverages for ERPs from three different auditory tasks (n=24 per group). matched on amount of alcohol consumed since previous reports have suggested that amount rather than frequency of alcohol consumption has the stronger effect on cognitive performance [16--18] and P3 latency [14,19].
Recording Conditions ERPs were elicited with binaural 1000 and 2000 Hz tones at 60 dB SPL or 1000 Hz tones presented at 40, 45, or 60 dB SPL having a 9.9 msec rise/fall and 50 msec plateau times. The tones or intensity changes were presented in a random series once every two seconds. Electroencephalographic activity was recorded at the Fz, Cz, and Pz electrode sites of the 10--20 system using gold-plated electrodes a t ~ e d with electrode paste and tape, referred to linked earlobes with a forehead ground. Additional electrodes were placed at the outer canthus and supraorbitally to the left eye with a bipolar recording made of the EOG. Impedance for all recording sites was 10 Kohms or less. The filter bandpass was 0.5 to 30 Hz (3 dB down, 12 dB octave/slope). The E E G was digitized at 1.5 msec per point for 768 msec with a 75 msec prestimulus baseline. Waveforms were averaged on-line by a Nicolet Pathfinder II which also controlled the stimulus presentation and artifact rejection. Trials on which the E E G or EOG exceeded ---45/xV were automatically rejected. Subjects kept their eyes closed during all ERP recording sessions.
Procedure Each subject participated in three different task situations designed to assess P3 latency and amplitude changes due to stimulus discrimination difficulty. The frequency discrimination task consisted of 1000 and 2000 Hz tones at 60 dB SPL with 80% and 20% probabilities, respectively. Subjects were required to move the index finger of their fight hand whenever they detected the infrequent, randomly occurring high tone (2000 Hz) which was designated as the target tone. Two successive blocks of trials were presented such that 20
P300 AND FAMILY HISTORY
303
400 FHN
T
350
300 15
-i-
after the N 1-P2-N2 complex between 250 and 400 msec was designated as the P3 component. Amplitude was measured relative to the prestimulus baseline with peak latency defined as the time point of maximum positive amplitude. These data points were then employed for statistical evaluation with analysis of variance procedures. Task performance was virtually perfect for the frequency and easy intensity discrimination tasks with 0.4% and 0.5% of the target trials misperceived, respectively. The hard intensity discrimination task yielded a 8.9% error rate for the target trials. F H + tended to produce fewer errors than F H - subjects (5.6 vs. 12.1%), but these group differences were not reliable, F(1,46)=3.7, p =0.06. Thus, ERPs were comprised of essentially all correctly processed stimulus trials with errors of the hard intensity discrimination task distributed over all pertinent subject conditions. Discrimination Tasks
Frequency Easy-intensity (lOOOvs 2000 Hz) (40 vs 60 dB)
Hard-Intensity (40 vs 45 dB)
FIG. 2. Mean (and 1 standard error) P3 amplitudes and latencies from the Pz electrode site for each FH group.
target tone trials comprised the ERP for each block. These waveforms were then averaged together to obtain the waveform used in subsequent analyses. All subjects received the frequency discrimination task first. Subjects were then presented with the intensity discrimination tasks. The easy intensity discrimination task consisted of a series of 1000 Hz tones such that 80% were 40 dB SPL and 20% were 60 dB SPL, with the latter occurring randomly throughout the sequence and designated as the target tone. The hard intensity discrimination task consisted of 1000 Hz tones presented in a series such that 80% were 40 dB and 20% were 45 dB, with the latter occurring randomly throughout the series and designated as the target tone. Subjects were instructed to move the index finger of their right hand whenever the target stimulus was detected. The order of the easy and hard task presentations was counterbalanced across subjects so that half the subjects of each family history group received the easy task first, while half received the hard task first. One block of trials was presented for each intensity condition yielding an ERP waveform comprised of the 20 target trials for that stimulus set. Before each condition, subjects practiced the task to ensure accurate performance. Rest periods were provided during the intervals between tasks in which subjects were allowed to stand and move about to avoid fatigue. All subjects were able to complete all procedures within a two hour session. RESULTS
P3 Measurement
Waveforms from each task condition were measured in the same fashion: the largest positive-going peak occurring
The grand ERP averages for each condition and subject group are presented in Fig. 1, with the mean P3 latencies and amplitudes (plus one standard error) for each task presented in Fig. 2. Each task was analyzed separately with a twofactor (Family History × Electrode Site) analysis of variance. For the frequency discrimination task, the amplitude data demonstrated no differences for FH variable, F(1,46)
304
POL1CH AND BLOOM
FH+ N=24
400 ~
400
35° ~-77~-.-. . . . . . . . . . . . .
350
300
300
250 ~-
[......
r 200
t .... I 2
Easy Intensity I Hard Intensity ]
t
L
4
6
demonstrated a significant correlation (r .... .46, p <0.02) between P3 amplitude and the amount of alcohol consumed for the hard intensity task relative to the two other stimulus discrimination paradigms. The analogous correlation for the F H - subjects was not significant ( r = - . 2 6 , p>0.20), nor were these correlations significantly different from one another (z=.703, p =0.76).
FHN=24
I 8
DISCUSSION
2501 J 200 10
I
20
20
15
15 I
10
10
i
I
--.....
............
E O'3 Ct-
5 0
I
I
I
I
2
,¢
6
8
lJg
0
I
I
I
J
J
2
4
6
8
10
NUMBER DRINKS/OCCASION
FIG. 3. Regression lines for each FH subject group from each task condition for P3 latency and amplitude plotted as a function of the number of drinks per drinking occasion. P3 amplitude for F H were obtained, F(1,46)<1,0, p=0.92. Both groups demonstrated an increase in P3 amplitude across the midline electrodes (FH +: 7.5, 8.6, 8.9/xV; F H - : 7.6, 8.8, 8.8/zV), F(2,92)=6.3, p<0.01, but no interaction between family history and electrode position was observed. The latency data demonstrated no effects for FH, F(1,46)<1.0, p=0.48, or electrode position for either group ( F H + : 348, 349,346 msec; F H - : 358,354,349 msec). Additional analysis performed on the data obtained from each recording site confirmed these results for both the amplitude and latency measures from the hard intensity discrimination task. Alcohol Consumption
The relationship between amount of alcohol typically consumed and P3 latency and amplitude was assessed by correlating these data separately for each family history group and task condition. The association between P3 latency and amplitude and amount of reported alcohol typically consumed for all three discrimination tasks were strongest at the Cz electrode site and are summarized in Fig. 3. When all three tasks are compared across both subject groups and dependent variables, only F H + individuals
No main effects for either latency or amplitude were obtained between the family history groups for any of the tasks employed in the present study. The primary finding was that P3 amplitude (and to a lesser extent P3 latency) were affected by the amount of alcohol typically consumed more for individuals who have a positive family history compared to individuals with a negative family history for alcoholism, but only when the task situation employed to elicit the P3 component was relatively difficult. For easier tasks, very little effect of alcohol consumption of P3 latency or amplitude was found. Previous reports have found that the F H + group displayed smaller amplitudes [1] or increased latencies when amount of alcohol consumption was considered [14,19] relative to the F H - control subjects. Although task difficulty appears to affect P3 responsivity across studies [6], only in the Begleiter et al. (1984) report were group differences pronounced. This study employed a difficult discrimination task involving a visual mental rotation paradigm coupled with the comparatively very young subjects (mean age= 12.2 years) from the general population. Hence, the results may reflect the operation of task and/or subject parameters quite different from those employed in the other investigations of general population subjects [7,14] or those which typically assessed college students with auditory stimuli [9]. Indeed, differences in subject populations, tasks, and recording procedures could contribute to the inconsistencies obtained across investigations. The minimal latency and decreased amplitude changes observed with reported increased alcohol consumption in the present study for both subject groups suggests that P3 may be somewhat sensitive to the cognitive decline associated with alcohol consumption [ 16--18]. The nature of the eliciting discrimination task, however, appears to be an important variable when the P3 ERP component is employed for this purpose. It is also unclear how different populations of subjects who may vary in their general degree of cognitive capability will affect family history group comparisons. Because P3 latency has been shown to be systematically correlated with general cognitive functioning in normal and impaired adults [21,22] and children [12], it is reasonable to suppose that different subject groups may yield different P3 latency/amplitude functions for the amount of alcohol typically consumed. Thus, the variegated findings obtained to date could very well be reflecting an interaction between task differences and fundamental population variables which have not yet been clearly delineated with ERPs.
REFERENCES
1. Begleiter, H., B. Porjesz, B. Bihari and B. Kissin. Event-related brain potentials in boys at risk for alcoholism. Science 225: 1493-1496, 1984. 2. Brown, W. S., J. T. Marsh and A. La Rue. Event-related potentials in psychiatry: Differentiating depression and dementia in the elderly. Bull LA Neurol Soc 47: 91-107, 1985.
3. Cloninger, R. C., M. Bohman and S. Sigvardsson. Inheritance of alcohol abuse. Arch Gen Psychiatry 38: 861-868, 1981. 4. Cotton, N. S. The familial incidence of alcoholism. J Stud Ah:ohol 40: 89-116, 1979. 5. Donchin, E. Surprise! . . . Surprise? Psychophysiology 18: 493-513, 1981.
P300 A N D F A M I L Y H I S T O R Y 6. Donchin, E., W. Ritter and C. McCailum. Cognitive psychophysiology: The endogenous components of the ERP. In: Brain Event-Related Potentials in Man, edited by E. Callaway, P. Tueting and S. Koslow. New York: Academic Press, 1978, pp. 439-441. 7. Elmasian, R., H. Neville, D. Woods, M. Schuckit and F. E. Bloom. Event-related brain potentials are different in individuals at high and low risk for developing alcoholism. Proc Natl Acad Sci USA 79: 7900-7903, 1982. 8. Gabrielli, W. F., S. A. Mednick, J. Volavka, V. E. Pollock, F. Schulsinger and T. M. Itil. Electroencephalograms in children of alcoholic fathers. Psychophysiology 19: 404--407, 1982. 9. Goodin, D., K. Squires and A. Starr. Long-latency eventrelated components of the auditory evoked potential in dementia. Brain 101: 635-648, 1978. 10. Goodwin, D. W. Alcoholism and hereditary. Arch Gen Psychiatry 36: 57-61, 1979. 11. Halgren, E., N. Squires, C. Wilson, J. Rohrbaugh, T. Bab and P. Crandall. Endogenous potentials generated in the human hippocampai formation and amygdala by infrequent events. Science 210: 803--805, 1980. 12. Howard, L. and J. Polich. P300 latency and memory span development. Dev Psychol 21" 283-289, 1985. 13. Karis, D., M. Fabiani and E. Donchin. "P300" and Memory: Individual differences in the yon Restorff effect. Cog Psychol 16: 177-216, 1984.
305 14. Neville, H. J. and A. L. Schmidt. Event-related brain potentials in subjects at a risk for alcoholism. In: Early Identification of Alcohol Abuse, edited by N. Chang and H. Chao (Res. Monogr. 17). Rockville, MD: National Inst. on Alc. Abuse and Alcoholism, 1985. 15. Okada, Y. C., L. Kaufman and S. J. Williamson. The hippocampal formation as a source of the slow endogenous potentials. Electroencephalogr Clin Neurophysiol 55: 417-426, 1983. 16. Parker, E. S. and E. P. Noble. Alcohol consumption and cognitive functioning in social drinkers. J Stud Alcohol 38: 12241232, 1977. 17. Parker, E. S., I. M. Birnbaum, R. Boyd and E. F. Noble. Neuropsychological decrements as a function of alcohol intake in male students. Alcohol: Clin Exp Res 4: 330-334, 1980. 18. Parker, D. A., E. S. Parker, J. A. Brody and R. Schoenberg. Alcohol use and cognitive loss among employed men and women. Am J Public Health 73: 521-526, 1983. 19. Polich, J. P300 latency reflects personal drinking history. Psychophysiology 21: 592-593, 1984. 20. Polich, J. and F. E. Bloom. P300 and alcohol consumption in normals and individuals at risk for alcoholism. Prog NeuroPsychopharrnacol Biol Psychiatr 10: 201-210, 1986. 21. Polich, J., L. Howard and A. Start. P300 latency correlates with digit span. Psychophysiology 20: 665-669, 1983. 22. Polich, J., C. E. Ehlers, S. Otis, A. J. Mandell and F. E. Bloom. P300 reflects the degree of cognitive decline in dementing illness. Electroencephalogr Clin Neurophysiol 63: 138--144, 1986. 23. Schuckit, M. A. Alcoholism and genetics: possible biological mediators. Biol Psychiatry 15: 437-447, 1980.
0741-8329/87 $3.00 + .00
P300 From Normals and Adult Children of Alcoholics I J O H N P O L I C H 2 A N D F L O Y D E. B L O O M
Division o f Preclinical Neuroscience and Endocrinology Scripps Clinic and Research Institute, La Jolla, CA 92037
POLICH, J. AND F. E. BLOOM. P3OOfrom normals and adult children of alcoholics. ALCOHOL 4(4) 301-305, 1987.The P300 event-related brain potential (ERP) was obtained from 24 pairs of undergraduate male subjects. One member of each pair reported having a father who was alcoholic (FH+), the other reported no alcoholic family member (FH-). Pairs were matched on height, weight, academic performance, and personal drinking history. Three auditory task situations were employed which manipulated stimulus discrimation difficulty. All tasks employed 20% target and 80% standard tones randomly presented with the subjects required to move their index finger whenever a target tone was detected. No differences in P300 amplitude or latency were obtained between the groups. FH+ subjects tended to demonstrate decreased amplitudes with increased amounts of reported alcohol consumption but only for the most difficult task. The results of the present study suggest that the relationship between the P300 and the heritability for alcoholism is not yet clear and may be modulated by differences in task requirements, subject populations, and personal drinking history. P300
ERP
Family history
Task difficulty
A variety of evidence suggests that alcoholism is an inherited disease [23]. In particular, children of alcoholic parents raised by nonalcoholic foster parents are at a much greater risk for developing alcoholism than are the biological children of nonalcoholic parents [3, 4, 10]. Recent electrophysiological reports comparing individuals with a family history of alcoholism ( F H + ) to individuals without such a history ( F H - ) have also demonstrated apparent genetic differences in background electroencephalographic (EEG) activity between these two groups: F H + boys were found to have more high-frequency activity compared to F H - male subjects [8]. Additional evidence supporting the contribution of genetic background to neuroelectric activity comes from event-related brain potentials (ERPs) recorded from the scalp. The P300 or P3 component has been used to study possible population differences for susceptibility to alcoholism. When male F H + individuals were compared to F H control subjects (n= 15 pairs) from the general population in an auditory vigilance task with or without an alcohol challenge and after a placebo control condition, the amplitude of the P3 component was reduced for the F H + subjects with no effect observed for the F H - subjects [7]. However, a subsequent sample (n=10 pairs) obtained no overall amplitude differences between groups, although peak latency of the P3 increased with increases in the reported amounts of alcohol
typically consumed by the F H + group. The matched group of F H - subjects demonstrated no effects of alcohol comsumption on P3 latency [14]. A similar relationship between P3 latency and amount of alcohol consumed has been observed in a larger sample (n=24) of male and female F H + college students, with the F H - subjects demonstrating the same, somewhat weaker effect. Amplitude of the P3 was not reliably related to alcohol consumption for either group of this study even though a tendency for the F H - subjects to decline in amplitude with increased amounts was observed [ 19]. Furthermore, an additional study has suggested that P3 amplitude is smaller in the sons of alcoholic fathers compared to age matched controls (n=25) before either subject group begins to drink. N o group differences for P3 latency were found [1]. Thus, family background for alcoholism and personal drinking history both appear to affect the P3 ERP component, albeit with some variation from test group to test group. Because the P3 potential may originate in hippocampal and associated brain regions---neural centers important for learning and memory [ 11,15], and since it is thought to reflect memory updating operations during information processing [5, 6, 13], variations in component amplitude and latency may reflect fundamental individual differences in cognitive capability. This assertion is supported by previous findings relating P3 latency to normal [12,21] and impaired [2, 9, 22]
~The data reported here were presented at the AlcohoFERP Symposium of the Eighth International Conference on Event-Related Potentials of the Brain at Stanford University, June 1986. This research was supported by the J. M. McDonald Foundation and NIAAA grant AA03504 and is publication 4428BCR from the Research Institute of Scripps Clinic. 2Requests for reprints should be addressed to John Polich, Preclinical Neuroscience---BCRl, Scripps Clinic, 10666 N. Torrey Pines Road, La Jolla, CA.
301
302
POLICH AND BLOOM TABLE 1
FREQUENCY 11000 vs 2000 Hz)
MEAN V A L U E S OF VARIABLES USED TO M A T C H I N D I V I D U A L S
EASY INTENSITY (40 vs 60 riB)
HARD INTENSITY 141"2vs 45 dR)
WITH A POSITIVEFAMILYHISTORYFOR ALCOHOLISM(FH+) AND THOSE WITH A NEGATIVE FAMILYHISTORY (FH-) ;14
FH+ Age (years) Height (inches) Weight (pounds) Grade Point Drinks/Occasion
20.4 70.6 160.6 2.9 4.1
FH20.4 70.7 155.5 2.8 4.0
i'4: ! Pz
~
j-
~,
/
~:v
J '::i
),z ij
memory function such that shorter P3 latencies are obtained from individuals with good cognitive function, while longer latencies are associated with individuals exhibiting poor cognitive function. Because neuropsychological studies have found that the amount of alcohol typically consumed is negatively correlated with performance on abstraction tasks for normal sober social drinkers [16-18], differential slowing of P3 latency in normal drinkers [14,19] suggests that memory function is related to alcohol's residual influences and that F H + individuals may be more sensitive to these effects than F H - individuals. If substantiated, such a relationship could provide a means of assessing individuals at risk for the effects of alcohol because their individual P3 latencies would exceed the normative data obtained from F H - individuals. In order to clarify the relationship between cognitive function, alcohol consumption, and familial background for alcoholism, the P3 ERP component was measured in F H + and F H - individuals who reported comparable social drinking patterns in a large sample. In addition, discrimination difficulty was also manipulated within subjects to assess the effects of task processing variables on P3 amplitude and latency with respect to alcohol consumption since a variety of stimulus parameters have been used in previous studies. No alcohol was administered at any time during the data collection procedures. METHOD
Subjects Twenty-four pairs of male subjects were obtained with the F H + member having a father who met the DSM III criteria for alcoholism (serious personal problems caused by alcohol consumption: divorce, j o b loss, arrests for intoxicated driving, participation in an alcohol treatment program, etc.). None of the F H + subjects' mothers were alcoholic. Familial background for alcoholism, personal drinking history, screening for neurologic and psychiatric disorders, etc. were assessed by means of standard survey and individual subject interviews. F H - members were defmed as having no first or second degree alcoholic relatives. Subject pairs were matched for sex, age (18-26 years), height, weight, educational background (all subjects were university undergraduates), academic performance, and on their reported amounts and frequency of alcohol consumption. The subject matching data are summarized in Table 1. The association between F H + and F H - subjects for the reported amount of alcohol typically consumed was quite high (r= .97, p <0.001). Although the correlation between the two groups for reported frequency of alcohol consumption was also strong (r=.42, p<0.05), subjects were more closely
+ 10~
10uV~
IO.V
[) 15o I[~o 450 fioo
ii ibl) 3~)u
~;L/L
Latency (msec)
FIG. i. Superaverages for ERPs from three different auditory tasks (n=24 per group). matched on amount of alcohol consumed since previous reports have suggested that amount rather than frequency of alcohol consumption has the stronger effect on cognitive performance [16--18] and P3 latency [14,19].
Recording Conditions ERPs were elicited with binaural 1000 and 2000 Hz tones at 60 dB SPL or 1000 Hz tones presented at 40, 45, or 60 dB SPL having a 9.9 msec rise/fall and 50 msec plateau times. The tones or intensity changes were presented in a random series once every two seconds. Electroencephalographic activity was recorded at the Fz, Cz, and Pz electrode sites of the 10--20 system using gold-plated electrodes a t ~ e d with electrode paste and tape, referred to linked earlobes with a forehead ground. Additional electrodes were placed at the outer canthus and supraorbitally to the left eye with a bipolar recording made of the EOG. Impedance for all recording sites was 10 Kohms or less. The filter bandpass was 0.5 to 30 Hz (3 dB down, 12 dB octave/slope). The E E G was digitized at 1.5 msec per point for 768 msec with a 75 msec prestimulus baseline. Waveforms were averaged on-line by a Nicolet Pathfinder II which also controlled the stimulus presentation and artifact rejection. Trials on which the E E G or EOG exceeded ---45/xV were automatically rejected. Subjects kept their eyes closed during all ERP recording sessions.
Procedure Each subject participated in three different task situations designed to assess P3 latency and amplitude changes due to stimulus discrimination difficulty. The frequency discrimination task consisted of 1000 and 2000 Hz tones at 60 dB SPL with 80% and 20% probabilities, respectively. Subjects were required to move the index finger of their fight hand whenever they detected the infrequent, randomly occurring high tone (2000 Hz) which was designated as the target tone. Two successive blocks of trials were presented such that 20
P300 AND FAMILY HISTORY
303
400 FHN
T
350
300 15
-i-
after the N 1-P2-N2 complex between 250 and 400 msec was designated as the P3 component. Amplitude was measured relative to the prestimulus baseline with peak latency defined as the time point of maximum positive amplitude. These data points were then employed for statistical evaluation with analysis of variance procedures. Task performance was virtually perfect for the frequency and easy intensity discrimination tasks with 0.4% and 0.5% of the target trials misperceived, respectively. The hard intensity discrimination task yielded a 8.9% error rate for the target trials. F H + tended to produce fewer errors than F H - subjects (5.6 vs. 12.1%), but these group differences were not reliable, F(1,46)=3.7, p =0.06. Thus, ERPs were comprised of essentially all correctly processed stimulus trials with errors of the hard intensity discrimination task distributed over all pertinent subject conditions. Discrimination Tasks
Frequency Easy-intensity (lOOOvs 2000 Hz) (40 vs 60 dB)
Hard-Intensity (40 vs 45 dB)
FIG. 2. Mean (and 1 standard error) P3 amplitudes and latencies from the Pz electrode site for each FH group.
target tone trials comprised the ERP for each block. These waveforms were then averaged together to obtain the waveform used in subsequent analyses. All subjects received the frequency discrimination task first. Subjects were then presented with the intensity discrimination tasks. The easy intensity discrimination task consisted of a series of 1000 Hz tones such that 80% were 40 dB SPL and 20% were 60 dB SPL, with the latter occurring randomly throughout the sequence and designated as the target tone. The hard intensity discrimination task consisted of 1000 Hz tones presented in a series such that 80% were 40 dB and 20% were 45 dB, with the latter occurring randomly throughout the series and designated as the target tone. Subjects were instructed to move the index finger of their right hand whenever the target stimulus was detected. The order of the easy and hard task presentations was counterbalanced across subjects so that half the subjects of each family history group received the easy task first, while half received the hard task first. One block of trials was presented for each intensity condition yielding an ERP waveform comprised of the 20 target trials for that stimulus set. Before each condition, subjects practiced the task to ensure accurate performance. Rest periods were provided during the intervals between tasks in which subjects were allowed to stand and move about to avoid fatigue. All subjects were able to complete all procedures within a two hour session. RESULTS
P3 Measurement
Waveforms from each task condition were measured in the same fashion: the largest positive-going peak occurring
The grand ERP averages for each condition and subject group are presented in Fig. 1, with the mean P3 latencies and amplitudes (plus one standard error) for each task presented in Fig. 2. Each task was analyzed separately with a twofactor (Family History × Electrode Site) analysis of variance. For the frequency discrimination task, the amplitude data demonstrated no differences for FH variable, F(1,46)
304
POL1CH AND BLOOM
FH+ N=24
400 ~
400
35° ~-77~-.-. . . . . . . . . . . . .
350
300
300
250 ~-
[......
r 200
t .... I 2
Easy Intensity I Hard Intensity ]
t
L
4
6
demonstrated a significant correlation (r .... .46, p <0.02) between P3 amplitude and the amount of alcohol consumed for the hard intensity task relative to the two other stimulus discrimination paradigms. The analogous correlation for the F H - subjects was not significant ( r = - . 2 6 , p>0.20), nor were these correlations significantly different from one another (z=.703, p =0.76).
FHN=24
I 8
DISCUSSION
2501 J 200 10
I
20
20
15
15 I
10
10
i
I
--.....
............
E O'3 Ct-
5 0
I
I
I
I
2
,¢
6
8
lJg
0
I
I
I
J
J
2
4
6
8
10
NUMBER DRINKS/OCCASION
FIG. 3. Regression lines for each FH subject group from each task condition for P3 latency and amplitude plotted as a function of the number of drinks per drinking occasion. P3 amplitude for F H were obtained, F(1,46)<1,0, p=0.92. Both groups demonstrated an increase in P3 amplitude across the midline electrodes (FH +: 7.5, 8.6, 8.9/xV; F H - : 7.6, 8.8, 8.8/zV), F(2,92)=6.3, p<0.01, but no interaction between family history and electrode position was observed. The latency data demonstrated no effects for FH, F(1,46)<1.0, p=0.48, or electrode position for either group ( F H + : 348, 349,346 msec; F H - : 358,354,349 msec). Additional analysis performed on the data obtained from each recording site confirmed these results for both the amplitude and latency measures from the hard intensity discrimination task. Alcohol Consumption
The relationship between amount of alcohol typically consumed and P3 latency and amplitude was assessed by correlating these data separately for each family history group and task condition. The association between P3 latency and amplitude and amount of reported alcohol typically consumed for all three discrimination tasks were strongest at the Cz electrode site and are summarized in Fig. 3. When all three tasks are compared across both subject groups and dependent variables, only F H + individuals
No main effects for either latency or amplitude were obtained between the family history groups for any of the tasks employed in the present study. The primary finding was that P3 amplitude (and to a lesser extent P3 latency) were affected by the amount of alcohol typically consumed more for individuals who have a positive family history compared to individuals with a negative family history for alcoholism, but only when the task situation employed to elicit the P3 component was relatively difficult. For easier tasks, very little effect of alcohol consumption of P3 latency or amplitude was found. Previous reports have found that the F H + group displayed smaller amplitudes [1] or increased latencies when amount of alcohol consumption was considered [14,19] relative to the F H - control subjects. Although task difficulty appears to affect P3 responsivity across studies [6], only in the Begleiter et al. (1984) report were group differences pronounced. This study employed a difficult discrimination task involving a visual mental rotation paradigm coupled with the comparatively very young subjects (mean age= 12.2 years) from the general population. Hence, the results may reflect the operation of task and/or subject parameters quite different from those employed in the other investigations of general population subjects [7,14] or those which typically assessed college students with auditory stimuli [9]. Indeed, differences in subject populations, tasks, and recording procedures could contribute to the inconsistencies obtained across investigations. The minimal latency and decreased amplitude changes observed with reported increased alcohol consumption in the present study for both subject groups suggests that P3 may be somewhat sensitive to the cognitive decline associated with alcohol consumption [ 16--18]. The nature of the eliciting discrimination task, however, appears to be an important variable when the P3 ERP component is employed for this purpose. It is also unclear how different populations of subjects who may vary in their general degree of cognitive capability will affect family history group comparisons. Because P3 latency has been shown to be systematically correlated with general cognitive functioning in normal and impaired adults [21,22] and children [12], it is reasonable to suppose that different subject groups may yield different P3 latency/amplitude functions for the amount of alcohol typically consumed. Thus, the variegated findings obtained to date could very well be reflecting an interaction between task differences and fundamental population variables which have not yet been clearly delineated with ERPs.
REFERENCES
1. Begleiter, H., B. Porjesz, B. Bihari and B. Kissin. Event-related brain potentials in boys at risk for alcoholism. Science 225: 1493-1496, 1984. 2. Brown, W. S., J. T. Marsh and A. La Rue. Event-related potentials in psychiatry: Differentiating depression and dementia in the elderly. Bull LA Neurol Soc 47: 91-107, 1985.
3. Cloninger, R. C., M. Bohman and S. Sigvardsson. Inheritance of alcohol abuse. Arch Gen Psychiatry 38: 861-868, 1981. 4. Cotton, N. S. The familial incidence of alcoholism. J Stud Ah:ohol 40: 89-116, 1979. 5. Donchin, E. Surprise! . . . Surprise? Psychophysiology 18: 493-513, 1981.
P300 A N D F A M I L Y H I S T O R Y 6. Donchin, E., W. Ritter and C. McCailum. Cognitive psychophysiology: The endogenous components of the ERP. In: Brain Event-Related Potentials in Man, edited by E. Callaway, P. Tueting and S. Koslow. New York: Academic Press, 1978, pp. 439-441. 7. Elmasian, R., H. Neville, D. Woods, M. Schuckit and F. E. Bloom. Event-related brain potentials are different in individuals at high and low risk for developing alcoholism. Proc Natl Acad Sci USA 79: 7900-7903, 1982. 8. Gabrielli, W. F., S. A. Mednick, J. Volavka, V. E. Pollock, F. Schulsinger and T. M. Itil. Electroencephalograms in children of alcoholic fathers. Psychophysiology 19: 404--407, 1982. 9. Goodin, D., K. Squires and A. Starr. Long-latency eventrelated components of the auditory evoked potential in dementia. Brain 101: 635-648, 1978. 10. Goodwin, D. W. Alcoholism and hereditary. Arch Gen Psychiatry 36: 57-61, 1979. 11. Halgren, E., N. Squires, C. Wilson, J. Rohrbaugh, T. Bab and P. Crandall. Endogenous potentials generated in the human hippocampai formation and amygdala by infrequent events. Science 210: 803--805, 1980. 12. Howard, L. and J. Polich. P300 latency and memory span development. Dev Psychol 21" 283-289, 1985. 13. Karis, D., M. Fabiani and E. Donchin. "P300" and Memory: Individual differences in the yon Restorff effect. Cog Psychol 16: 177-216, 1984.
305 14. Neville, H. J. and A. L. Schmidt. Event-related brain potentials in subjects at a risk for alcoholism. In: Early Identification of Alcohol Abuse, edited by N. Chang and H. Chao (Res. Monogr. 17). Rockville, MD: National Inst. on Alc. Abuse and Alcoholism, 1985. 15. Okada, Y. C., L. Kaufman and S. J. Williamson. The hippocampal formation as a source of the slow endogenous potentials. Electroencephalogr Clin Neurophysiol 55: 417-426, 1983. 16. Parker, E. S. and E. P. Noble. Alcohol consumption and cognitive functioning in social drinkers. J Stud Alcohol 38: 12241232, 1977. 17. Parker, E. S., I. M. Birnbaum, R. Boyd and E. F. Noble. Neuropsychological decrements as a function of alcohol intake in male students. Alcohol: Clin Exp Res 4: 330-334, 1980. 18. Parker, D. A., E. S. Parker, J. A. Brody and R. Schoenberg. Alcohol use and cognitive loss among employed men and women. Am J Public Health 73: 521-526, 1983. 19. Polich, J. P300 latency reflects personal drinking history. Psychophysiology 21: 592-593, 1984. 20. Polich, J. and F. E. Bloom. P300 and alcohol consumption in normals and individuals at risk for alcoholism. Prog NeuroPsychopharrnacol Biol Psychiatr 10: 201-210, 1986. 21. Polich, J., L. Howard and A. Start. P300 latency correlates with digit span. Psychophysiology 20: 665-669, 1983. 22. Polich, J., C. E. Ehlers, S. Otis, A. J. Mandell and F. E. Bloom. P300 reflects the degree of cognitive decline in dementing illness. Electroencephalogr Clin Neurophysiol 63: 138--144, 1986. 23. Schuckit, M. A. Alcoholism and genetics: possible biological mediators. Biol Psychiatry 15: 437-447, 1980.