The distribution of illicit drug use: correlations between extent of use, heavy use and problems

Sections

Method
Results
Discussion
Bibliography

Details

Author: Reginald G. SMART
Pages: 33 to 41
Creation Date: 1978/01/01

The distribution of illicit drug use: correlations between extent of use, heavy use and problems

Reginald G. SMART
Addiction Research Foundation, 33, Russell Street, Toronto, Canada

A great deal of research has shown that the distribution of per capita alcohol consumption is continuous, unimodal and log normal (see Bruun et al., 1975, for a review). This involves a distribution with many light users, fewer moderate users and even fewer heavy users. It has also been shown that certain alcohol problems such as liver cirrhosis are more common in countries where per capita alcohol consumption is high. These findings have important implications for the prevention of alcohol problems. They suggest that in order to reduce problems such as liver cirrhosis it will be necessary to reduce per capita alcohol consumption.

Data on the distribution of drug use is less extensive. However, studies by Smart et al. (1971) have shown that composite drug use scores among high school students in a variety of Canadian cities and at several points in time described a log normal distribution. Later studies showed that the same distribution held for Canadian adults and British university students (Smart and Whitehead, 1973) and for the use of individual drugs in several cities (Smart and Whitehead, 1972). More recently, McDermott and Scheurich (1977) found the same distribution among adults in Kansas. What has not been done so far is to establish points in the distribution of drug use above which some psychological or physical pathology occurs. Schmidt and DeLint (1970) have shown that about 15 cl of absolute alcohol is consumed by alcoholics and that regular consumption of 10 cl per day is associated with pathologies such as liver cirrhosis. Smart et al. (1971) argued earlier that "some such damaging level of drug consumption will also have to be determined in order to give meaning to the term 'drug abusers' ". The present paper reports a study of the distribution of illicit drug use in 7 areas of Ontario, the correlation between per capita or average drug scores and percentage of heavy users and problem drug users. A scale for the measurement of problem drug use has been developed which attempts to assess the student's problems with drugs by asking about police arrests, treatment for drug use and concern about over use of drugs on the part of the student and his parents.

It was predicted that:

  1. The distributions of illicit drug use scores in the 7 areas would be continuous, unimodal and log normal.

  2. That the correlation between the frequency of illicit drug use and drug problems would be positive: the more drugs a student used the more likely he was to have drug problems.

  3. That areas in which there was more illicit drug use would also have high proportions of heavy users and problem drug users.

Method

Sampling

A survey of drug use was made among 4,687 students in Ontario. The details of the sampling methods are described in the report by Smart and Goodstadt (1977) only a few of which are relevant here. The sampling was done in grades 7, 9, 11 and 13 in all areas of Ontario. From each of 7 districts, Boards of Education were selected at random with their numbers proportionate to the total Ontario student enrolment. Schools were chosen from selected boards by a random method. The selection of classes was made in consultation with the school principal to be representative of all students in a grade.

In general, the sampling methods gave a large representative sample of Ontario students, except for one area of the province where severe weather conditions decreased the expected sample size. Some schools were closed on the day of the survey in mid-western Ontario.

Most schools required parental consent for students to participate, but some did not. Over-all, 70 per cent of the students contacted did participate. There were no significant differences in alcohol or tobacco use between the two types of schools. However, students in the ones requiring parental consent reported a slightly lower rate of marijuana use.

Questionnaire

Questions were asked on an anonymous questionnaire. The questionnaire covered the following areas:

  1. Demographic characteristics (e.g., age, sex, occupation of parents).

  2. The frequency of use of 16 drugs in the past year including alcohol, tobacco, cannabis, LSD and other hallucinogens, glue and other solvents, tranquillizers, stimulants and sleeping pills with and without prescription, heroin, "speed" (generally understood to mean methamphetamines) and cocaine.

  3. The availability of certain drugs.

  4. Exposure to educational materials about drugs.

The latter two areas will be reported on at a later date.

Illicit drug use scores

Illicit drug use scores were computed for each student who used any of the following nine drugs during the past year: cannabis, barbiturates, stimulants or tranquillizers with a prescription, heroin, "speed", LSD, other hallucinogens or cocaine. Total scores varied between 0 and 54 with the following scores given for each:

0 = did not use
4 = used 10 to 19 times
1 = used 1 or 2 times
5 = used 20 to 39 times
2 = used 3 to 5 times
6 = used 40 or more times
3 = used 6 to 9 times
 

Drug problem scores

Problem drug scores were computed only for student users (i.e., those with non-zero illicit drug use scores). The score varied from 0 to 4 according to the number of positive responses to the following questions:

  1. Have you ever been arrested or warned by the police because of your use of a drug other than alcohol?

  2. Have you ever seen a doctor, talked to a school counsellor (nurse, teacher) or been in a hospital because of your use of a drug other than alcohol?

  3. Do your parents think that you use drugs other than alcohol too often?

  4. Do you wish you could use drugs less than you do now?

FIGURE 1 - Distribution of drug use scores for total Ontario population.

Full size image: 16 kB, FIGURE 1 - Distribution of drug use scores for total Ontario population.

Results

Distributions of illicit drug use scores

Of the total sample (n = 4,687) 1,309 or 27.9 per cent had used an illicit drug and hence had an illicit drug score varying from 1 to 54. The distribution of scores in the total population of illicit drug users is shown in figure 1. It can be seen that the distribution is uni-modal and continuous; the skewness value is __.01 (Croxton and Cowden, 1955). The distributions for the various areas are shown in figures 2 to 8. They, too, are unimodal and continuous; the skewness values vary from - .14 to .12 and hence the distributions are log normal.

FIGURE 2 - Distribution of drug use scores for northern Ontario students.

Full size image: 11 kB, FIGURE 2 - Distribution of drug use scores for northern Ontario students.

FIGURE 3 - Distribution of drug use scores for western Ontario students.

Full size image: 10 kB, FIGURE 3 - Distribution of drug use scores for western Ontario students.

Drug problem scores

Of 1,309 students who used an illicit drug 296 or 22.6 per cent had a drug problem score of 1 to 4. Most (93 per cent) had scores of 1 or 2. Only 6.32 per cent of the total sample had a higher drug problem score.

The relationship between illicit drug scores and drug problem scores is shown in the table (p. 40). It can be seen that students with a drug problem score of 0 had a mean illicit drug use score of 4.20, those with a score of 1-8.04, 2-12.66, and 3-14.11. Only one student had a problem score of 4 with an illicit score of 7.00. Apart from this one student, the higher the problem score the higher the frequency of illicit drug use.

FIGURE 4 - Distribution of drug use scores for mid-western Ontario students.

Full size image: 10 kB, FIGURE 4 - Distribution of drug use scores for mid-western Ontario students.

FIGURE 5 - Distribution of drug use scores for Niagara region students.

Full size image: 11 kB, FIGURE 5 - Distribution of drug use scores for Niagara region students.

FIGURE 6 - Distribution of drug use scores for central Ontario students.

Full size image: 16 kB, FIGURE 6 - Distribution of drug use scores for central Ontario students.

FIGURE 7 - Distribution of drug use scores for eastern Ontario students.

Full size image: 10 kB, FIGURE 7 - Distribution of drug use scores for eastern Ontario students.

FIGURE 8 - Distribution of drug use scores for Ottawa Valley students.

Full size image: 10 kB, FIGURE 8 - Distribution of drug use scores for Ottawa Valley students.

From the table (p. 40) it can be seen that a score of 9 or more could be chosen as a cut-off point indicating "heavy" or problem drug use. Only 16.4 per cent of illicit users have a score of 9 or more. Scores less than 9 include 91 per cent of those with no drug problem score. Scores of 9 or more include 31.8 per cent of those with 1 drug problem, 70.6 per cent of those with two and 60 per cent of those with 3 or 4. Therefore a score of 9 or more identifies most of those with frequent drug problems and misclassifies relatively few of those with no problems. In order to achieve a score of 9 or more a student would have had to have used more than one illicit drug or several on a few occasions each. Only 4.48 per cent of the total sample (210 of 4,687) had a drug use score of 9 or more.

Correlations between illicit drug scores, drug use and drug problems scores

As predicted, there was a positive correlation ( r = .42, p < .001) between students' drug problem scores and their illicit drug use scores. There was a positive correlation (r = .67, p < .05) between average drug problem scores in the seven areas and average illicit drug use scores. Also, there was a very high correlation ( r = .91, p < .002) between mean illicit drug use scores and the proportions of students who reported a drug problem in the seven areas. That is, students who used illicit drugs (more often) were more likely to have drug problems. Also, in areas where there was more illicit use, there were more students with drug problems and with more serious drug problems.

However, there was no correlation between proportions of students with heavy drug use (i.e., 9+ scores) in a given area, and (i) mean drug use score ( r = .06, p < .45) or (ii) percentage of users with drug problems ( r = .02, p < .48). One reason for the lack of correlation may be that the proportions of heavy users in the various areas did not differ greatly (13.46 to 18.39) and correlations are more difficult to find when the numeric range of the variables is highly restricted. Also since there are only seven areas, a very close association would be necessary for significance.

TABLE

Relationship between illicit drug use score and drug problem scores for total sample

 

Drug problem score

Illicit drug use score

0

1

2

3

4

1
294 (N)
26      
2 143 26 2    
3 120 13 2 1  
4 117 16 4 1  
5 84 24 2 1  
6 76 20 1 1  
7 46 13 5   1
8 42 14 1 3  
9 16 8 4    
10 8 6 1    
11 10 5 4 3  
12 10 7 5 1  
13 8 3 4 1  
14 7 4 1    
15 4 4 1    
16 3 4      
17 4 3 1    
18 5 7 3    
19 1 2 2    
20 1 3   2  
21 1 4   1  
22 2   4 1  
23 3 4 2    
24 3 1 1 1  
25 1 1 1    
26          
27 1        
28          
29 1 1 1 1  
30 2 1      
31          
32   1   1  
33          
34     1    
35   1      
(...)
         
54   1      
X
4.20 8.04 12.66 14.11 7.00
N
1013 223 53 19 1

Discussion

The results of this study indicate substantial support for the hypotheses. In Ontario as a whole, and in each of the 7 areas the distributions of drug use were uni-modal, continuous and log normal. Few students reported many drug problems (6.32 per cent) or heavy drug use (4.48 per cent). There was a positive correlation between drug problem scores and illicit drug use scores across areas. There is also a high correlation between mean illicit drug use scores and the proportions of students who reported a problem.

The findings indicate that students who have more often used illicit drugs are more likely to have problems. In geographic areas where there was more illicit drug use, there were more students with problems and more students with a variety of problems. This suggests conclusions similar to those made for alcohol, i.e., where consumption is high the likelihood of problems from use is also high.

The distribution data show clearly, that there is no clear-cut dividing line between infrequent, moderate and heavy users, except that which is arbitrarily defined. The distributions in all areas of Ontario are continuous and relatively smooth, without any suggestion of bi-modality or separate groups of "normal" and "abnormal" users. The data, in total, suggest that efforts to reduce drug problems or frequent drug use without reducing drug use in general will likely be ineffective. In order to have fewer drug problems we will need to have less drug use in the total population. In general, findings about the distribution of drug use are similar to those found in earlier studies in Canada, Britain and the USA.

In this study, only limited dimensions of "drug problems" have been assessed and only few geographic areas have been studied. It would be well to repeat the study using additional criteria for "drug problems" and having a larger number of geographic areas. However, data are accumulating from various countries which strongly suggest that the distribution of drug use is continuous and uni-modal and that the total volume of drug use in an area governs the extent of drug problems.

Bibliography

Bruun, K., G. Edwards, M. Lurnio, K. Makela, Pam, Lyn, R.E. Popham, R. Room, W. Schmidt, O. Skog, P. Sulkunen and E. Osterberg. Alcohol Control Policies in Public Health Perspective, Finnish Foundation for Alcohol Studies, Helsinki, 1975.

Croxton, F.R., and D. J. Cowden. Applied General Statistics, Englewood Cliffs, Prentice Hall, 1955.

McDermott, Diane, and J. Scheurich. The logarithmic normal distribution in relation to the epidemiology of drug abuse, Bulletin on Narcotics, XXIX:1, 13-19, 1977.

Schmidt, W., and J.E. DeLint. Estimating the prevalence of alcoholism from alcohol consumption and mortality data, Quarterly Journal of Studies on Alcohol, 31, 957-964, 1970.

Smart, R.G., and M.S. Goodstadt. Alcohol and drug use among Ontario students in 1977: preliminary findings, Addiction Research Foundation, Toronto, 1977.

- - - - and P.C. Whitehead. The consumption patterns of illicit drug use and their implications for prevention of abuse, Bulletin on Narcotics, XXIV:1, 39-47, 1972.

- - - - and - - - - . The prevention of drug abuse by lowering per capita consumption: distribution of consumption in samples of Canadian adults and British university students, Bulletin on Narcotics, XXV:4, 49-55, 1973.

- - - - , - - - - and L. Laforest. The prevention of drug abuse by young people: an argument based on the distribution of drug use, Bulletin on Narcotics, XXIII:2, 11-15, 1971.