Changes over time in heroin and cocaine use among injecting drug users in Amsterdam, The Netherlands, 1985-1989
British Journal of Addiction (1991) 86, 1091-1097 RESEARCH REPORT
CHRISTINA HARTGERS,1 ANNEKE (J.A.R.) VAN DEN HOEK,1 PIETA KRIJNEN,1 GIEL H. A. VAN BRUSSEL 2 &
ROEL A. COUTINHO 1
Correspondence to: Christina Hartgers, Municipal Health Ser vice, Department of Public Health and Environment, P.O. Box 20244, 1000 HE Amsterdam, The Netherlands.
1 Municipal Health Service, Department of Public Health and Environment, & 2 Municipal Health Service, Drugs Department, Amsterdam, The Netherlands
Some injecting-related HIV risk reduction has occurred among injecting drug users (hereafter called IDUs) in Amsterdam.1-3
In the USA, next to HIV risk reduction among drug users, 4-5 there seems to be an increase in HIV risk due to an increase in cocaine use.6 Cocaine injecting has been shown to be related to risky injecting behaviour7 and to HIV-seropositivity.8 Smoking or inhaling cocaine ('crack') appears to be related to risky sexual behaviour 9,10 and to HIV infection.' 11,12 In Europe, an increase in the use of smokable forms of cocaine has been reported, 13 but data on the relation with HIV infection are absent so far.
It seems therefore relevant to monitor drug use trends, especially with regard to cocaine use. The authors surveyed oral drug use and injecting variables in IDUs in Amsterdam enroled in an HIV study through `low threshold' methadone clinics from December 1985 until March 1989. The objective of the present study is to assess whether changes over time have occurred in ways of drug use and in the kind of drugs used.
Drug users and low threshold methadone programs in Amsterdam
It is estimated that on average about 5500 hard drug users (DUs) were present in Amsterdam in each quarter during the years 1987 and 1988. This estimate does not include short term visits by drug tourists. Among these 5500 in both years approximately 2400 DUs, i.e. 40%, had injected in the previous month. 14
The Drugs Department of the Municipal Health Service in Amsterdam operates seven 'low threshold' methadone clinics: four neighbourhood clinics, two clinics for extremely problematic DUs and one clinic for DUs who work as prostitutes or who are non-residents." These clinics do not aim at treatment of the addiction; they aim primarily at contacting heroin users and 'regulating' (stabilizing) their use.15 Illicit opiate or cocaine use is not a reason for dismissal. The clinics daily provide onsite primary medical and social care and supply methadone to a yearly population of approximately 3500 DUs (both IDUs and non-IDUs). In 1987, the average prescribed methadone dose was 35 milligrams. 1 6
Study population and variables
In 1985 an epidemiologic study of HIV-infection among drug users was initiated. This ongoing study, aspects of which have been previously de scribed',","," involves voluntary, confidential HIV antibody testing and counselling for drug users in Amsterdam, combined with an interview by speci ally trained nurses using a standard demographic and behavioural questionnaire. Most DUs (83%) entering the study have never injected (IDUs), while 55% have injected in the previous month. DUs, who can participate once or more often, enrol in the HIV study mainly through one of the above mentioned seven methadone clinics or through the separate STD-clinic for addicted prostitutes.
The present study concerns the intake-visits in the period December 1985 to March 1989 of all 386 IDUs who enroled for the HIV-study through one of the previously mentioned seven methadone clinics. The research team was distinct from methadone clinic workers.
Limited data are available on DUs who get methadone on prescription in Amsterdam,1 4, 1 6 but not separately for injecting DUs. With regard to age, there was no significant difference between the 386 IDUs in the sample and all DUs who got methadone prescribed in Amsterdam in 1987. With regard to the sex-ratio there was a difference: there were relatively fewer males in the study than among DUs attending low threshold methadone clinics in 1987 (respectively male/female ratio 2:1 versus 3:1).
All variables, except HIV-serostatus, are selfreported. All 386 participants are IDUs, which means they have injected drugs at some previous time. Current behaviour is behaviour in the 6 months preceding intake. Long term methadone (LTM) use was defined as daily methadone use in the last 5 years. Heroin smoking in the present study comprises both smoking heroin and tobacco in a cigarette as well as 'chasing the dragon', a method in which the heroin vapours are inhaled. Cocaine freebasing is the smoking or inhaling of cocaine alkaloid. Use of benzodiazepines and/or sedatives was determined by asking about the current use of different benzodiazepines, barbiturates and metha qualones (indicated by brand name). To study changes over time the total study period was split up in four consecutive intake-periods of 10 months each: I=12/85-9/86, II=10/86-7/87, III=8/87-5/88, IV=6/88-3/89.
Statistical analysis and serology
Statistics used included chi-square (X2), Fisher's exact test, Mann-Whitney test (M-W-test) and Spearman rank correlation coefficient. P-values less than 0.05 were considered significant.
Overall increases or decreases in drug use over time were studied; for each drug use variable a logistic regression model was fitted with drug use as dependent variable. 20 Moment of intake (in days as from start of the study) and six potentially confounding variables (age, sex, German or South European nationality, time living in Amsterdam, post at which participant enroled in the study and long term methadone use) were forced into the models as independent variables. Adjusted odds ratios (ORs) and 95% confidence intervals (CIs') were calculated for the period length of 303 days, which is the length of the four consecutive intake periods.
It was decided not to include drug use behaviours which were practised by less than 45 participants during the total study period, since estimates from models based upon these data might be unstable. Serological tests used for HIV-testing were ELISA's. Confirmation of positive specimens was performed by immuno blotting and competitive ELISA's, as has been described before."
1These intervals were calculated as follows: e raised to the power (beta + 1.96 x standard error).
Table 1 presents the general characteristics of the total group and of the four consecutive intake groups.
Table 2 presents drug use in the 6 months preceding intake. Most prevalent with regard to oral and intranasal use after methadone are the smoking of heroin and the use of benzodiazepines. Three quarters of the IDUs currently inject, of whom approximately one half inject daily. Most current IDUs mainly inject heroin and cocaine, either simultaneously ('speedball') or apart.
Drug use in the four consecutive intake-groups
Table 2 shows that the injecting behaviour of the four intake-groups over time, adjusted for the six possible confounders, is similar. However, daily methadone use and use of benzodiazepines decrease over time, while cocaine freebasing and heroin smoking increase. Fig. 1 shows the non-adjusted proportions of cocaine freebasing, heroin smoking, current injecting and current daily injecting in the four intake groups.
When comparing changes over time in current daily injecting among different subgroups, an adjusted OR of 0.65 was found in the subgroup of 93 cocaine freebasers (CI=0.40-1.07), as compared to an adjusted OR of 0.99 among the 291 non-basers (CI=0.77-1.29) (see fig. 2 for non-adjusted data). In the subgroups of heroin smokers and nonheroin smokers the adjusted ORs for current daily injecting were 0.95 (CI=0.67-1.34) and 0.91 (CI=0.67-1.24) respectively
Since the four intake-groups differed with regard to current daily methadone use we calculated ORs which were also adjusted for this variable. No noteworthy changes occurred among the earlier found adjusted ORs or CIs. The possibility exists that IDUs with high risk behaviour entered the HIV study preferentially at the start. However, as can be seen in table 1, the HIV-seroprevalence in each of the four intake groups is similar. Current borrowing is the same in the four intake-groups (adjusted OR=0.97, CI=0.78-1.20) and there was no relation between intake group and frequency of ever borrowing: in intake group I to IV respectively 37%, 364'x, 37% and 41% had borrowed 10 times or more (X2 for trend=0.17, df=l,p=0.68).
The HIV-prevalence among the 93 cocaine freebasers is 19% compared to 32% among the 291 non basers (X2=4.6, df=1, p=0.03). Cocaine freebasing in the present study was found to be related to three of nine previously determined HIV-risk factors: 1 7 ,18 cocaine freebasers have more recently started injecting (X==4.7, df=1, p=0.03), they have borrowed less often (M-W-test, p=0.005) and most of them (90%) smoke heroin (as compared to 36% among non-basers, X 2=84.0, df=1, p<0.0001), all of which has been found to be related to a smaller risk of HIV-infection.
Among the 93 cocaine freebasers there are 19 (20%) current prostitutes as compared to 41 (14%) among the 291 non-basers (X2=2.1, df=1, p=0.14). Among the subgroup of 60 current prostitutes 42% of freebasers report using condoms always in vaginal contacts with clients as compared to 20% of non-basers (/„ 2=3.4, df=1, p=0.07). Current prostitutes who freebase have as many clients per month as non-basers (means 55 and 58, medians 24 and 50 respectively, M-W-test, p=0.49).
In a group of 31 DUs, who enroled in the study from September 1988 to March 1989, extra questions concerning sexual behaviour were asked. Thirteen of these 31 currently had vaginal contact with a casual partner. There was no difference between freebasers and non-basers in this regard (X 2=0.9, df=1, p=0.35). The mean number of different casual partners in the previous six months is as high in the 16 cocaine freebasers as in the 15 non-basers: 1.2 versus 0.6, median 0.5 and 0 respectively (M-W-test, p=0.27). There is also no significant difference between cocaine freebasers and non-basers with regard to always using condoms in vaginal contact with casual partners (13% and 20% respectively, Fisher's exact test p=1.0).
The most striking findings in our sample were firstly the increase over time in cocaine freebasing and in heroin smoking, and, secondly, that there were no indications of significant changes in injecting over the years 1985 to 1989. The decrease over time in daily methadone use was not related to the changes found in other drugs. Thus, the findings pertain both to daily methadone users as to others. We do not have a clear explanation for the decrease in daily methadone use. The decrease in benzodiazepine use might be a reflection of restrictions in prescribing tranquillizers to drug users, which started in 1986. 21
Potential limitations of the present study include, first of all, that the representativeness of the sample for IDUs in Amsterdam or IDUs in low threshold methadone programs in Amsterdam cannot be known; no data on both populations are available. A second potential limitation is that our sample is self-selected; it consists of IDUs who voluntarily entered an HIV study. A third potential limitation is the lower number of participants in the later intake groups; only 37% of the sample is in the third or fourth intake group. Therefore, if IDUs with an high HIV-risk entered the study preferentially at the beginning, this would be a serious potential confounder. However, there are no indications for this: no changes over time were found for HIV sero status, current injecting, current borrowing and ever borrowing. These findings are in agreement with earlier studies,' -; which found that neither needle sharing nor HIV sero status at intake are related to time of intake in the HIV study.
A higher proportion of current cocaine freebasers and current heroin smokers is found in the later intake-groups. In the total group, there are no indications of changes over time in current injecting, in current daily injecting or in the kind of drugs injected. There is an indication for a decrease in current daily injecting among cocaine freebasers: in 1985/1986 half inject daily, while in 1988/1989 among a relatively larger groups of cocaine freebasers only one in eight inject daily.
These findings suggest that in spite of an increase in number of needles and syringes exchanged in the years 1985 to 1989,14 no parallel increase in current injecting or current daily injecting among IDUs has occurred, which is in agreement with earlier studies.1'2
It was found that almost all cocaine freebasers in the present study also smoke heroin. In Amsterdam, traditionally, the combination of cocaine freebasing and heroin smoking is practised by Surinam/Antillian drug users. This ethnic group, which injects rarely, is hardly represented in the present study (only 3%). Our findings seem to indicate that non ethnic IDUs are adopting the combination of heroin smoking and cocaine freebasing. Heroin smoking has been found to be independently associated with HIV seronegativity.12 The increase in heroin smoking seems therefore favourable from an HIV prevention point of view.
Freebasing cocaine became popular in Amsterdam in 1981?r In the USA crack rocks (the dried end product of a heated mixture of cocaine hydrochloride, sodium bicarbonate (baking soda) and water are sold ready for use by dealers, while an Amsterdam DU who freebases cocaine buys cocaine hydrochloride and prepares the 'freebase' him or herself, either in the above described manner or by adding aqueous ammonia, The increase in cocaine freebasing among IDUs at intake over time does not seem to be related to a decrease in price. A spokesman of the 'Junkiebond' (the union of DUs in Amsterdam) mentioned prices of cocaine hydrochloride between 140-200 Dutch guilders per gram in 1989, while in 1985 prices between 115-170 Dutch guilders per gram were reported.23 Up till the beginning of 1990, the Municipal Police had no indications that ready for use crack was being sold in Amsterdam (personal communication R, Jellema, Amsterdam Municipal Police).
In the present sample there were no indications that sexual risk behaviour was related to cocaine freebasing. Also, the IDUs in the sample who freebase cocaine have a relatively low HIV seroprevalence. Of importance is furthermore that they are primary opiate users.
Accordingly, the present findings suggest that the relationship between crack use, risky sexual behaviour and high HIV-seroprevalence reported in the USA6,9,10,11,12 is not only related to the pharmacological effects of crack but also to the specific cultural and social setting and to the primary use of cocaine, In conclusion, our study shows that cocaine
freebasing and heroin smoking in recent years have become more popular among IDUs in our sample and that injecting does not increase. Whether these changes occur due to the AIDS-threat is difficult to assess. From an HIV-prevention point of view, these changes could have a favourable effect on the spread of HIV among this high risk group, However, the present picture may altogether change if ready for use crack did become available and reach new or existing primary cocaine users,
This study was supported by the Netherlands Foundation for Preventive Medicine (grant no. 28-1258), The authors thank the nurses B, Frolich, B. Scheeringa-Troost, G, Stienstra and J. Teeuwissen for interviewing and collecting blood samples; Dr J. Goudsmit and M. Bakker for performing the laboratory tests; R, Jellema and the 'Junkiebond' for information they provided; H, J, A, van Haastrecht for management of data and critical comments and M, ter Pelle and T, Maruanaya for preparing the manuscript.
1. HOEK, J. A. R. VAN DEN, HAASTRECHT, H. J. A. VAN & COUTINHO, R. A. (1989) Risk reduction among intravenous drug users in Amsterdam under the influence of AIDS, American Journal of Public Health, 79, pp. 1355-1357.
2. HARTGERS, C,, BUNING, E, C., SANTEN, G. W. VAN, VERSTER, A, D. & COUTINHO, R. A. (1989) The impact of the needle and syringe-exchange pro gramme in Amsterdam on injecting risk behaviour, AIDS, 3, pp. 571-576,
3. HAASTRECHT, H. J. A. VAN, HOEK, J. A. R. VAN DEN, BARDOUX, C., LEENTVAAR-KUYPERS, A. & COUTINHo, R. A. (1991) The course of the HIV epidemic among intravenous drug users in Amster dam, the Netherlands, American Journal of Public Health, 81, pp. 59-62.
4. DES JARLAIS, D. C., FRIEDMAN, S. R. & HOPKINS, W. (1985) Risk reduction for the acquired immuno defrciency syndrome among intravenous drug users, Annals of Internal Medicine, 103, pp. 755-759.
5. BECKER, M. H. & JOSEPH, J. G. (1988) AIDS and behavioral change to reduce risk, A review, American Journal of Public Health, 78, pp. 394-410.
6. WEISS, S. H. (1989) Links between cocaine and retroviral infection, Journal of the American Medical Association, 261, pp. 607-608.
7. AMSEL, Z., BATTJES R, & PICKENS, R, (1990) Cocaine use and HIV-risk among intravenous opiate addicts. Paper presented at the VIth International Conference on AIDS, San Francisco.
8. CHAISSON,. R. E,, BACCHETTI, P., OSMOND, D., BRODIE, B,, SANDE, M. & Moss, A. R, (1989) Cocaine use and HIV infection in intravenous drug users in San Francisco, Journal of the American Medical Association, 261, pp. 561-565.
9, GOLDSMITH, M, F. (1989) Sex tied to drugs=STD spread, journal of the American Medical Association, 260, pp. 2009.
10, ROLFS, R, T., GOLDBERGH, M. & SHARRAR, R. G. (1990) Risk factors for syphilis: cocaine use and prostitution, American Journal of Public Health, 80, pp.853-857.
11. STERK, C. (1988) Cocaine and HIV seropositivity, The Lancet, pp. 1052-1053.
12. SCHOENBAUM, R. R., HARTEL, D, & FRIEDLAND, G. H, (1990) Crack use predicts incident HIV seroconver sion. Paper presented at the VIth International Conference on AIDS, San Francisco.
13. STRANG, J., GRIFFITHS, P. & GOSSOP, M. (1990) Crack and cocaine use in South London drug addicts: 1987-1989, British Journal of Addiction, 85, pp. 193-196.
14. BUNING, E. C. (1987, 1988, 1990) de GG&GD en her drugprobleem in cijfers, 11, III, IV (Amsterdam, Municipal Health Service).
15. BUNING, E. C. (1989) AIDS related interventions among drug users in the Netherlands. Paper presented at the European Symposium on AIDS and Drug Abuse, Stockholm, 25-28 September.
16. PLOMP, H. N. & REIJNEVELD, S. A. (1991) Methadon verstrekking aan druggebruikers in Amsterdam, ana lyse van gegevens van de Centrale Methadon Registra tie over de jaren 1984 tot 1987 (Amsterdam, VU uitgeverij).
17. HOEK, J. A. R. VAN DEN, COUTINHO, R. A., HAAS TRECHT, H. J. A. VAN, ZADELHOFF, A. W. & GOUDSMIT, J. (1988) Prevalence and risk factors of HIV infections among drug users and drug using prostitutes in Amsterdam, AIDS, 2, pp. 55-60.
18. HOEK, J. A. R. VAN DEN, HAASTRECHT, H. J. A. VAN, ZADELHOFF, A. W. VAN, GOUDSMIT, J. & COUTINHO, R. A. (1988) HIV infectie onder druggebruikers in Amsterdam; prevalentie en risicofactoren, Nederlands Tijdschrift voor Geneeskunde, 132, pp. 723-728.
19. HOEK, J. A. R. VAN DEN, HAASTRECHT, H. J. A. VAN, SCHEERINGA-TROOST, B., GOUDSMIT, J. & COU TINHO, R. A. (1989) HIV infection and STD in drug addicted prostitutes in Amsterdam: potential for heterosexual HIV transmission, Genitourinary Medi cine, 65, pp. 146-150,
20. HOSMER, D. W. & LEMESHOW, S. (1989) Interpreta tion of the coefficients of the logistic regression model, in: HOSMER, D. W. & LEMESHOW, S. Applied Logistic Regression, pp. 56-57 (New York, John Wiley & Sons).
21, LIONARONS, W. & WEIJERS-EVERHARD, J. P. (1986) Her voorschrijven van kalmerings- en slaapmiddelen aan druggebruikers, Geneesmiddelen Bulletin, 20, pp, 57-62.
22. BoETJE, J. (1984) "Free-basen": een nieuwe methode van cocalnegebruik, Nederlands Tijdschrift voor Gen eeskunde, 128, pp. 2237-2238.
23, LEWIs, R. (1989) European markets in cocaine, Contemporary. Crises, 13, pp. 35-52