ARU home page | ARU Profile | Addiction Primer | Biological Basis | Research Findings
University Courses | Opportunities | Research Reports | Feedback

From M.A. Bozarth and R.A. Wise (1985). Toxicity associated with long-term intravenous heroin and cocaine self-administration in the rat. Journal of the American Medical Association, 254, 81-83.

Toxicity Associated With Long-term Intravenous Heroin and Cocaine Self-Administration 
in the Rat


Michael A. Bozarth & Roy A. Wise 

Center for Studies in Behavioral Neurobiology
Department of Psychology 
Concordia University 
Montréal, P.Q. H3G 1M8 CANADA


Laboratory rats were given unlimited access to intravenous cocaine hydrochloride or heroin hydrochloride. Animals self-administering cocaine quickly developed a pattern of episodic drug intake, with periods of excessive cocaine self-administration alternating with brief periods of abstinence. Subjects allowed continuous access to intravenous heroin showed stable drug self-administration, with a gradual increase in daily heroin intake over the first two weeks of testing. The general health of the animals became markedly different: those self-administering heroin maintained grooming behavior, pretesting body weight, and a good state of general health; rats self-administering cocaine tended to cease grooming behavior, loose up to 47% of their pretesting body weight, and to show a pronounced deterioration in general health. The mortality rate for 30 days of continuous testing was 36% for animals self-administering heroin and 90% for those self-administering cocaine. These results suggest that cocaine is a much more toxic compound than heroin when animals are given unlimited access to intravenous drug.


Opiate addiction has traditionally been associated with a moderate fatality rate. Cocaine use, on the other hand, is considered by many to be a relatively safe habit. Reported causes of death directly attributable to cocaine use seem to be less frequent. With the current rise in cocaine use, it is appropriate to estimate the mortality rate from addiction to this compound relative to that of heroin, which is widely recognized as a moderately dangerous drug addiction. The possibility exists that factors limiting availability and use are responsible for the apparent safety of cocaine and that, given the availability of large quantities of high-grade drug, cocaine toxicity has been underestimated.

Studies involving intravenous (IV) drug self-administration in laboratory animals have typically limited access to drug during the experimental session. One reason for this is that early tests of cocaine self-administration with unlimited access resulted in a high subject fatality rate.

While this observation is sufficient to prompt investigators to limit periods of drug access during experimental testing of stimulant drugs, there has not been a systematic comparison of the toxicity of cocaine and heroin during continuous access. The present study reports tests where animals were allowed continuous access to either IV cocaine hydrochloride or heroin hydrochloride for 30 days. The hourly drug intake was measured for each 24-hour period of testing, and observations were made regarding the general health of the animals.


Male Long-Evans rats were anesthetized with pentobarbital sodium, 60 mg/kg intraperitoneally, and long-term IV catheters (Silastic tubing: outside diameter, 1.2 mm) implanted in the right external jugular vein. Penicillin G procaine, 60,000 units administered intramuscularly, was administered prophylactically, and the subjects were allowed seven to ten days to recover from the surgical procedure.

The rats were randomly divided into two groups and placed in standard operant test chambers with food and water available ad libitum. One group of subjects (n = 12) had access to cocaine hydrochloride, 1 mg/kg per infusion, while the other group (n = 11) had access to heroin hydrochloride, 100 m g/kg per infusion. Both drugs were dissolved in a physiological saline containing 0.3% sodium metabisulfite and sterilized by filtration. Each IV catheter was connected to a 50-mL syringe by polyethylene tubing, and a fluid commutator located between the subject and the infusion pump permitted unrestricted movement of the animal during testing. Pressing a lever activated a motor-driven syringe pump that delivered 0.25 mL of drug solution over 10 s. Pressing the lever during the infusion interval had no scheduled effect, but rats could self-administer another infusion of drug immediately after completion of an injection. Testing for drug self-administration continued 24 hours a day for a maximum of 30 days. The hourly drug intake was measured for each rat.


Both groups of rats quickly learned to self-administer the drugs. Rates of learning for the two groups were comparable although the dose of cocaine tested produced a somewhat higher rate of lever pressing than the dose of heroin tested. All 11 rats tested for heroin self-administration reliably self-administered drug with approximately the same hourly intake of heroin as rats tested for two- and six-hour sessions (M.A.B., unpublished data). Ten (83%) of the 12 rats tested for cocaine self-administration showed reliable responding for cocaine. Two animals in the cocaine group did not learn to self-administer significant quantities of the drug (i.e., took < 10% of the normal hourly intake of subjects tested in two- and six-hour sessions; M.A.B., unpublished data), and their data were omitted from subsequent analysis.

Animals self-administering heroin showed a gradual increase in 24-hour drug intake during the first two weeks of testing (Fig 1). After this time, the mean daily intake of heroin remained constant through the remaining weeks of testing, with no evidence of further learning or tolerance to the rewarding effects of the drug injections. Rats in the heroin group tended to take approximately the same number of injections each day and to distribute their responding evenly throughout the 24 hours of testing. Body weights of these subjects remained stable during the 30 days of testing.

24-hr heroin self-administration

  Figure 1: Daily intake of drug for typical subject self-administering heroin hydrochloride, 100 mg/kg per infusion. Note gradual increase in drug intake across first two weeks of testing and stable level of self- administration through termination of testing. Data are expressed as mean number of infusions per hour of testing, and error bars represent SEM.  

The group tested for cocaine self-administration demonstrated erratic patterns of drug intake across days of testing. Many rats would take large amounts of drug during one 24-hour period and much less during the next period (Fig 2). The episodic pattern of responding developed early and continued throughout the 30 days of testing. Cocaine self-administration was associated with a reduction in body weight (mean loss, 29%), and several animals lost about 40% of their pretesting body weight. Grooming behavior diminished over days of testing, and there was an obvious deterioration in the animals general health. Several animals displayed full, clonicotonic seizures during testing but would reinitiate drug self-administration as soon as the convulsions subsided. The pattern of drug self-administration and the behavioral effects of chronic intake have been reported in detail elsewhere.

24-hr cocaine self-administration

Figure 2: Daily intake of drug for typical subject self-administering cocaine hydrochloride, 1 mg/kg per infusion. Note erratic pattern of responding across days. Data are expressed as mean number of infusions per hour of testing, and error bars represent SEM.  

One of the most striking differences in the effects of continuous drug access in animals self-administering cocaine or heroin was the mortality rate (Fig 3). By the end of the second week of testing, 60% of the subjects self-administering cocaine had died, while only 9% of those self-administering heroin had died (c 2[1] = 37.7, p < .005). At the completion of the 30-day test protocol, the mortality rate for the cocaine group was 90% and the mortality rate for the heroin group was 36% (c 2[1] = 23.1, p < .005). Animals in the heroin group surviving the full 30 days of testing were still in good health, while the surviving animal in the cocaine group showed marked deterioration (e.g., 32% weight loss).

 Comparison of cocaine & heroin toxicity


Figure 3: Cumulative number of fatalities from intravenous cocaine hydrochloride and heroin hydrochloride self- administration during unlimited access to drug. Percentage of animals lost are depicted as function of days of continuous testing. Subject attrition due to catheter failure is not shown. Solid circles indicate deaths in cocaine group; solid squares, deaths in heroin group.





Cocaine self-administration was accompanied by a substantially higher incidence of mortality than was heroin self-administration. This is reflected both by the number of days necessary to produce a 50% mortality rate and by the number of animals surviving the 30-day testing protocol. Rats terminally deprived of food have been reported to die after a loss of about 40% of their pretreatment body weight. In the present study, 56% of the subjects responding for cocaine died after a loss of less than 30% of their pretesting body weight. This suggests that the mortality associated with long-term cocaine use was not merely the result of a reduction of body weight; the high mortality rate is likely to be related to other factors, possibly the toxic effects of cocaine and its depression of medullary function.

The fact that the number of fatalities following unlimited access to cocaine was over twice that following unlimited access to heroin has obvious implications for human drug abuse. While many drug users recognize the inherent danger of opiate addiction, they fail to recognize the potential danger of long-term cocaine use. The facts that cocaine is usually not available to humans in unlimited quantities and that the available cocaine tends to be of low purity probably account for the relatively few deaths associated with its use. As the intake of cocaine increases and the concentration of this compound becomes greater (through the availability of higher-purity drug, by IV injections, or by "free-basing") the number of fatalities associated with cocaine abuse is likely to increase.


This research was supported in part by grants from the Medical Research Council (Canada) and by the National Institute on Drug Abuse. The technical assistance of Aileen Murray is gratefully acknowledged. Dr. Bozarth is a University Research Fellow sponsored by the Natural Sciences and Engineering Research Council of Canada.


1. Abelson PH: Death from heroin. Science 1970;168:1289.

2. Jaffe JH: Drug addiction and drug use, in Goodman LS, Gilman A (eds): The Pharmacological Basis of Therapeutics. New York, Macmillan Publishing Co Inc, 1975, pp 284-324.

3. Finkle BS, McCloskey KL: The forensic toxicology of cocaine, in Petersen RC, Stillamn RC (eds): Cocaine: 1977. Washington, DC, National Institute on Drug Abuse, 1977, pp 153-178.

4. Adams EH, Durell, J: Cocaine: A growing public health problem, in Grabowski J (ed): Cocaine: Pharmacology, Effects, and Treatment of Abuse. Washington, DC, National Institute on Drug Abuse, 1984, pp 9-14.

5. Deneau GW, Yanagita T, Seevers MH: Self-administration of psychoactive substances by the monkey: A measure of psychological dependence. Psychopharmacology 1969;16:30-48.

6. Johanson CE, Balster RL, Bonese K: Self-administration of psychomotor stimulant drugs: The effects of unlimited access. Pharmacol Biochem Behav 1976;4:45-51.

7. Pickens R, Thompson T: Characteristics of stimulant reinforcement, in Thompson T, Pickens R (eds): Stimulus Properties of Drugs. New York, Appleton-Century-Crofts, 1971, pp 177-192.

8. Campbell BA, Teghtsoonian R, Williams RA: Activity, weight loss, and survival time of food-deprived rats as a function of age. J Comp Physiol Psychol 1961;54:216-219.

9. Ritchie JM, Cohen PJ: Local anesthetics, in Goodman LS, Gilman A (eds): The Pharmacological Basis of Therapeutics. New York, Macmillan Publishing Co Inc, 1975, pp 379-403.

10. Evans Ma, Dwivedi C, Harbison RD: Enhancement of cocaine-induced lethality by phenobarbital, in Ellinwood EH, Kilbey MM (eds): Cocaine and Other Stimulants. New York, Plenum Press, 1977, pp 253-267.


©1985 American Medical Association

ARU home page | ARU Profile | Addiction Primer | Biological Basis | Research Findings
University Courses | Opportunities | Research Reports | Feedback