Early clinical manifestations of cannabis dependence in a community sample

Abstract

Background: In this research, the ‘natural history’ of cannabis dependence is probed, using data from a large epidemiological sample of cannabis users followed from 1981 through 1996, until most of these users had passed through the empirically derived period of risk for developing cannabis dependence. Methods: The Baltimore Epidemiologic Catchment Area research group sampled, recruited, and assessed 3481 adults age 18+ years in 1981. Survivors were re-assessed roughly 13 years later. Among 599 cannabis users, a total of 37 had become cases of cannabis dependence during the follow-up interval; 41 had developed DSM-IIIR cannabis abuse without dependence; 521 had suffered insufficient problems to qualify as cases. Survival analysis methods were used to plot and study the cumulative occurrence of these problems in a primary contrast of the 37 dependence cases and 562 non-cases. Results: Subjectively felt loss of control over cannabis and continued cannabis use despite knowledge of harm seem to appear most rapidly among cases of cannabis dependence. In contrast, subjectively felt withdrawal symptoms tended to emerge later and for a much smaller proportion of dependence cases and non-cases. Conclusion: The 5000+ person-years of follow-up experience provided by the cannabis users in this study give an unprecedented look at the natural history and clinical course of cannabis dependence. Distinctive features of early cannabis dependence may help differentiate cannabis users who progress to clinically significant dependence from those who remain non-dependent.

Introduction

Based upon evidence from recent surveys in the US an estimated 5800 people start using cannabis every day (Department of Health and Human Services, 1999), and among cannabis users an estimated one in 11 (9%) become cannabis dependent (Anthony et al., 1994). In their descriptions of their first cannabis intoxications, users generally stress feelings of euphoria or well-being; feelings apparently reinforced with repeated use of the drug (Joy et al., 1999). Neuroadaptive changes to cannabis exposure might start after just a few doses (Bass and Martin, 2000), perhaps resulting in use of larger doses of the drug to achieve the same ‘high’. For reasons such as these, it might be surmised that subjectively-felt tolerance to cannabis effects emerges early in the course of cannabis dependence, and might help differentiate the experience of cannabis dependence cases from the experience of cannabis users who do not develop dependence upon this drug. Alternately, differences in early subjectively experienced effects of cannabis might help to differentiate users with biological predisposition to make more rapid progression to cannabis dependence, perhaps in synchrony with biological variation in enzyme activity (Pianezza et al., 1998).

For reasons described by Schuckit and colleagues in their research reports on the clinical course of alcohol dependence (Schuckit et al., 1993, Schuckit et al., 1995, Schuckit et al., 1998), it is important to discover the time-sequencing of clinical features that emerge during the insidious onset of chronic diseases of non-infectious origin. Studies of this type started with the work of Sydenham and other 18th century clinicians who sat by the bedside of affected patients and made careful clinical observations about which signs and symptoms tended to emerge first, and which appeared later, in the absence of specific clinical intervention. In the 20th century, similar methods have been applied to the ‘natural history’ and ‘clinical course’ of many different diseases (e.g. see Buckwalter et al., 1999, Boos et al., 2000, Hauser et al., 2000), but for the most part the psychoactive drug use disorders have been neglected in this research.

The work of Schuckit and colleagues has addressed this gap in evidence with respect to alcohol dependence. Martin et al. (1996) and Langenbucher and Chung (1995) have added to the picture with respect to alcohol problems during adolescence and adulthood.

In this research, the epidemiological spotlight is turned on cannabis dependence, building upon a line of research initiated by Kandel and colleagues (Kandel et al., 1992, Chen and Kandel, 1995, Kandel and Chen, 2000). Both traditional analyses of age of onset data (e.g., numerical summaries of the type presented by Schuckit and colleagues), as well as boxplots and survival analyses that shed light on the time course of clinical features associated with cannabis dependence are used.

This study represents a new step in our research group's continuing effort to understand the natural history of drug involvement and development of drug use disorders. Prior research on the precursors of alcohol and cannabis dependence has focused primarily upon clinical samples. The work uses a community sample of cannabis users assessed in 1981 and followed up through 1996, and allows for the identification of clinical features associated with cannabis use prior to the development of disorder. In addition, this study gains strength because it has followed cannabis users through most of the empirically derived ‘period of risk’ for developing cannabis dependence in the contemporary United States. As shown recently by Wagner and Anthony (submitted), the vast majority of recent cannabis dependence cases have emerged within 10 years of initial cannabis use. In this epidemiological sample of cannabis users, almost all had started smoking marijuana for more than 10 years (97%). Hence, for most of the cannabis users under study, cannabis dependence would already have developed by the time of follow-up in the early 1990s.