Evolution of the Theory
The 1950s was a time of great advances in the development of behaviorally active drugs, and a great deal of effort was spent trying to understand why drugs altered behavior in specific ways. During this period, drug action on behavior was interpreted, for the most part, in terms of the drug's capacity to reduce a primary drive (e.g., hunger, fear, or thirst) or by the drug's capacity to interfere with a learned association between a stimulus in the environment and one of the primary drives. The work of Clark L. Hull and Neal Miller figured prominently during this period (for review see McMillan and Katz 2002). Thus, when drugs with clinically relevant effects on anxiety were examined in laboratory animals and shown to produce differential effects on behavior controlled by food presentation as compared to behavior controlled by electric shock, the observations were viewed as being consistent with the reduction of primary drives and confirmed the validity of the drive reductionist concepts of drug action. It was in this historical context that Peter B. Dews initiated a series of studies that would eventually lead to the development of what is now known as rate-dependency theory.
In 1955, Dews published the first of a series of papers in which he examined the effect of drugs in food-deprived pigeons responding under different operant schedules of food presentation (Dews 1955). In his first paper, he showed that doses of ► pentobarbital that produced dramatic decreases in the rate of responding under a fixed-interval 900 s (FI900) schedule did not affect the rate of responding under a fixed-ratio 30 (FR30) schedule of food presentation. Since responding under the two schedules was maintained by the same event, it was difficult to explain such profound differences based upon drive reduction theories proposed by the current theorists of the day. Alternatively, Dews suggested that rather than interacting with drive states, pentobarbital appeared to be producing the differential effect by interacting with the control of the rate of responding. In a subsequent paper, Dews showed that the effects of ► methamphetamine in pigeons responding under four different schedules of food presentation depended upon the length of the time between individual responses (IRT; rate of responding). When control performance was characterized by low rates of responding, low to moderate doses of metham-phetamine increased rates of responding. At higher doses, methamphetamine did the same thing but it also decreased the high rates of responding (Dews 1958). Over the next several years Dews along with colleagues, William Morse and Roger Kelleher, as well as investigators from other labs showed that the control rate of responding was a key factor in determining drug effects, and that the control rate of the behavior was an important independent variable. However, it was not until 1964 that Dews formally proposed a way to express this relationship
(Dews 1964). Dews argued that since one frequently had to deal with a range of doses that spanned several log units, by analogy to dose-response curves in classical pharmacology, the control rate of responding should be plotted on the abscissa using a log scale. He went on to show that if one plotted the log of the rate of responding following drug administration divided by the rate of responding under control conditions on the ordinate (Fig. 1), not only did this log-log plot aid in the uniform distribution of the data points, it yielded a linear relationship that had predictive value. As a result, one typically sees a log-log linear relationship with a negative slope. Thus, very low rates of responding tend to be increased more than moderate rates, and high rates of responding are either increased less than moderate rates or are actually decreased compared to the control rate of responding. This log-log linear relationship and the theory behind it are referred to as the rate-dependent theory of drug action.
During the 1960s and early 1970s, numerous studies showed that the rate of responding was more important in determining the effect of drugs than the specific reinforcing event used to maintain the behavior. Furthermore, in addition to the ► amphetamines and ► barbiturates, rate-dependent drug effects were shown for many drugs including ► benzodiazepines, ► meprobamate, ► phe-nothiazines, ► atypical antipsychotics, anticholinergics, ► phencyclidine, and ► ketamine.
Rate-Dependency Theory. Fig. 1. Typical rate-dependency plot. Abscissa: control rate of responding (log scale); Ordinate: Rate of responding following drug administration divided by the control rate and converted to a percentage (log scale).
Was this article helpful?