In my prior post, I overviewed a pair of papers which suggested the possibility that rats provided with running wheels might be used to model exercise addiction. The application hinged on a finding that when rats are provided with longer term 4 or 24 hr access to a wheel they gradually escalate their running over the course of about three weeks; this effect is greater than any increases seen when rats have wheel access for only an hour or two. As I alluded to, however, confidence in wheel running as a model of human exercise addiction akin to substance dependence is going to require a lot more converging evidence.
Kanarek and colleagues have provided some of this converging evidence. The authors examined the effects of challenge with the opiate antagonist naloxone in groups of male and female rats which had been permitted to run on an activity wheel.
This study relies on an effect* which has been known for quite some time, namely that the acute administration of low doses of drugs which block mu opiate receptors can rapidly precipitate withdrawal signs in rats or mice which have been treated chronically with morphine, heroin or other mu opiate agonists. Withdrawal signs that are similar in appearance to those that emerge with spontaneous withdrawal of the animal from chronic exposure to opiates. As Marshall and Weinstock observed in 1969, withdrawal symptoms could be quantified as an index of opiate dependence.
The purpose was to determine whether the number of jumps elicited by nalorphine in groups of mice could be used as a method of measuring the intensity of the withdrawal syndrome...The number of jumps was a monotonic increasing function of both the number of injections and the total dose [of morphine-DM] injected...In conclusion it is suggested that the number of jumps elicited by an antagonist in chronically narcotized mice can be used as a quantitative measure of the withdrawal syndrome.
Kanarek and colleagues were thus not just hypothesizing that they could precipitate withdrawal differentially in exercised animals, but also that the neuropharmacological change associated with exercise involved endogenous opioids. To wit, the endorphins which have been speculated in common use to underlie the so-called runner's high.
The study is a bit complicated because it includes a manipulation termed Activity-Based Anorexia. Apparently if you give rats access to food for only an hour a day, they can survive with approximately normal maintenance of weight but if you also provide them with an activity wheel, they stop eating and drop weight-even to the point of death. This is a mere distraction for the present purpose, however, since the effect of challenge with the opiate antagonist was not qualitatively changed by the feeding condition. Nevertheless, the designs were between groups with factors of wheel access and feeding condition (24 hr food vs 1 hr food). There was also an additional yoked-pair feeding group which was inactive but fed the same amount of food consumed by the 1-hr / wheel access animals. This is by way of explaining the graphs, but the key effect for today's discussion lies in the main effect of exercise condition.
The first experiment was conducted in female rats permitted wheel access for 7 days (plus sedentary groups) and then initiated on the 1 hr / 24 hr / yoked feeding conditions. The naloxone challenge (1 mg/kg) was initiated after 3-6 days when the 1 hr / activity group had dropped to 80% of their initial bodyweight. Since individuals took different numbers of days to reach this criterion, matched numbers of animals from the other groups were challenged with naloxone on the days over which the critical group reached criterion. Traditional withdrawal symptoms were scored.
As you can see in the Figure, withdrawal was precipitated more robustly in the group which had been permitted to exercise on the wheel and received 1 hr access to food, relative to the remaining groups. The authors also reported a correlation between total withdrawal signs exhibited by an individual and the wheel activity on the day before naloxone challenge in all activity rats but this was attributable to the food restricted subgroup. Similar results were found when they assessed the number of rats expressing a given withdrawal symptom, instead of the overall withdrawal score, as shown in the Table.
The second experiment was conducted in males with similar wheel access and feeding groups. In this case, however, the males in the exercise groups were permitted 25 days of wheel access (instead of the 7 used for the females) prior to initiation of the feeding conditions. Again, naloxone challenges were conducted when the 1-hr feeding / Wheel access group dropped to 80% of their prior weight.
Effects of naloxone challenge were most pronounced in exercised rats however in this case the effect did not depend on feeding condition. The graph of the mean total withdrawal scores shows that naloxone precipitated more signs of withdrawal in the 24-hr feeding / Wheel access group than in the sedentary groups.
So why the difference in the 24-hr feeding / wheel access condition between the experiments? I think the most likely issue is the difference in wheel access duration prior to the food conditions. The males, although they ran less than the females, escalated their running through about 16 days of access and had plateaued by the start of the food-access manipulation at day 25. The females were still increasing their running at the end of 7 days and into the food manipulation condition, but very likely had not completely expressed the commonly observed increase in daily running associated with 24 hr access over ~3-4 weeks duration.
In some senses these two experiments are not discordant but rather complement each other. Together they point out that the amount of activity on a wheel is not sufficient to increase liability for precipitated withdrawal. The females in the 24-hr feeding condition peaked at about 21,000 revolutions per day whereas the males in the 1-hr feeding condition peaked at about 8,500 revolutions per day. Only the latter exhibited increased withdrawal signs after naloxone when compared with their inactive control group. This suggests that it is the relative increase from baseline activity levels that is most important, rather than the spontaneous difference in baseline running.
And that interpretation, DearReader, is consistent with the idea that repeatedly engaging in physical activity can disrupt the neuronal mechanisms that subserve the rewarding aspects of that exercise. This disruption can then be observed as withdrawal signs, given acute administration of an opiate antagonist. This further suggests that endogenous opioids may be critically involved in the rewarding, and therefore addictive, aspects of repetitive exercise.
As with the behavioral escalation papers I previously discussed, this is not in and of itself proof that rats are addicted to, or become dependent on, wheel running.
A PubMed search for naloxone precipitated withdrawal finds 1135 references.
Kanarek RB, D'Anci KE, Jurdak N, & Mathes WF (2009). Running and addiction: precipitated withdrawal in a rat model of activity-based anorexia. Behavioral neuroscience, 123 (4), 905-12 PMID: 19634951
Marshall I, & Weinstock M (1969). A quantitative method for the assessment of physical dependence on narcotic analgesics in mice. British journal of pharmacology, 37 (2) PMID: 5388579