The "Follow up experiments" section of the Wikipedia article somewhat contradicts you:
> Studies that followed up on the contribution of environmental enrichment to addiction produced mixed results. A replication study found that both caged and "park" rats showed a decreased preference for morphine compared to Alexander's original study; the author suggested a genetic reason for the difference Alexander initially observed.[9] Another study found that while social isolation can influence levels of heroin self-administration, isolation is not a necessary condition for heroin or cocaine injections to be reinforcing.[10]
> Other studies have reinforced the effect of environmental enrichment on self-administration, such as one that showed it reduced re-instatement of cocaine seeking behavior in mice through cues (though not if that re-instatement was induced by cocaine itself)[11] and another that showed it can eliminate previously established addiction-related behaviors.[12] Furthermore, removing mice from enriched environments has been shown to increase vulnerability to cocaine addiction[13] and exposure to complex environments during early stages of life produced dramatic changes in the reward system of the brain that resulted in reduced effects of cocaine.[14]
> Broadly speaking, there is mounting evidence that the impoverished small cage environments that are standard for the housing of laboratory animals have undue influence on lab animal behavior and biology.[15] These conditions can jeopardize both a basic premise of biomedical research—that healthy control animals are healthy—and the relevance of these kinds of animal studies to human conditions.[16]
