It can be argued that there are no 'enhancement' drugs per se, for enhancement is not an objective property of any drug but rather a meaning assigned to the drug on the basis of the expected effects of that drug. Thus all drugs could be enhancement drugs if deemed to be, for the definition of enhancement depends on the meaning we, as a society, give to the drugs' effects. If a drugs' effects are deemed to be detrimental that drug may be considered a 'poison' or 'toxin'. If the effects of a drug are deemed valuable or beneficial, and the drug is used to restore or sustain 'health' or 'normality', the drug may be deemed 'therapeutic' or a 'medicine'. If the effects of the drugs are experienced as beneficial but those benefits are not seen as necessary to restore or sustain health, these drugs may be deemed 'recreational' (if the drug is used primarily for 'pleasure') or 'enhancing'. Whether or not drug use is considered enhancement depends in large part on our definition of 'normality', and on what we, as a society, value. The term 'enhancement' refers to changes that are considered to be in a positive direction, that is, they are 'improvements' or changes that are viewed as increasing value, and that alter an individual to a state which society has deemed above 'normal'. Enhancement is commonly defined by bioethicists as "interventions that are used to improve human form or functioning beyond what is necessary to restore or sustain health" (Juengst & Moseley, 2019). However, whilst there are established definitions of ‘health’, such as the WHO "Health is a state of complete physical, mental and social well-being and not merely the absence of disease or infirmity", these are typically broad and so open to interpretation (e.g. is it merely repair and maintenance or does it include enhancing well-being?) which makes enhancement challenging to define.

The nonmedical use of anabolic–androgenic steroids (AAS) by athletes and the general population is a public health concern. One particular concern relates to the link between AAS use and violence. Although there is a growing body of research on the association between AAS and violence, results as to what the relationship is seem to be mixed. The aim of this meta-analysis is to improve our understanding of what the current evidence base indicates regarding the relationship between the use of AAS and interpersonal violence. In total, 14 studies with a total of 16 samples met the inclusion criteria, comprising a total sample size of 137,055 participants. The findings show that there is a significant association between AAS use and interpersonal violence (r= .21 [95% confidence intervals (CI): 0.15, 0.27], p< .00001), although small. It is however unclear what the direction of this relationship is as almost all studies failed to demonstrate whether AAS use occurred prior or subsequent to the violent behavior. Risk of bias for most studies was moderate or high. Three potential moderators, AAS use via self-report or testing, interpersonal violence via self-report or criminal records of violent behavior, and proportion of males, were nonsignificant. Although a meta-analytic relationship between AAS use and interpersonal violence was found, it is unclear to what extent AAS causes violent acts. We need to carefully consider the role of mediating factors, and that the relationship between the two could be part of a broader context as opposed to AAS being the cause of violence.