The most important finding of this study was that patients requiring rescue medication after arthroscopic knee surgery differed significantly in age, weight, BMI and smoking compared to those not requiring rescue medication. Multinomial logistic regression, showed a higher frequency of pain among young patients and smokers. These findings are in accordance with earlier studies [
3,
5]. The differences were due to physiological dissimilarities, or socioeconomic factors such as social status and education level in the different age groups and among smokers/non-smokers [
3‐
5]. To properly research these disparities, further studies are needed. In contrast with an earlier study, we found no significant difference between genders [
7]. An expected finding was a significant association between duration of surgery and degree of pain. There were furthermore no statistically significant differences between the RM and NRM groups regarding type of surgery (Table
1). It could, therefore, be speculated that surgical manipulation of tissue for a longer period of time would cause a higher level of pain due to increased tissue trauma. Interestingly, the increased pain was not reflected by subsequent detectable changes in the measured inflammatory and metabolic markers. Glucose consumption, interpreted as a sign of a hypermetabolic state [
9,
10], glycerol, a marker of cell membrane degradation, and therefore if increased a marker of cell death and tissue injury [
16,
17], and PGE
2, a mediator of central and peripheral pain sensitization [
18]. The results also demonstrate how subjective pain measured by VAS declined with time post-surgery. VAS exhibited an initial significant difference between RM and NRM groups as was predicted and the base for this study. PGE
2, known as a pain mediator and a marker of inflammation, showed a visual difference between the NRM and the NRM group (Fig.
4). However, this did not reach statistical significance. This result is not in accordance with earlier studies that showed an increase of PGE
2 in patients requiring opioids [
10]. The decline with time in PGE
2 was significant in both groups post-surgery. This suggests a primary elevation associated with the initial pain and/or surgical trauma. Glucose levels showed no statistical difference between RM and NRM groups which is in contrast with other studies that demonstrate an increased consumption of glucose in the synovial membrane after minor arthroscopic surgery, but no change after ACL reconstruction [
10,
13,
19]. Glycerol, the marker of cell death and cell wall degradation exhibited no difference between groups. These findings along with the finding of a higher level of pain during long, but not necessarily complex surgeries lead us to the following speculations. First, that harm done to the synovial cells is irrespective of the magnitude of the knee arthroscopic procedure itself. Second, that postoperative pain is due to extension of the knee cavity caused by the pressure of irrigation fluid, and not primarily due to the surgical trauma itself. Arthroscopy for a long period of time, with a corresponding time of distension, could, therefore, result in more pain. Limitations are that the study is based on cross-sectional data for prevalent patients undergoing knee arthroscopy. Also, there is a smaller number of patients in the RM group compared to the NRM group, which suggests a generally low level of pain caused by arthroscopic surgery.