Patellofemoral pain (PFP) is highly prevalent in both adolescent and adult populations and as much as 91% of those affected experience ongoing knee pain for up to 20 years after the initial diagnosis.

Exercise-based treatment focusing on strengthening the muscles around the hip and knee is usually effective when treating patient with PFP. However, this approach has proven ineffective in a subgroup of females who have suffered from PFP for multiple years.

This group was found to have a significantly lower tolerance for pressure stimuli compared to healthy controls. This indicating that central mechanisms within the nervous system rather than the problem being the knee itself.

Exercising a non-painful muscle distant to a painful part of the body has previously been effective in deceasing pain in affected area. This mechanism has been investigated in patients with long-standing musculoskeletal pain as well as healthy populations.

Related research has suggested that a possible link between patients perception of painful stimuli an the pain-reducing effect of exercise exists. As such, patients witch experienced a pain reducing effect of exercise was found to be able to tolerate more pain than before the intervention.

The aim of this study is to assess if exercising a distant non-painful muscle around the hip has a larger acute pain-reducing effect on knee pain compared to knee specific exercises in female patients with long-standing PFP.

Background: Patellofemoral pain (PFP) has a high prevalence in both adolescent and adult populations (7% and 15%) and as much as 91% of those affected experience ongoing knee pain for up to 20 years after the initial diagnosis. Exercise-based treatment focusing on strengthening the muscles around the hip and knee is usually effective when treating patient with PFP. However, this approach has proven ineffective in a subgroup of females with long-standing PFP. This group was found to have altered condition pain modulation indicating facilitated central mechanisms. A localized analgesic effect by exercising a distant non-painful muscle has been observed previously in patients with long-standing musculoskeletal pain. This is known as exercise induced hypoalgesia. In relation, an increased condition pain modulation was found to be a predictor for an increase in the effect of exercise induced hypoalgesia, indicating a link between the two mechanisms. The aim of this study is to assess if exercising a distant non-painful muscle around the hip has a larger acute analgesic effect on localized knee pain compared to knee specific exercises in female patients with long-standing PFP.

Study design: This study is a randomized, controlled, participant and assessor blinded crossover study. The study will be conducted at the Research Unit of General Practice in Aalborg. Reporting of the study will follow the CONSORT guidelines for randomized trials of Non-pharmacologic Treatment.

Method: Thirty women between the age of 20 and 25 with current knee pain which they have experienced ongoing or recurrent for a duration of 5 years will be randomized to either a knee or hip specific resistance exercise as their first intervention. After a washout period of 15 minutes, the participant will perform the other exercise. The participants will undergo a test-procedure before the first exercise, between the exercises and after the final exercise. The test-procedure consist of handheld pressure algometry and cuff pressure algometry assessing pain pressure threshold, pain detection threshold and pain tolerance threshold. Condition pain modulation and temporal summation of pain will be assessed by cuff algometry. In order to assess knee pain in relation to the exercises, the participants will be asked to rate their knee pain before and after each exercise set. The primary outcome will be pain pressure threshold assessed by handheld pressure algometry at the center of patella of the test limb.

Recruitment: A random sample of participants from the APA2011 cohort who experience ongoing PFP in 2016 will be invited to participate in this study. The participants will be contacted by telephone and invited to participate in the study. Verbal information regarding the study will be provided over the phone and written material will be send by e-mail. In case the cohort provide an insufficient number of participants, patients from general practice will be sought out and invited to participate. This process will be done in collaboration with general practice clinics in the area. If the combination of the above-mentioned strategies does not provides the required number of participants, online social-medias like Facebook and Twitter will be used as a mean to obtain a sufficient number of participants.

Allocation sequence: The participants will be randomized to do either sitting knee extension or sidelying hip abduction first. A computer generated allocation sequence will be conducted at www.random.org by an independent researcher. This individual will put slips of paper, which says either "Hip" or "Knee" into opaque sealed envelopes. The envelopes will be marked with sequential numbers. When the first participant have underwent the baseline testing he or she will be assigned the ID number 1 (ID1). After completing the baseline test this participant will proceed to the exercise room where the exercise instructor will be waiting. The exercise instructor will then select the envelope marked "ID1" and note if the participant was assigned to the hip or knee exercise as the first exercise.

Familiarization session: Before each exercise, the load at which the participants can do exactly 12 repetitions will be established by using elastic band with different elastic abilities. This will also work as a familiarization session where the participants can get to know the exercises. For the familiarization session, a black elastic band will be used as default for the sitting knee extension exercise, whereas a yellow elastic band will be used in the sidelying hip abduction exercise. If the participants are capable of performing more or less than 10-12 repetitions with the default elastic band, the length or color will be changed to ensure the participants exercise at 12RM.

Precautions: Participants will be asked to withstand from caffeine, alcohol, nicotine and physically exhausting activities for up to 24 hours before undertaking the study protocol as these have been shown to have a dampening effect on pain. Due to the nature of the study, participants are encouraged to withstand from taking any analgesics on the day for undergoing the study protocol.

Data analysis: A preliminary analysis to see if order of exercise has an effect on the results will be conducted. To investigate if there is any statistically significant difference in PPT, pain detection threshold (PDT) and pain tolerance threshold (PTT) and VAS scores measured at the three time-points, a repeated measure analysis of variance (ANOVA) will be undertaken if data is found to be normally distributed. If data is not normally distributed the Kruskal-Wallis test will applied instead. This data will be presented with a mean, a standard deviation, an f-value and a p-value. For the PPT, PDT and PTT a mean difference of 50 kPa is considered to be clinically relevant.

In addition, the association between knee pain intensity at baseline testing and EIH response for hip versus knee exercises will be investigated. By doing this test it will be possible to assess if those with greater knee pain intensity has a larger EIH response from hip exercises compared to knee exercises.

To assess flare up in pain during exercise the mean change in numeric pain ratings from start to finish of each set within each exercise (hip or knee) will be calculated.

To investigate if there is any correlation between handled PPT measured at the three sites (center of patella on the test limb, the belly of the m. tibialis anterior muscle og the test limb and the epicondyle of the contralateral elbow) a test for correlation will be conducted.

In order to examine if the flare up in pain during exercise behave the same way as TSP, a test for correlation between these two variables will conducted.

Sample size: A group of 10 healthy subjects volunteered to participate in a pilot study prior to recruitment of the current study. Data regarding handheld PPT measured at the center of patella was used to calculate the sample size required to power the study. The mean difference between handheld PPT after hip exercise and after knee exercise within the group was 44 kPa with a corresponding standard deviation of 80 kPa. These data was used in a two-sided power analysis for two paired means which was conducted in STATA. With a significant level of 0.05 and the power set to 0.8, a minimum of 28 participants would be required to power the study. However, 30 participants will be included in order to insure the study is adequately power in case of missing data due to system malfunction or if the effects size turns out to be lower than the anticipated 0.55.

This randomized study is embedded within a larger cross-sectional study which is approved by the ethical committee under the same approval number (N-20160058). This can be accessed on clinicaltrials.gov under the title "Pain mechanisms in Young Adolescent Females With Longstanding Patellofemoral Pain" with Dr. Sinead Holden as the primary investigator.

Inclusion Criteria:

• Participants with PFP are required to have current non-traumatic anterior or retropatellar knee pain which they have experienced ongoing or recurrently for a duration of at least 5 years.
• At least two of the following daily activities should provoke knee-related symptoms: prolonged sitting or kneeling, squatting, running, hopping, or stair walking, tenderness on palpation of the patella or double leg squatting.
• Worst pain during the previous week have to be of more than 3 cm on a 10-cm visual analog scale (VAS).

Exclusion Criteria:

• Patients who have sustained a traumatic injury to the hip, knee, ankle or the lumbar spine within the past 3 month will be excluded.
• Other identifiable knee conditions in isolation (that is, not occuring concurrently with PFP. As an example, if the subject has isolated patellar tendinopathy they will be excluded. If the subject has PFP and concurrent pain at the patellar tendon, they will be included).
• Individuals with rheumatoid arthritis, knee joint effusion, self-reported patellofemoral instability, known malign conditions, neurological disease or previously knee surgery will be excluded as well.
• In cases with suspicion of serious pathology a rheumatologist will be consulted.

IPD data will be available on request after manuscript has been published.

The outcome assessor will be blinded to the order of the exercises and the participants will be blinded to the study-hypothesis.