Patellofemoral pain syndrome is what people call “runner's knee” in conversation. A dull ache around or under the kneecap that gets worse on descents, on stairs and after sitting for a long time. It is the most common running complaint, and the prescription for it on the internet is always the same: your glutes are weak, the pelvis drops, this increases the Q-angle — the angle between the line of pull of the quadriceps and the patellar tendon — and the kneecap stops running in its groove. Train hip abduction.

The logic looks flawless. In September 2026 PLoS ONE published a study that tested it on marathon runners.

What was done

Bai, Cao and Feng selected 487 marathon runners with diagnosed patellofemoral syndrome, of whom 312 met the criteria. Then came propensity score matching at a 1:1 ratio, so that the groups would not differ in their baseline characteristics. That left 94 people in each:

Hip-STAB — a hip stability programme;
control — standard care.

Three things were measured at entry and at 4, 8 and 16 weeks: the standing Q-angle, pain while running on a visual analogue scale (VAS-Run, 0–100 mm) and the anterior knee pain scale AKPS (0–100 points).

What came out

Both groups improved. Mixed linear models showed a significant effect of time on all three measures (all p < 0.001): the Q-angle decreased, pain while running fell, function rose.

There is no difference between the groups. The time × group interaction was non-significant: Q-angle p = 0.190, pain while running p = 0.302, AKPS p = 0.317. Post hoc comparisons with a Bonferroni correction found no differences at week 16 either.

The Q-angle and pain are related — but equally in both groups. A reduction in the Q-angle correlated with a reduction in pain both in the hip exercise group (r = 0.35; 95% CI 0.16–0.52) and in the control group (r = 0.44; 95% CI 0.26–0.59).

The authors word their conclusion carefully: the link between the Q-angle and pain should be read as an association, not as proof of a biomechanical mechanism. Put simply, the Q-angle goes down in people who are getting better — but that does not mean it was the cause.

What the study does not prove

This is the point at which it is important not to jump to the opposite extreme.

It is a retrospective cohort, not an RCT. Propensity score matching evens out known differences, but not the ones that are missing from the database: motivation, adherence, running volume during treatment.

“Standard care” is not doing nothing. People in the control group received treatment too, and judging by the numbers it worked. The result reads as “the hip programme is no better than ordinary care”, not as “exercise is unnecessary”.

An absence of difference is not proof of equivalence. 94 people per group is enough for a noticeable effect, but a design like this could well have missed a small advantage.

The Q-angle was measured standing, statically. Dynamic knee valgus while running is a different measurement, and the authors state outright that the clinical value of the static Q-angle remains unclear.

What to do if your knee already hurts

The consensus from the international patellofemoral pain research retreats (British Journal of Sports Medicine, 2016, part 2, Crossley's group) still stands: exercise is first-line therapy, and it makes sense to work on both the hip and the knee rather than picking one. The new study does not overturn that recommendation — it removes the excessive promise that a glute programme is by itself better than everything else.

The second practical conclusion is about time. Both in this study and in the consensus the time frames are 6 to 16 weeks. Expecting it to “pass in two weeks” fits no dataset at all.

How to apply this

Do not stop running completely, but remove the provocation. Descents, long staircases and a sharp rise in volume are the three things that usually drive this pain. The guideline is simple: if the pain during a run is no higher than 3 out of 10 and is not worse by the next morning, the volume is tolerable.
Do exercises for both the hip and the knee. Hip abduction and external rotation plus knee extension through a pain-free range — that is what the consensus recommends, and it is what both groups in the new study were built on.
Do not chase the Q-angle number. Its change accompanies recovery but does not drive it. There is no point measuring the angle with a ruler and celebrating a couple of degrees.
Count the time in weeks, not days. Judging a programme before the sixth week is pointless; a sensible checkpoint is 12–16 weeks.
Record your starting point objectively. Progress in the strength tests and the trend in pain on a simple 0–10 scale will tell you more than a feeling that things are “sort of better”.
If the pain does not go away after 16 weeks of honest work, that is a reason not to change your exercise routine for the third time but to rule out other causes of pain at the front of the knee.

The bottom line

In the 2026 study, 188 marathon runners with patellofemoral syndrome were split between a hip stability programme and standard care; over 16 weeks both groups improved to the same extent (p = 0.302 for pain, p = 0.317 for function).
The static Q-angle decreased in those who were getting better, in both groups — that is an association, not a mechanism, and aiming at it separately makes no sense.
The design is retrospective, so the conclusion is more correctly read as “the hip programme does not beat ordinary care” rather than “exercise is useless”.
Exercise remains first-line therapy under the international consensus, the work should cover both the hip and the knee, and the horizon for judging it is 6–16 weeks, not days.

Sources: Bai D., Cao X., Feng S. “Effects of hip stability training versus standard care on Q-angle and pain in marathon runners with patellofemoral pain syndrome: A retrospective cohort study”. PLoS ONE, 2026;21(9):e0357057. DOI: 10.1371/journal.pone.0357057 · Collins N.J., Barton C.J., van Middelkoop M. et al. “2018 Consensus statement on exercise therapy and physical interventions for patellofemoral pain”. British Journal of Sports Medicine, 2018;52(18):1170-1178. DOI: 10.1136/bjsports-2018-099397 · Crossley K.M., van Middelkoop M., Callaghan M.J., Collins N.J., Rathleff M.S., Barton C.J. “2016 Patellofemoral pain consensus statement from the 4th International Patellofemoral Pain Research Retreat, Manchester. Part 2: recommended physical interventions”. British Journal of Sports Medicine, 2016;50(14):844-852. DOI: 10.1136/bjsports-2016-096268