REF-575 · Cobb, Bachrach, Sowers, Nieves et al., 2007 · Medicine & Science in Sports & Exercise, · Paper · PubMed
2-year RCT in female runners: randomisation to the oral pill was unrelated to bone-density change and did not significantly change stress-fracture incidence (direction protective, inconclusive)
Participants
n=150 competitive female runners aged 18–26 yr: 50 oligo/amenorrhoeic and 100 eumenorrhoeic.
Protocol
Randomised to OC (30 µg ethinyl estradiol + 0.3 mg norgestrel) or control (no intervention) for 2 yr; BMD and BMC measured yearly by DXA; stress fractures confirmed by x-ray, MRI or bone scan.
Key Findings
Intention-to-treat: randomisation to OC was unrelated to changes in BMD or BMC in oligo/amenorrhoeic (n=50) or eumenorrhoeic (n=100) runners.
Treatment-received analysis (actual OC use): oligo/amenorrhoeic runners who used OC gained about 1%/yr in spine BMD and whole-body BMC (P<0.005) — similar to runners who regained periods spontaneously and significantly more than those who stayed oligo/amenorrhoeic (P<0.05). Susceptible to selection bias.
Dietary calcium intake and weight gain independently predicted bone-mass gains in oligo/amenorrhoeic runners.
Stress fractures: randomisation to OC was not significantly related to incidence, but the direction was protective in both menstrual groups (HR 0.57, 95% CI 0.18–1.83) and became stronger in treatment-received analyses.
Authors' conclusion: OC may reduce stress-fracture risk, but the data are inconclusive; oligo/amenorrhoeic athletes with low bone mass should increase dietary calcium and take steps to resume menses, including weight gain — they may benefit from OC, but the evidence is inconclusive.
Practical Application
RCT, 150 competitive female runners, 2 years: randomisation to the pill did not change bone density (ITT), and stress fractures were not significantly reduced (HR 0.57, CI 0.18–1.83, protective direction, inconclusive); only those who actually took the pill and were oligo/amenorrhoeic gained ~1%/yr spine BMD — while dietary calcium and weight gain independently predicted bone gains.
Do not sell the pill as a proven bone-protection strategy for a runner with menstrual disturbance — the randomised evidence is inconclusive, not positive.
For an oligo/amenorrhoeic athlete with low bone mass, the paper's own advice: increase dietary calcium and take steps to resume normal menses, including weight gain; refer to a physician. The pill is at best a possible adjunct.
Frame stress-fracture prevention around load management, energy availability, calcium/vitamin D and menstrual status — not contraceptive choice (not from this paper; the base's own practice, consistent with the trial's conclusion).
Limitations
Power reduced by non-compliance → per-protocol "benefits" selection-biased, genuinely inconclusive for fractures; single OC formulation (older progestin), may not generalise.
How this entry was checked and changed
Why this level
regraded 2026-09-27 from B: parallel RCT with n=150 (rubric: well-powered RCT ≥40 = A-); statistical power reduced by higher-than-anticipated non-compliance, so not graded A
Content correction 2026-09-27
Title/lead said the pill "did NOT change ... stress-fracture rate" -> abstract: not significantly related, direction protective, data inconclusive; reworded. Added the treatment-received BMC gain and the authors' calcium/weight-gain advice.
Level B -> A- per rubric (RCT n=150, power reduced by non-compliance). Sport line set to "running".