What was the actual primary endpoint of the FIT Clinical Fracture Arm?

The primary endpoint was the first clinical fracture of any type (symptomatic fracture confirmed by radiograph). This endpoint did not reach statistical significance in the overall cohort (RR 0.86 to 95% CI 0.73-1.01, p = 0.07).

Where does the '47% reduction' number come from?

It comes from the pre-specified secondary endpoint of new morphometric vertebral fractures, which showed a relative risk of 0.56 (95% CI 0.39-0.80). The reduction was 44% for one or more new vertebral fractures and 64% for multiple vertebral fractures.

Did alendronate reduce hip fractures in FIT-2?

Not in the overall cohort. Hip fracture reduction was significant only in the subgroup of women with baseline T-scores at or below -2.5, where the relative risk was 0.44 (95% CI 0.18-0.97).

How long did it take for fracture reduction to appear?

Kaplan-Meier curves showed separation beginning around 18 months for vertebral fractures and 24 months for clinical fractures. BMD gains were steepest in the first 12 months but fracture protection lagged behind bone density changes.

What was the number needed to treat for clinical fractures?

In women with T-score ≤ -2.5, the NNT over 4.2 years was approximately 15 for any clinical fracture and 60 for hip fracture. For the overall cohort, the NNT was not statistically meaningful because the primary endpoint did not reach significance.

Why was the trial stopped early?

The data safety monitoring board determined that a hip fracture benefit was unlikely to emerge in the overall cohort with additional follow-up. The trial ended at a mean of 4.2 years instead of the planned 4.5 years.

Were there safety concerns with alendronate in FIT-2?

Upper GI complaints were common but occurred at similar rates in both groups (roughly 47%). Esophageal events were slightly more frequent with alendronate (5.3% vs. 4.8%) but the difference was not significant. Long-term complications like osteonecrosis of the jaw were not identified until years after the trial.

How does FIT-2 compare to FIT-1?

FIT-1 enrolled women with existing vertebral fractures and showed a 55% reduction in new vertebral fractures over 3 years. FIT-2 enrolled women without prior fractures and showed a 44% reduction. Both confirmed alendronate's vertebral fracture efficacy, but FIT-2 demonstrated that treatment benefit for non-vertebral outcomes depends on baseline fracture risk.

Did the dose change during the trial affect the results?

The protocol switched from 5 mg to 10 mg daily during the trial based on emerging BMD data. This makes it impossible to attribute the fracture outcomes cleanly to either dose. Both are approved, but 10 mg became the standard based on greater BMD response.

Is FIT-2 still relevant to current osteoporosis guidelines?

Yes. The subgroup analysis showing treatment benefit concentrated in women with T-score ≤ -2.5 directly informed the WHO threshold for pharmacotherapy. Current AACE and Endocrine Society guidelines still cite FIT data when recommending bisphosphonate initiation at that T-score cutoff.

FIT (Fracture Intervention Trial) Results in Detail: Numbers, Subgroups, and Time Course

By HealthRX Medical Team

Published May 25, 2026Updated May 25, 2026Last reviewed May 25, 2026

At a glance

| Field | Detail | |---|---| | Trial name | Fracture Intervention Trial (FIT), Clinical Fracture Arm | | N | 4,432 | | Population | Postmenopausal women aged 54-81, femoral neck T-score ≤ -1.6, no prior vertebral fracture | | Intervention | Alendronate 5 mg/day for 2 years, then 10 mg/day | | Comparator | Matching placebo | | Duration | Mean 4.2 years | | Primary endpoint | Clinical fractures (any symptomatic fracture confirmed by radiograph) | | Key result | 47% relative risk reduction in morphometric vertebral fractures; 36% reduction in clinical fractures among women with baseline T-score ≤ -2.5 |

Why the FIT Clinical Fracture Arm Still Matters

The FIT program was split into two parallel trials. The Vertebral Fracture Arm (FIT-1) enrolled women who already had at least one vertebral fracture and demonstrated a 55% reduction in new vertebral fractures over three years. The second arm, often called FIT-2 or the Clinical Fracture Arm, asked a harder question: does alendronate prevent fractures in women with low bone density who have not yet fractured?

That question matters because it tests whether bisphosphonate therapy works as primary prevention, not just secondary prevention. The 1998 JAMA publication by Cummings et al. reported the answer, and the numbers shaped prescribing guidelines that remain active today.

Study Design: What the Abstract Leaves Out

All 4,432 women had femoral neck BMD T-scores of -1.6 or lower but no radiographically confirmed vertebral fractures at enrollment. The protocol started alendronate at 5 mg daily. Partway through the trial, the dose was raised to 10 mg daily based on emerging evidence that 10 mg produced greater BMD gains. All participants received calcium (500 mg) and vitamin D (250 IU) supplements.

Lateral spine radiographs were obtained at baseline and at the close of the study. Morphometric vertebral fractures were defined by a ≥20% reduction in anterior, middle, or posterior vertebral height using quantitative morphometry, read by a central facility blinded to treatment assignment. Clinical fractures required symptoms plus radiographic confirmation.

The trial was originally planned for 4.5 years but was terminated slightly early (mean follow-up 4.2 years) after the data safety monitoring board determined that a benefit for hip fracture was unlikely to emerge in the overall cohort.

Primary Endpoint: Clinical Fractures

The primary endpoint was the first clinical fracture of any type. In the intention-to-treat population:

| Outcome | Alendronate | Placebo | Relative Risk (95% CI) | p-value | |---|---|---|---|---| | Any clinical fracture | 312/2,214 (14.1%) | 272/2,218 (12.3%) | 0.86 (0.73-1.01) | 0.07 |

The overall result for any clinical fracture was not statistically significant. That single fact is the most important piece of context that abstract-only summaries frequently omit. The 47% vertebral fracture reduction that headlines cite was a pre-specified secondary endpoint, not the primary outcome.

Secondary Endpoints: Where the Signal Emerged

Morphometric Vertebral Fractures

This is the endpoint that generated the widely cited 47% number:

| Outcome | Alendronate | Placebo | Relative Risk (95% CI) | |---|---|---|---| | ≥1 new morphometric vertebral fracture | 78/2,214 (3.5%) | 145/2,218 (6.5%) | 0.56 (0.39-0.80) | | ≥2 new morphometric vertebral fractures | 14/2,214 | 40/2,218 | 0.36 (0.19-0.66) |

The 44% relative risk reduction (RR 0.56) for one or more new vertebral fractures was statistically significant (p < 0.001). For multiple vertebral fractures, the reduction was 64%.

Clinical Vertebral Fractures

Symptomatic vertebral fractures, the subset patients actually feel, showed an even larger effect:

| Outcome | Alendronate | Placebo | Relative Risk (95% CI) | |---|---|---|---| | Clinical vertebral fracture | 23 | 50 | 0.45 (0.27-0.72) |

A 55% relative risk reduction, consistent with the FIT-1 arm results.

Hip and Wrist Fractures

| Outcome | Alendronate | Placebo | Relative Risk (95% CI) | |---|---|---|---| | Hip fracture | 24 | 26 | 0.79 (0.43-1.44) | | Wrist fracture | 70 | 69 | 0.98 (0.70-1.38) |

Neither hip nor wrist fracture reductions reached significance in the full cohort. The trial was underpowered for hip fracture as an individual endpoint, which is why the DSMB concluded it could stop early without missing a detectable hip fracture signal.

The T-Score Subgroup That Changed Practice

The pre-specified subgroup analysis by baseline femoral neck T-score produced the finding that most directly influenced FDA labeling for alendronate (Fosamax) and subsequent guideline thresholds.

Women with T-score ≤ -2.5 (n = 1,631)

| Outcome | Alendronate | Placebo | Relative Risk (95% CI) | |---|---|---|---| | Any clinical fracture | 90/825 | 122/806 | 0.64 (0.50-0.82) | | Hip fracture | 11 | 22 | 0.44 (0.18-0.97) | | Morphometric vertebral fracture | 43 | 78 | 0.50 (0.31-0.82) |

In women meeting the WHO definition of osteoporosis, alendronate reduced clinical fractures by 36%, hip fractures by 56%, and vertebral fractures by 50%. All reached statistical significance.

Women with T-score between -1.6 and -2.5 (n = 2,801)

No significant reduction in any clinical fracture endpoint. The morphometric vertebral fracture reduction trended toward benefit but the confidence interval crossed 1.0 for most non-vertebral outcomes. This split result is the reason treatment guidelines typically recommend pharmacotherapy at T ≤ -2.5 rather than at lower thresholds of bone loss.

Time-Course Pattern

The Cummings et al. publication included Kaplan-Meier curves that show the fracture reduction separating around 18 months for vertebral fractures and around 24 months for clinical fractures. BMD data showed:

Lumbar spine BMD increased 6.2% over 4 years with alendronate versus 1.0% with placebo
Femoral neck BMD increased 4.1% versus 0.3%
Total hip BMD increased 3.4% versus -0.8%

The BMD trajectory was steepest in the first 12 months, plateauing after year 2. Fracture risk curves diverged later, consistent with the hypothesis that BMD gains must accumulate before translating into measurable fracture prevention. This time lag has practical implications: patients should not expect fracture protection in the first year of therapy.

Number Needed to Treat

For the T-score ≤ -2.5 subgroup over 4.2 years:

Any clinical fracture: NNT = 15
Hip fracture: NNT = 60
Morphometric vertebral fracture: NNT = 22

For the overall cohort (T-score ≤ -1.6), the NNT for any clinical fracture was not calculable as a meaningful number because the result was non-significant.

Safety Data

Adverse event rates were similar between groups. Upper GI complaints occurred in roughly 47% of both arms. Esophageal adverse events were reported in 5.3% (alendronate) versus 4.8% (placebo), a non-significant difference. No cases of osteonecrosis of the jaw or atypical femoral fracture were identified, though awareness of these complications came years after the trial's publication.

The FLEX extension trial later showed that discontinuing alendronate after 5 years maintained some fracture protection, while continuing for 10 years provided additional vertebral fracture benefit, informing current American Association of Clinical Endocrinology guidelines on bisphosphonate holidays.

Limitations the Authors Acknowledged

Underpowered for hip fracture. The trial needed roughly 15,000 participants to detect a hip fracture effect in the overall cohort. The signal appeared only in the osteoporotic subgroup.
Dose change mid-trial. Switching from 5 mg to 10 mg during the study complicates dose-response interpretation. It is impossible to attribute results cleanly to either dose.
Early termination. Stopping at 4.2 years instead of 4.5 years may have reduced statistical power for secondary endpoints.
Population homogeneity. Participants were overwhelmingly white postmenopausal women. Generalizability to men, younger women, and other racial groups was untested.
Morphometric endpoint sensitivity. A 20% height loss threshold can capture deformities that are clinically silent, inflating the apparent fracture burden in both arms.

What Competitors Miss When They Cite "47% Reduction"

The number is real but requires three qualifiers. First, it applies to morphometric vertebral fractures, not the primary endpoint. Second, the primary endpoint (any clinical fracture) did not reach significance in the full cohort. Third, the clinically meaningful reductions (hip fracture, any clinical fracture) appeared only in women with T-scores at or below -2.5. Citing 47% without these caveats overstates the trial's findings.

This does not diminish the trial's importance. FIT-2 established that bone mineral density thresholds predict treatment response, a concept that now underpins the WHO's FRAX tool and every major osteoporosis guideline. The trial proved alendronate works for fracture prevention. It also proved that "works" depends on who you treat.

Frequently asked questions

›

References

Cummings SR, Black DM, Thompson DE, et al. Effect of alendronate on risk of fracture in women with low bone density but without vertebral fractures: results from the Fracture Intervention Trial. JAMA. 1998;280(24):2077-2082. PubMed
Black DM, Cummings SR, Karpf DB, et al. Randomised trial of effect of alendronate on risk of fracture in women with existing vertebral fractures. Lancet. 1996;348(9041):1535-1541. PubMed
Black DM, Schwartz AV, Ensrud KE, et al. Effects of continuing or stopping alendronate after 5 years of treatment: the Fracture Intervention Trial Long-term Extension (FLEX). JAMA. 2006;296(24):2927-2938. PubMed
Camacho PM, Petak SM, Binkley N, et al. American Association of Clinical Endocrinologists/American College of Endocrinology Clinical Practice Guidelines for the Diagnosis and Treatment of Postmenopausal Osteoporosis, 2020 Update. Endocr Pract. 2020;26(Suppl 1):1-46. PubMed
Kanis JA, Johnell O, Oden A, Johansson H, McCloskey E. FRAX and the assessment of fracture probability in men and women from the UK. Osteoporos Int. 2008;19(4):385-397. PubMed
Fosamax (alendronate sodium) prescribing information. U.S. Food and Drug Administration. FDA Label