Case-control studyindependent cross-checkfull text checked★ key study I-BAR-07 in-depth review
Fractures after weight-loss surgery in Quebec
Change in fracture risk and fracture pattern after bariatric surgery: nested case-control study.
Overview
Rousseau C et al.. BMJ 2016 · PMID 27814663 · DOI 10.1136/bmj.i3794
Fractures were about 38% more frequent than in obese controls, clearly so only after biliopancreatic diversion.
Does the study support the specific claim? Partly · Methodological quality: moderate
Plain-language summary
Study explained simply
- What did the researchers want to find out?
Fracture
- How was the study done?
Nested case-control
- How many people, animals or samples were studied?
177,464
- What was actually found?
Higher risk
- How large was the observed effect?
RR 1.38
- What do the results mean in everyday life?
Bone health needs follow-up
- What are the study's limitations?
Administrative data; controls not BMI-matched; short follow-up for sleeve; few bypass patients
- What does it say about Elawady's specific claim?
Partly supports
- Do the results apply to Germany and Europe?
Applicable
- What can explicitly not be concluded from this study?
Sleeve-specific risk
Scientific in-depth review
1. Full reference
- Authors
- Rousseau C et al.
- Journal
- BMJ
- Year
- 2016
- Identifier
- PMID 27814663 DOI 10.1136/bmj.i3794
- Verification
- verified – PubMed efetch 2026-10-25
- Retracted
- no
- Corrections
- none
- Extent checked
- full text
- Conflicts of interest and funding
No support for the submitted work; author scholarships (Laval University, Fondation du CHU de Quebec, FRQ-S, Kidney Foundation of Canada KRESCENT). Competing interests: grants from Johnson & Johnson, Valen Tx, GI Windows; honoraria from Amgen, AbbVie, Eli Lilly, BMS; equipment from Roche; advisory boards for Amgen, Eli Lilly, GI Windows.
2. Research question
Fracture risk after bariatric surgery
3. Study design and population
- Study design
- Case-control study – Population-based retrospective matched cohort analysis labelled by the authors as a nested case-control study (Quebec Integrated Chronic Disease Surveillance System administrative data); before-after and between-group comparisons with conditional Poisson regression
- Population
Severely obese adults undergoing bariatric surgery in Quebec (72.3% women, mean age 42.6 years), matched by age (+-3 years) and sex 1:3 to obese and 1:10 to non-obese controls
- Country
Canada
- Period
1 April 2001 - 31 March 2014; procedure type available 2006-2014
- Sample size
12,676 bariatric patients; 38,028 obese and 126,760 non-obese controls; procedure-type subgroup 2006-2014: 9300 surgeries (3887 AGB, 2554 SG, 873 RYGB, 1986 BPD)
- Inclusion and exclusion criteria
Bariatric group: billing code for bariatric surgery plus obesity diagnosis, hospital admission and intervention code; obese controls: >=3 obesity claims within 2 years, one within 6 months of index date, no bariatric surgery; non-obese: no record of obesity or bariatric surgery. Craniofacial, hand, finger and toe fractures excluded
4. Methods and endpoints
- Exposure / intervention
Bariatric surgery
- Comparator
Age- and sex-matched obese (not BMI-matched) and non-obese controls; each group's own pre-index period
- Endpoints
Incident fracture (validated claims algorithm, sensitivity and PPV >=80% for non-vertebral fractures) overall and by site (distal lower limb, clinical spine, pelvis/hip/femur, upper limb)
- Statistical methods
Conditional Poisson regression for matched data adjusted for follow-up duration, material/social deprivation, area of residence, fracture history and number of comorbidities; within-group before/after Poisson models for clustered data; unadjusted life tables
5. Actual results
- Results
Before surgery: fractures 10.5% (bariatric) vs 8.1% (obese) vs 6.6% (non-obese); adjusted RR vs non-obese 1.30 (95% CI 1.21-1.39) and 1.18 (1.13-1.23). After surgery (mean 4.4 years): 514 (4.1%) vs 1013 (2.7%) vs 3008 (2.4%); adjusted RR 1.44 (1.29-1.59) vs non-obese and 1.38 (1.23-1.55) vs obese; median time to first fracture 3.9 years. Within bariatric group after vs before: distal lower limb RR 0.66 (0.56-0.78), upper limb 1.64 (1.40-1.93), clinical spine 1.78 (1.08-2.93), pelvis/hip/femur 2.52 (1.78-3.59). Post-surgery vs non-obese by site: upper limb 1.65 (1.42-1.91), spine 1.70 (1.06-2.73), pelvis/hip/femur 1.88 (1.37-2.58). By procedure (post-surgery adjusted RR vs non-obese): AGB 1.08 (0.83-1.39), sleeve 1.27 (0.86-1.88; median follow-up 1.3 years), RYGB 1.13 (0.67-1.92; n=873, median 2.7 years), BPD 1.60 (1.25-2.03).
- Absolute risks
Post-index fractures over mean 4.4 years: 4.1% bariatric vs 2.7% obese vs 2.4% non-obese; by procedure 2.1% AGB, 1.4% sleeve, 2.6% RYGB, 4.7% BPD vs 1.5% non-obese (2006-2014 subgroup).
- Authors' conclusion
Bariatric patients more likely to fracture; pattern shifts to osteoporotic sites; only BPD clearly associated.
6. Strengths
- Largest population-based study of fracture after bariatric surgery at the time
- Validated fracture algorithm; two control groups and within-person before/after comparison
- Site-specific analysis
- Very low and similar loss to follow-up
7. Weaknesses
- Obese controls not BMI-matched; bariatric group had more comorbidities and baseline fractures (authors)
- Administrative data; no information on medications, physical activity, smoking, alcohol, trauma level, weight loss (authors)
- Only clinical spine fractures captured
- Short follow-up for sleeve (median 1.3 years) and few RYGB (n=873)
- Procedure subgroups not comparable in BMI/comorbidities
- Second peak at year 11 based on small numbers (n=518)
- Mostly Europid population
8. Risk of bias
- Orienting tool
- ROBINS-E
- Type of assessment
- structured assessment based on the full text by a single reviewer (no formal double review)
- Full formal assessment
- no
- Overall risk of bias
- some concerns
- Rationale
Structured single-reviewer judgement from full text; a formal assessment would require two independent reviewers and a protocol.
| Domain | Judgement | Rationale |
|---|---|---|
| Confounding | some concerns | Adjusted for deprivation, residence, fracture history and number of comorbidities; BMI, smoking, drugs, activity not available; baseline fracture risk higher in bariatric group (partly addressed by before/after comparison). |
| Exposure measurement | low | Surgery identified by concordant billing, admission and intervention codes; procedure type only from 2006. |
| Selection of participants | some concerns | Obese controls defined by repeated obesity claims (may be less obese, healthcare users). |
| Post-exposure interventions | some concerns | Supplement and physical activity adherence unknown. |
| Missing data | low | Population coverage 95.3%; low, similar loss to follow-up. |
| Outcome measurement | some concerns | Validated algorithm (sensitivity/PPV >=80%) but vertebral fractures under-ascertained; possible differential surveillance after surgery. |
| Selection of reported result | low | Hypotheses stated a priori; all main comparisons reported. |
Full-text review (25 October 2026): Added pre-surgery RRs, adjusted post-surgery RRs vs both control groups, site-specific RRs vs non-obese, procedure-specific RRs and absolute proportions (sleeve 1.27, 0.86-1.88; BPD 1.60), case definitions, limitations and COI; bias domains added, overall unchanged (some_concerns).
9. Assessment of the statistical interpretation
Moderate relative increase (RR about 1.4) with a site shift towards osteoporotic fractures; absolute excess about 1.4 percentage points over 4.4 years vs obese controls. Clear signal only for BPD; sleeve and RYGB estimates imprecise and non-significant, not evidence of absence.
10. Relevance to the specific claim
Supports C-BAR-06 partly; procedure-specific.
11. Independent contrary findings
12. Transferability to Germany and Europe
Applicable.
13. Overall methodological judgement
Supports increased fracture risk mainly after malabsorptive procedures.
| Multidimensional assessment | |
|---|---|
| Authenticity of the source | verified |
| Methodological quality | moderate |
| Risk of bias | some concerns |
| Precision of results | moderate |
| Directness | direct |
| Reproducibility | replicated |
| Consistency with the overall evidence | mixed |
| Regional transferability | moderate |
| Adequacy of the interpretation (by the archive) | not applicable |
14. Open questions
Fracture risk after sleeve and RYGB with longer follow-up; effect of supplementation and exercise programmes on fractures; BMI-matched comparisons.
15. Linked fact checks
- Bariatric surgery causes measurable hip and spine bone loss (and fracture risk). – Studies supporting the claim