Key takeaways
- A 2015 meta-analysis found implant failure in 6.35% of implants in smokers and 3.18% in non-smokers.
- The same authors report a failure risk ratio of 2.23 and tell readers to interpret it cautiously.
- A 2022 meta-analysis found an odds ratio of 1.777 for failure in people with diabetes, which the authors describe as a 77.7% higher risk.
- Neither paper is a personal prognosis. Diabetes care and smoking are medical facts to bring to the consultation.
The smoking percentages below are implants that failed inside the studies, not the chance that your implant fails.[1]
| Finding | Limit |
|---|---|
| Smokers: 1,259 failures among 19,836 implants (6.35%). Non-smokers: 1,923 failures among 60,464 implants (3.18%) | 107 studies. Risk ratio 2.23 (95% CI 1.96–2.53). Authors warn about uncontrolled confounding |
| Postoperative infection risk ratio 2.01 (95% CI 1.09–3.72) | From the studies in that review that reported infection |
| Diabetes: odds ratio 1.777 for implant failure, described by the authors as 77.7% higher | 89 publications. 5,510 implants in people with diabetes and 62,780 in people without |
| Type 1 versus type 2 diabetes: odds ratio 4.477 (p = 0.032) | A subgroup inside that diabetes review, not a reason to guess your type from a blog |
Not sure which option fits your mouth? Nine questions, no obligation, and a personalized next step.
Start with the implant candidate quizWhat do the pooled smoking numbers say?
Chrcanovic, Albrektsson, and Wennerberg included 107 studies. Implants in smokers failed more often: risk ratio 2.23 (95% CI 1.96 to 2.53). Postoperative infection, in the studies that reported it, had a risk ratio of 2.01 (95% CI 1.09 to 3.72). They also found more marginal bone loss, mean difference 0.32 mm, but that analysis was highly heterogeneous (I² 95%), so this page does not treat 0.32 mm as a stable number. The authors say the results should be interpreted with caution because of uncontrolled confounding.[1]
What did the diabetes review measure?
Al Ansari, Shahwan, and Chrcanovic included 89 publications: 5,510 implants in people with diabetes and 62,780 in people without. The odds ratio for failure was 1.777 (p < 0.001). The authors' conclusion calls that a 77.7% higher risk. Failures were more likely in type 1 than type 2 diabetes (odds ratio 4.477, p = 0.032). The failure difference was statistically significant in the maxilla and not in the mandible. Mean marginal bone loss differed by 0.776 mm (p = 0.027); that bone-loss analysis was also highly heterogeneous, so it is a pooled difference, not a measurement of your jaw.[2]
This is not medical advice to change a diabetes medication, and it is not a statement that implants will fail if you have diabetes.
Why these figures are not a yes or a no?
A meta-analysis averages studies that used different patients, jaws, and follow-up. Your physician manages diabetes. The consultation is where the implant surgeon and the denturist hear the history. Tom Szarski, DD, is the denturist. He does not place implants and he does not treat diabetes.
Candidacy in general, including bone, is on what to expect. If the question is bone after years without teeth, read dental implants with bone loss. Healing in the weeks after All-on-4 is a different page: All-on-4 recovery in the first weeks. The quiz sorts which treatment people in a similar situation explore. It does not diagnose diabetes or count cigarettes.
Sources
Frequently asked questions
Does smoking mean a dental implant will fail?
No. In a 2015 meta-analysis, 6.35% of implants in smokers failed and 3.18% of implants in non-smokers failed. The risk ratio was 2.23. The authors say uncontrolled confounding means the result should be read with caution. It is not a personal failure rate.
Does diabetes rule out implants?
Not as a blanket rule. A 2022 meta-analysis of 89 publications found a higher pooled failure risk in people with diabetes (odds ratio 1.777), which the authors describe as 77.7% higher than in people without diabetes. The difference was significant in the upper jaw and not in the lower jaw. It is not a forecast for one person.
Should I tell the clinic?
Yes. Bring your diabetes care and whether you smoke to the consultation. The quiz does not replace that conversation or your physician.
