Associations between periodontitis and CHD/ Associations between periodontitis and CHD/ stroke – how strong is the evidence stroke – how strong is the evidence Thomas Dietrich Thomas Dietrich Boston University School of Dental Medicine Boston University School of Dental Medicine
62
Embed
Associations between periodontitis and CHD/stroke – how ...
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Associations between periodontitis and CHD/Associations between periodontitis and CHD/stroke – how strong is the evidencestroke – how strong is the evidence
Thomas DietrichThomas Dietrich
Boston University School of Dental MedicineBoston University School of Dental Medicine
“old news” 1997
Associated Press, July 9, 1997
Meta-analysis of nine studies on CHD RR= 1.19 (95% CI: 1.08 - 1.32)
rr.5 1 1.5
Combined
DeStephano
Hujoel
Matilla1
Wu_cvd
Morrison
Howell
Joshipura
Beck
Genco
Janket et al., O.O.O. 2003;95:559
Meta-analysis (6 studies) for CHD risk in younger adults (RR = 1.4)
rr.5 1 1.5
Combined
Morrison2
Wu_2
Hujoel2
DeStephano2
Matilla1
Genco
Janket et al., O.O.O. 2003;95:559
Meta-analysis of 5 studies using fatal cardiac events as outcome (RR = 1.5)
rr.5 1 1.5
Combined
Howell_fatal
Hujoel_fatal
Morrison_fatal
Wu_fatal
Beck_fatal
Janket et al., O.O.O. 2003;95:559
Meta-analysis of studies using Stroke/TIA events as outcome
Summary RR = 2.85
95% CI: 1.78 – 4.56Wu et al., Arch Intern Med 2000;160:2749-2755
Beck et al., J Periodontol 1996;67:1123-1137
Janket et al., O.O.O. 2003;95:559
HyperinflammatoryPhenotype
Chronic Periodontitis
CHD
Indirect(e.g. CRP )
Established CVD risk factors
(e.g. smoking, diabetes)
Direct(bacteremia)
Beck et al. 1996, Danesh 1997
Conceptual model
Issues with observational studies
• Issues with epidemiologic studies of periodontitis-CHD association– Exposure measure
• misclassification/measurement error– Self-report vs. clinical measures– Baseline exposure status only
• Conceptualization– (e.g. clinical vs. serological measures)
– Measurement error/misclassification of confounder• Spiekerman et al., J Dent Res 82(5): 345-349, 2003
– Imperfect modeling of confounder• Dietrich et al., J Dent Res 83(11):859-863, 2004
Confounding
Spiekerman et al., J Dent Res 82(5): 345-349, 2003
linear association remaining between mean AL and serum cotinine after adjustment for self-reported intensity (cig/d), age, race and gender (1507 current smokers, NHANES III) .
Confounding
Spiekerman et al., J Dent Res 82(5): 345-349, 2003
• Measurement error/misclassification of confounder– Simulation experiments using joint distribution
of smoking and periodontal variables
– Time to fictional morbid event:t = X•exp(-ß•cotinine/287)
– Cox PH models to estimate association between AL and morbid event (true independent RR = 1)
– Adjustments for self-reported smoking vs. cotinine
• AL and PD of mesiobuccal sites from 12,976 subjects in NHANES III• Bootstrap samples (n=5000)• time to fictional morbid events: t = X * exp (-ln(1.5) * AL)
– X : exponential random variable– AL: mean AL per subject– RR of morbid event for an increase in mean AL of 1 mm was 1.5
• fixed censoring time was chosen to yield a cumulative incidence of 5%
• Periodontal disease defined using three previously used criteria: – (1) at least 1 site with PD 4 mm– (2) at least 4 sites with PD 4 mm– (3) at least 1 site with PD 6 mm.
Meta-analysis of nine studies on CHD RR= 1.19 (95% CI: 1.08 - 1.32)
rr.5 1 1.5
Combined
DeStephano
Hujoel
Matilla1
Wu_cvd
Morrison
Howell
Joshipura
Beck
Genco
Janket et al., O.O.O. 2003;95:559
New Epidemiologic Studies
• Case control study• 303 incident cases of acute cerebral ischemia• 300 population controls• 168 hospital controls• Periodontal status assessed clinically and
radiographically• periodontal probing at 4 sites in all teeth• Subjects were categorized into 4 categories based on
means clinical attachment levels (CAL), which allowed evaluation of the dose-response pattern.
Grau et al., Stroke 2004
Grau et al., Stroke 2004
1 a SR with homogeneity of RCTs b Individual RCT with narrow Confidence Interval c All or none
2 a SR with homogeneity of cohort studies b Individual cohort study (including low quality RCT; e.g., <80% follow-up) c "Outcomes" Research; Ecological studies
3 a SR with homogeneity of case-control studies b Individual Case-Control Study
12 Case-series and poor quality cohort and case-control studies
13 Expert opinion without explicit critical appraisal, or based on physiology, bench research or "first principles"
Hierarchy (levels) of evidence
Efficacy of AB for secondary CVD prevention: major RCT
• antibiotic arm of the PROVE-IT trial• Inclusion: patients with an acute coronary syndrome• randomized to placebo (n=2,086) or monthly courses of
• no other subgroup could be identified in which the treatment was effective
Cannon et al., New Engl J Med 2005
Efficacy of AB for secondary CVD prevention: major RCT
• Inclusion: patients with documented stable CHD• Randomiyed to placebo (n=2,008) or 600mg azithromycin once per
week for a year (n=2,004)• Patients followed for up to 4.5 years • composite primary endpoint (fatal MI, revascularization, unstable
AP)• Cumulative incidence identical: placebo: 22.4%; treatment: 22.3%• Conclusion: no clinically significant benefit in the secondary
prevention of CHD events from one year of weekly azithromycin treatment
• neither Chlamydia pneumoniae nor another organism susceptible to azithromycin plays an important role in the pathogenesis of CHD events in subjects with advanced CHD.