periodontal disease

(1)

Cardiovascular risks associated with incident and prevalent

periodontal disease

Yu Y-H, Chasman DI, Buring JE, Rose L, Ridker PM. Cardiovascular risks associated with incident and prevalent periodontal disease. J Clin Periodontol 2015;

42: 21–28. doi: 10.1111/jcpe.12335 Abstract

Aim: While prevalent periodontal disease associates with cardiovascular risk, lit- tle is known about how incident periodontal disease influences future vascular risk. We compared effects of incident versus prevalent periodontal disease in developing major cardiovascular diseases (CVD), myocardial infarction (MI), ischaemic stroke and total CVD.

Material and Methods: In a prospective cohort of 39,863 predominantly white women, age≥45 years and free of cardiovascular disease at baseline were followed for an average of 15.7 years. Cox proportional hazard models with time- varying periodontal status [prevalent (18%), incident (7.3%) versus never (74.7%)] were used to assess future cardiovascular risks.

Results: Incidence rates of all CVD outcomes were higher in women with prevalent or incident periodontal disease. For women with incident periodontal disease, risk factor adjusted hazard ratios (HRs) were 1.42 (95% CI, 1.14–1.77) for major CVD, 1.72 (1.25–2.38) for MI, 1.41 (1.02–1.95) for ischaemic stroke and 1.27 (1.06–1.52) for total CVD. For women with prevalent periodontal disease, adjusted HRs were 1.14 (1.00–1.31) for major CVD, 1.27 (1.04–1.56) for MI, 1.12 (0.91–1.37) for ischaemic stroke and 1.15 (1.03–1.28) for total CVD.

Conclusion: New cases of periodontal disease, not just those that are pre-existing, place women at significantly elevated risks for future cardiovascular events.

Yau-Hua Yu¹, Daniel I. Chasman^2,3, Julie E. Buring², Lynda Rose²and Paul M Ridker^2,3

1Division of Periodontology, Department of Oral Medicine Infection and Immunity, Harvard School of Dental Medicine, Boston, MA, USA;²Center for Cardiovascular Disease Prevention and the Division of Preventive Medicine, Brigham Women’s Hospital, Boston, MA, USA;³Harvard Medical School, Boston, MA, USA

Key words: cardiovascular disease;

C-reactive protein; diabetes; family history of MI; periodontal disease; smoking; survival analysis

Accepted for publication 3 November 2014

Evidence accumulating over the past decade demonstrates a significant association between prevalent periodontal disease and future cardiovascular

events (Heidenreich et al. 2011, Eke et al. 2012, Lockhart et al. 2012, To- netti & Van Dyke 2013). Prevalence studies, however, are limited as they typically include wide variation in terms of disease duration prior to study enrolment, a source of variation that can lead to considerable uncertainty with regard to the true magnitude of association between an exposure of interest and a future clinical outcome (Joshipura et al.

1996b, 2003, Hujoel et al. 2000, Wu et al. 2000, Howell et al. 2001, Tu- ominen et al. 2003, Hung et al.

2004, Tu et al. 2007, Dorn et al.

2010, Holmlund et al. 2010, de Oliveira et al. 2010). In contrast, analysis of incident events that accrue during the course of a prospective observational study often provides a better measure of exposure duration and hence a more accurate estimate of risk. Research efforts examining effects of periodontal therapy on subclinical cardiovascular disease (CVD) surrogates such as endothelial function and carotid artery intima–media thickness (Tonetti et al. 2007, Jung Conflict of interest and source of

funding statement

The authors declare no conflict of interest. This study is supported by HL043851, HL080467, HL099355 from the National Heart, Lung and Blood Institute, by CA047988 from the National Cancer Institute, and by the Donald W. Reynolds Foundation.

(2)

et al. 2014), or reducing inflamma- tory markers were abundant (Bokhari et al. 2012, Caula et al. 2014);

however, evidence is still limited to support the relief of CVD burden after periodontal treatment (D’Aiuto et al. 2013). To date, analyses of incident periodontal disease as a determinant of future vascular risk have been sparse (Dietrich et al.

2008, Jimenez et al. 2009).

Common risk factors such as diabetes and smoking are shared between cardiovascular and periodontal diseases (Lockhart et al.

2012, Dietrich et al. 2013, Tonetti &

Van Dyke 2013). Therefore, efforts have been made in prior investiga- tions to examine the association between periodontal disease and cardiovascular outcomes in non- smokers (Dorn et al. 2010) and in non-diabetic patients (Kodovazenitis et al. 2013). However, it remains an open question as to how the inter- relationships between these risk factors and periodontal disease may modify the process of developing cardiovascular disease. Furthermore, genetic studies (Mucci et al. 2009, Schaefer et al. 2009, Ernst et al.

2010) have suggested shared links between periodontal disease and coronary heart disease. Thus, it is of additional interest to evaluate the interaction between family history of myocardial infarction (MI), periodontal disease and future vascular risks.

To address these issues, we exam- ined the relationships of both prevalent and incident periodontal disease to the occurrence of first ever cardiovascular events in a prospective cohort of 39,863 American women followed for an average period of 15.7 years. We further sought evidence of effect modification between periodontal disease, cardiovascular disease and common exposures such as obesity, smoking, diabetes and family history of MI.

Material and Methods

Study population

The Women’s Health Study (WHS) is an NIH-funded prospective cohort that was initiated as a randomized, placebo-controlled trial examining low-dose aspirin and vitamin E in a 2 9 2 factorial design for the primary

prevention of cardiovascular disease and cancer in initially healthy middle- aged women. Details of the study design have been described previously (Ridker et al. 2005a). Between Sep- tember 1992 and May 1995, 39,876 female healthcare professionals in the US, aged 45 years or older, free of cardiovascular disease and cancer were enrolled and have been followed prospectively since that time for incident cardiovascular events. Approxi- mately 72% of WHS participants provided a blood sample at baseline that was stored in liquid nitrogen for subsequent measurement of blood- based biomarkers. The study was approved by the institutional review board of Brigham and Women’s Hos- pital, Boston, Massachusetts.

Self-reported periodontal disease

At study entry, participants in the WHS were asked about whether they had prevalent periodontal disease.

New incident periodontal disease cases were then assessed at 36, 48, 60, 72, 84, 96, 108, 120 months during the randomized trial, and then subsequently at the first follow-up (about 1 year after trial completion) that occurred during the observational extension study. The month and year of the periodontal diagnosis as well as the number of teeth loss during the study interval were requested for the participants report- ing incident periodontal disease.

Cardiovascular events

Participants were followed for cardiovascular endpoints of by annual follow-up questionnaires, letters or telephone calls. Medical records were obtained and reviewed in a blinded fashion for reported endpoints. Car- diovascular deaths were confirmed by family members, postal authori- ties, autopsy reports, medical records and the National Death Index. Myo- cardial infarction was assessed by physician review of medical records according to WHO criteria as well as associated abnormal levels of cardiac enzymes or diagnostic electrocar- diograms. Stroke was confirmed if the participant had a new focal neurological deficit of sudden onset that persisted for more than 24 h.

Clinical information, computed tomographical scan and magnetic

resonance images were used to dis- tinguish haemorrhagic from ischaemic events. Reports of coronary revascularization procedures (bypass surgery or percutaneous coronary angioplasty) were confirmed by record reviews. Composite cardiovascular outcomes were used in the analysis. Major cardiovascular disease (CVD) events included non-fatal myocardial infarction, ischaemic stroke or death from cardiovascular causes. Total CVD events were defined as major CVD as well as bypass surgery, or percutaneous coronary angioplasty.

Assessment of covariates

Details of the study protocol have been described in the primary end- point report of the randomized control trial (Ridker et al. 2005a).

Covariates of interests were collected at the baseline by validated questionnaires and used in the analysis– age, body mass index, race/ethnicity, education, diabetes, hypertension, hypercholesterolaemia, smoking, family history of myocardial infarction and physical activities. Hypertension was defined as a systolic blood pressure of at least 140 mm Hg, a diastolic blood pressure of at least 90 mm Hg or self-reported physician-diagnosed hypertension; hyperlipidaemia was defined as a total cholesterol level of at least 240 mg/dl (6.2 mmol/l) or self-reported physician-diagnosed high cholesterol levels; other covariates were self-reported diagnosis of diabetes or estimation of physical activities. In the subset of WHS participants who provided a blood sample at baseline, plasma lipid fractions and high-sensitivity C-reactive protein (hsCRP) were measured as described (Ridker et al. 2005b).

Statistical analyses

We used time-varying survival analysis to investigate the effects of prevalent as well as incident periodontal disease as determinants of future vascular events. Prevalent periodontal disease (PD) was defined as PD reported as having occurred at or before baseline; while incident periodontal disease, defined as PD having occurred during the follow-up and before incident CVD if any, was encoded as time-varying independent

(3)

variable in the data structure for the survival analysis with Cox proportional hazard model. Similarly, ever- having had PD status was combined from the baseline prevalent PD as well as the time-varying incident PD.

Other covariates used in the analysis were based on the baseline characteristics (Therneau & Crowson 2014). All analyses were performed using the “survival” package of R software. Analyses were adjusted on an a priori basis for age alone as well as for age, race/ethnicity, body mass index, education, smoking, diabetes, hypertension, hypercholesterolaemia, family history of myocardial infarction and physical activity

(model 1). Given that smoking is a strong risk indicator in both periodontal and cardiovascular disease, we adjusted for never- vs. current/

past-smokers as has been shown to be more robust in recent reports (Dorn et al. 2010, Jung et al. 2014).

Estimates were also obtained after further adjustment for C-reactive protein, a biomarker of inflammation (model 2). To assess effect modification by relevant risk factors, analyses were stratified by the pres- ence or absence of obesity, smoking, hypertension, hypercholesterolaemia, a family history of myocardial infarction and diabetes. Survival curves were generated separately for

the component endpoints of myocardial infarction and ischaemic stroke, as well as for the joint endpoints of major CVD and total CVD.

Results

Baseline characteristics

Women having either prevalent or incident periodontal diseases were older compared to those without periodontal disease (mean age: 54.5 and 53.3 versus 52.5; Table 1), more likely to be over-weight or obese, more likely to be current or former smokers (63.6% and 51.2% versus 45.2%), and exercised less frequently.

Table 1. Baseline characteristics of study participants according to the periodontal disease status

Characteristics* No periodontal disease Baseline prevalent periodontal disease Incident periodontal disease pValue^†

No. (%) of patients n= 29,787 (74.7) n= 7185 (18.0) n= 2891 (7.3)

Age 52.5 (48.7–58.4) 54.5 (50.0–60.3) 53.3 (49.1–59.2) <0.001

Body mass index (kg/m²)

<25 14,907 (51.1) 3569 (50.7) 1367 (48.1) 0.012

25~ <30 9020 (30.9) 2159 (30.7) 900 (31.7)

≥30 5240 (18.0) 1308 (18.6) 574 (20.2)

Race/ethnicity

White 28,061 (95.0) 6746 (94.8) 2662 (92.8) <0.001

Other 1481 (5.0) 368 (5.2) 205 (7.2)

Highest education level

<Bachelor’s degree 16,879 (57.7) 3978 (56.3) 1586 (55.6) <0.001

Bachelor’s degree 6800 (23.2) 1572 (22.2) 671 (23.5)

Master’s degree or doctorate 5593 (19.1) 1520 (21.5) 593 (20.8)

Smoking

Current 3254 (11.0) 1501 (20.9) 474 (16.4) <0.001

Past 10,191 (34.2) 3067 (42.7) 1005 (34.8)

Never 16,313 (54.8) 2612 (36.4) 1410 (48.8)

Hypertension

Yes 7574 (25.4) 1983 (27.6) 756 (26.2) <0.001

No 22,206 (74.6) 5201 (72.4) 2134 (73.8)

Hypercholesterolaemia

Yes 8470 (28.4) 2402 (33.4) 868 (30.0) <0.001

No 21,306 (71.6) 4780 (66.6) 2021 (70.0)

Family Hx MI

Yes 3375 (12.6) 891 (13.9) 366 (14.1) 0.003

No 23,449 (87.4) 5523 (86.1) 2228 (85.9)

Hx of diabetes

Yes 780 (2.6) 274 (3.8) 88 (3.0) <0.001

No 29,007 (97.4) 6911 (96.2) 2803 (97.0)

Physical activities

Rarely/Never 11,218 (37.7) 2879 (40.1) 1179 (40.8) <0.001

<1 time per week 5899 (19.8) 1454 (20.3) 571 (19.8)

1–3 times per week 9406 (31.6) 2135 (29.7) 864 (29.9)

4+ times per week 3250 (10.9) 712 (9.9) 276 (9.6)

Total cholesterol (mg/l)^‡ 207 (183–235) 212 (187–238) 209 (183–238) <0.01

HDL (mg/l)^‡ 52.0 (43.2–62.5) 51.5 (42.7–62.0) 51.4 (42.5–61.2) 0.021

LDL (mg/l)^‡ 120.7 (100.1–143.5) 124.0 (102.4–147.1) 123.0 (100.2–146.2) <0.001

Triglyceride (mg/l)^‡ 118 (83–175.0) 124 (87–178) 117 (82–174) <0.001

hsCRP (mg/l)^‡ 2.0 (0.8–4.3) 2.1 (0.9–4.7) 2.0 (0.8–4.3) <0.001

BMI, body mass index; hsCRP, high-sensitivity C-reactive protein; HDL, high-density lipoprotein; LDL, low-density lipoprotein.

*Data are expressed as median (inter-quartile range) for continuous variables and as number (percentage) of participants unless otherwise indicated. Number across categories may not sum to the given total because of the missing data.

†pvalues are based on Kruskal–Wallis test for continuous variables and of the chi-square tests for categorical variables.

‡Number of women having blood samples collected at the baseline and tested were 27,939.

(4)

Serum levels of total cholesterol, LDL cholesterol, triglyceride and hsCRP were higher in women with prevalent periodontal disease at the baseline, and a higher prevalence of hypertension, diabetes and hypercholesterolaemia were also noted in participants with periodontal disease.

We also provide demographical information for women correspond- ing to the different statistical analyses in Table S2. Women having blood samples tended to have more education and fewer cardiovascular events.

Periodontal disease and future cardiovascular events

Overall in the full cohort, compared to women without periodontal disease, age-adjusted hazard ratios for the major CVD events were higher

in women with prevalent [hazard ratio (HR), 1.27 (95% CI, 1.12–

1.43); p< 0.001] or incident [HR, 1.40 (95% CI, 1.15–1.72); p < 0.001]

periodontal disease. These effects remained comparable after adjustment for traditional vascular risk factors (multivariate model 1, Table 2). For women with incident periodontal disease, hazard ratios (HRs) were 1.42 (95% CI, 1.14–1.77;

p= 0.002) for major CVD, 1.72 (95% CI, 1.25–2.38; p < 0.001) for MI, 1.41(95% CI, 1.02–1.95;

p= 0.04) for ischaemic stroke and 1.27(95% CI, 1.06–1.52; p = 0.01) for total CVD after adjusting for established risk factors and physical activity. After further accounting for inflammation indices such as serum level of hsCRP, effects were attenuated but still significant for major CVD [HR, 1.31 (95% CI, 1.01–

1.71); p= 0.045] and MI [HR, 1.65 (1.11–2.45); p = 0.01]. Likewise, for women with prevalent periodontal disease at baseline, adjusted hazard ratios were 1.14 (95% CI, 1.00–1.31;

p= 0.05) for major CVD, 1.27 (95%

CI, 1.04–1.56; p = 0.02) for MI, 1.12 (95% CI, 0.91–1.37; p = 0.3) for ischaemic stroke and 1.15 (95% CI, 1.03–1.28; p = 0.01) for total CVD (Multivariate model 1, Table 2).

Similarly, effects were attenuated after further adjusting for hsCRP, but remained statistically significant for MI [HR, 1.35 (95% CI, 1.05–

1.73); p= 0.02] and total CVD [HR, 1.12 (95% CI, 1.05–1.36); p = 0.007].

Further including aspirin vs. placebo allocation in the multivariate analysis had negligible effects on the hazard ratios (data not shown). The cumulative incidences of these events are presented graphically in Fig. 1

Table 2. Risk of cardiovascular events among women with periodontal disease

Outcome N(Events) No periodontal

disease

Prevalent periodontal disease at baseline

Incident periodontal disease during follow-up*

Major CVD

Incidence rate (95% CI)^† 39,863 (1549) 2.25 (2.11–2.38) 3.24 (2.90–3.57) 2.98 (2.49–3.48)

HR (95% CIs) p HR (95% CIs) p

Age-adjusted^‡ 39,863 (1549) [Reference] 1.27 (1.12–1.43) <0.001 1.40 (1.15–1.72) 0.001 Multivariate model 1^§ 34,228 (1274) [Reference] 1.14 (1.00–1.31) 0.05 1.42 (1.14–1.77) 0.002 Multivariate model 2^¶ 24,033 (883) [Reference] 1.17 (0.99–1.37) 0.06 1.31 (1.01–1.71) 0.045 Myocardial Infarction (MI)

Age-adjusted^‡ 39,863 (642) [Reference] 1.34 (1.12–1.62) 0.002 1.56 (1.15–2.12) 0.005

Multivariate model 1^§ 34,228 (525) [Reference] 1.27 (1.04–1.56) 0.02 1.72 (1.25–2.38) <0.001 Multivariate model 2^¶ 24,033 (353) [Reference] 1.35 (1.05–1.73) 0.02 1.65 (1.11–2.45) 0.01 Ischaemic stroke

Age-adjusted^‡ 39,863 (677) [Reference] 1.28 (1.07–1.54) 0.006 1.40 (1.04–1.90) 0.03

Multivariate model 1^§ 34,228 (558) [Reference] 1.12 (0.91–1.37) 0.3 1.41 (1.02–1.95) 0.04 Multivariate model 2^¶ 24,033 (394) [Reference] 1.05 (0.82–1.35) 0.7 1.28 (0.87–1.88) 0.2 Total CVD

Age-adjusted^‡ 39,863 (2387) [Reference] 1.27 (1.15–1.4) <0.001 1.25 (1.06–1.48) 0.009 Multivariate model 1^§ 34,228 (1968) [Reference] 1.15 (1.03–1.28) 0.01 1.27 (1.06–1.52) 0.01 Multivariate model 2^¶ 24,033 (1388) [Reference] 1.12 (1.05–1.36) 0.007 1.2 (0.97–1.49) 0.1 HR, hazard ratio; CI, confidence interval; CVD, cardiovascular disease; p, p-value, with significance level set at 0.05, highlighted bold and italic.

*Follow up time of incident periodontal disease ended on 31 March 2006 and follow-up of cardiovascular outcomes ended on 14 March 2012.

†Per 1000 person-years of follow-up.

‡Models were adjusted for the age at randomization. These models were based on 1549, 642, 677, 2387 events of major CVD, MI, ischaemic stroke and total CVD, respectively, among 39, 863 women.

§Additionally adjusted for race/ethnicity, body mass index, education, smoking, diabetes, hypertension, hypercholesterolaemia, family history of myocardial infarction and physical activities. These multivariable models were based on 1274, 525, 588, 1968 events of major CVD, MI, ischaemic stroke and total CVD, respectively, among 34,228 women, due to missing data.

¶Additionally adjusted for natural-log-transformed serum levels of C-reactive protein. These multivariable models were based on 883, 353, 394, 1388 events of major CVD, MI, ischaemic stroke and total CVD, respectively, among 24,033 women, due to missing data.

(5)

for those with any evident (i.e. either prevalent or incident) periodontal disease. Supplementary analyses examining risks of cardiovascular events in women having either prevalent or incident periodontal disease versus never are provided in Table S1. As a sensitivity analysis, we also repeated the analyses restricted to women having blood samples (the reduced set). Results were similar that effects were slightly attenuated after further adjusting for hsCRP and that higher CVD risks were found in women having incident vs. prevalent periodontal disease.

As shown in Table 3, the magnitude of association between any evident periodontal disease and future vascular events was less prominent in women who were obese, diabetic or did not smoke. However, none of these differential effects were statisti-

cally significant in analyses assessing for a multiplicative interaction (P > 0.05), except smoking vs. total CVD (p= 0.02); obesity versus major CVD (p= 0.03) and total CVD (p = 0.02). Of note, confidence intervals still overlap between these stratified subsets of women as well as that care should be taken into the consideration of multiple testing given the stratifications (p< 0.05/

6 = 0.008).

Discussion

In this large, prospective cohort of middle-aged women, incident as well as prevalent periodontal disease (PD) was associated with statistically significant increased risks of developing future cardiovascular events.

Moreover, in these data, the magnitude of vascular risk associated with incident periodontal disease was at

least as large as that for prevalent periodontal disease. Thus, these data not only confirm prior work for prevalent periodontal disease, but importantly provide new evidence that incident periodontal disease is also associated with high vascular event rates. We believe these data likely to have relevance for the prac- tice of periodontal medicine as they demonstrate that new cases of periodontal disease, not just those that have been present for long periods of time, put women at significantly elevated risks for vascular disease. It is possible that the lower hazard ratios shown in women having prevalent PD compared to women with incident PD may be due to treatment history or greater awareness of the oral hygiene care. However, this issue is complex and will require additional research. Lastly, our esti- mated cardiovascular risks among

0 5 10 15

0.000.020.040.060.080.10

MAJCVD

Years of follow−up

Never [Reference]

Ever HR(95%CI): 1.20(1.06−1.36) p = 0.0036

0 5 10 15

0.000.020.040.060.080.10

TotalCVD

Never [Reference]

Ever HR(95%CI): 1.17(1.06−1.29) p = 0.0016

0 5 10 15

0.000.020.040.060.080.10

MI

Never [Reference]

Ever HR(95%CI): 1.36(1.13−1.64) p = 0.0011

0 5 10 15

0.000.020.040.060.080.10

iStroke

Never [Reference]

Ever HR(95%CI): 1.177(0.978−1.417) p = 0.085

Fig. 1.Cumulative incidence rates of major cardiovascular disease (CVD), total CVD, myocardial infarction (MI) and ischaemic stroke (iStroke) between women having either prevalent or incident periodontal disease versus never. Hazard ratios and p-values were calculated after accounting for established cardiovascular risk factors, and physical activities (Multivariate model 1 in Table S1).

(6)

women having prevalent periodontal disease were similar to previous pub- lished population studies examining effects of existing periodontal disease on CVD risks (Helfand et al. 2009).

In addition to providing incidence data, the current analysis also provides a comprehensive analysis of effect modification by established risk factors. In these data, risks of vascular events associated with periodontal disease were generally similar in most clinical subgroups, but tended to be preferentially elevated among women who smoked and who were not obese.

In contrast, the magnitude of association between periodontal disease and cardiovascular disease risk generally did not vary in our data according to family history of myocardial infarction, except for a slightly higher risk of MI among women ever having had periodontal disease as well as a family history of MI (Table 3 and Table S4) although this enhanced effect did not meet statistical significance. While this observation is somewhat discordant with prior work (Mucci et al. 2009, Schaefer et al. 2009, Ernst et al.

2010), questionnaire data on family history are qualitative only and not equivalent to or representative of shared genetic risks. In this regard, recent genome-wide association studies have suggested that some loci might have effects on periodontal risk, albeit most of the signals were below genome-wide significance level (Schaefer et al. 2010, Divaris et al.

2012, Teumer et al. 2013). Further research of periodontal disease versus cardiovascular risks in genomic medicine could lead to more insight of the underlying biology.

Strength and limitation

Our study has several strengths including its prospective design, large sample size and confirmation of all incident vascular events. Our approach using time-varying periodontal disease status in the survival analysis is an additional strength as it provides a more accurate assessment of future CVD risk associated with periodontal disease than available from retrospective case–control designs or studies that employ only prevalent periodontal disease. None- theless, the generalizability of our

findings may be limited due to the study population of middle-aged female healthcare professionals. Pre- vious reports from male healthcare professionals in the Physician Health Study (Howell et al. 2001) and the Healthy Professional Follow-Up Study (Joshipura et al. 1996b) did not identify self-reported periodontal disease status as a significant risk factor for cardiovascular disease. Sec- ond, while self-reported periodontal status has been widely used and previously validated for its discrimina- tory power among healthcare professionals (Joshipura et al. 1996a) as well as for a high correlation with PD defined by clinical examination among the general population in the National Health and Nutrition Examination Survey (Eke et al.

2013), it is a far less useful method than direct examination which can address disease severity. Given the design of the questionnaire, we unfor- tunately did not have information regarding tooth loss for the majority of women. Furthermore, history of periodontal treatment, unavailable in our data set, could be a potential con- founder in the association. Third, censored or misclassified periodontal disease status might have led to an underestimation of the magnitude of true effects. Such misclassification, however, would not have resulted in a false-positive finding.

Conclusion

Incident periodontal disease confers at least as high a risk of future cardiovascular events as prevalent periodontal disease in middle-aged women.

Acknowledgement None.

References

Bokhari, S. A., Khan, A. A., Butt, A. K., Azhar, M., Hanif, M., Izhar, M. & Tatakis, D. N.

(2012) Non-surgical periodontal therapy reduces coronary heart disease risk markers: a randomized controlled trial. Journal of Clinical Periodontology 39, 1065–1074.

Caula, A. L., Lira-Junior, R., Tinoco, E. M. &

Fischer, R. G. (2014) The effect of periodontal therapy on cardiovascular risk markers: a 6-month randomized clinical trial. Journal of Clinical Periodontology 41, 875–882.

D’Aiuto, F., Orlandi, M. & Gunsolley, J. C.

(2013) Evidence that periodontal treatment Table3.Age-adjustedriskofcardiovasculareventsamongsubgroupsofwomeneverhavinghadperiodontaldiseases OutcomesMajorCVDMIIschaemicstrokeTotalCVD StratificationaNEHR(95%CIs)aEHR(95%CIs)aEHR(95%CIs)aEHR(95%CIs)a Obese+71233561.04(0.82–1.32)1491.22(0.85–1.75)1481.02(0.7–1.47)5871.05(0.87–1.27) –31,92311471.38(1.22–1.56)***4701.46(1.2–1.78)***5151.39(1.15–1.67)**17271.34(1.21–1.49)*** Smoking+19,4938971.33(1.16–1.53)***4011.39(1.13–1.7)**3681.34(1.09–1.66)**13481.33(1.19–1.49)*** –20,3366481.05(0.87–1.27)2391.08(0.79–1.48)3071.16(0.88–1.51)10351.02(0.88–1.19) Hypertension+10,3147371.21(1.03–1.42)**2981.34(1.05–1.72)*3331.26(1–1.59)11271.18(1.04–1.34)* –29,5428111.37(1.18–1.59)***3441.41(1.13–1.78)**3431.35(1.07–1.7)*12591.33(1.18–1.5)*** Hypercholesterolaemia+11,7416321.43(1.21–1.69)***2751.54(1.2–1.97)**2861.23(0.96–1.58)10631.35(1.19–1.53)*** –28,1089171.18(1.02–1.37)*3671.25(1–1.57)3911.35(1.09–1.68)**13241.17(1.04–1.32)* FamilyHxMI+46332041.28(0.95–1.73)1031.64(1.1–2.45)*751.03(0.62–1.71)3551.14(0.91–1.44) –31,20111481.28(1.13–1.45)***4601.38(1.13–1.69)**5141.3(1.08–1.58)**17271.25(1.13–1.39)*** Hxofdiabetes+11421751.21(0.88–1.67)811.13(0.71–1.82)681.44(0.88–2.36)2931.14(0.89–1.46) –38,72113741.27(1.13–1.43)***5611.39(1.16–1.66)***6091.27(1.06–1.51)**20941.25(1.14–1.37)*** N,numberofwomeninthestratifiedsubset;E,numberofevents;HR,hazardratio;CVD,cardiovasculardisease;MI,myocardialinfarction;CHD,coronaryheartdisease. a***p<0.001;**p<0.01;*p<0.05ashighlightedinbold.

(7)

improves biomarkers and CVD outcomes.

Journal of Clinical Periodontology 40 (Suppl 14), S85–S105.

Dietrich, T., Jimenez, M., Krall Kaye, E. A., Vokonas, P. S. & Garcia, R. I. (2008) Age- dependent associations between chronic periodontitis/edentulism and risk of coronary heart disease. Circulation, 117, 1668–1674.

Dietrich, T., Sharma, P., Walter, C., Weston, P.

& Beck, J. (2013) The epidemiological evidence behind the association between periodontitis and incident atherosclerotic cardiovascular disease. Journal of Clinical Periodontology 40 (Suppl 14), S70–S84.

Divaris, K., Monda, K. L., North, K. E., Olshan, A.

F., Lange, E. M., Moss, K., Barros, S. P., Beck, J. D. & Offenbacher, S. (2012) Genome-wide association study of periodontal pathogen colo- nization. Journal of Dental Research 91, 21S–28S.

Dorn, J. M., Genco, R. J., Grossi, S. G., Falkner, K. L., Hovey, K. M., Iacoviello, L. & Trevisan, M. (2010) Periodontal disease and recurrent cardiovascular events in survivors of myocardial infarction (MI): the Western New York Acute MI Study. Journal of Periodontology 81, 502–511.

Eke, P. I., Dye, B. A., Wei, L., Slade, G. D., Thornton-Evans, G. O., Beck, J. D., Taylor, G. W., Borgnakke, W. S., Page, R. C. & Gen- co, R. J. (2013) Self-reported measures for sur- veillance of periodontitis. Journal of Dental Research 92, 1041–1047.

Eke, P. I., Dye, B. A., Wei, L., Thornton-Evans, G.

O. & Genco, R. J. (2012) Prevalence of periodontitis in adults in the United States: 2009 and 2010. Journal of Dental Research 91, 914–920.

Ernst, F. D., Uhr, K., Teumer, A., Fanghanel, J., Schulz, S., Noack, B., Gonzales, J., Reichert, S., Eickholz, P., Holtfreter, B., Meisel, P., Linden, G. J., Homuth, G. & Kocher, T. (2010) Repli- cation of the association of chromosomal region 9p21.3 with generalized aggressive periodontitis (gAgP) using an independent case- control cohort. BMC Medical Genetics 11, 119.

Heidenreich, P. A., Trogdon, J. G., Khavjou, O.

A., Butler, J., Dracup, K., Ezekowitz, M. D., Finkelstein, E. A., Hong, Y., Johnston, S. C., Khera, A., Lloyd-Jones, D. M., Nelson, S. A., Nichol, G., Orenstein, D., Wilson, P. W. &

Woo, Y. J. (2011) Forecasting the future of cardiovascular disease in the United States: a policy statement from the American Heart Association. Circulation 123, 933–944.

Helfand, M., Buckley, D. I., Freeman, M., Fu, R., Rogers, K., Fleming, C. & Humphrey, L. L.

(2009) Emerging risk factors for coronary heart disease: a summary of systematic reviews con- ducted for the U.S. Preventive Services Task Force. Annals of Internal Medicine 151, 496–507.

Holmlund, A., Holm, G. & Lind, L. (2010) Num- ber of teeth as a predictor of cardiovascular mortality in a cohort of 7,674 subjects followed for 12 years. Journal of Periodontology 81, 870–876.

Howell, T. H., Ridker, P. M., Ajani, U. A., Hennekens, C. H. & Christen, W. G. (2001) Periodontal disease and risk of subsequent cardiovascular disease in U.S. male physicians.

Journal of the American College of Cardiology 37, 445–450.

Hujoel, P. P., Drangsholt, M., Spiekerman, C. &

DeRouen, T. A. (2000) Periodontal disease and coronary heart disease risk. JAMA 284, 1406–

1410.

Hung, H. C., Joshipura, K. J., Colditz, G., Man- son, J. E., Rimm, E. B., Speizer, F. E. & Willett, W. C. (2004) The association between tooth loss and coronary heart disease in men and

women. Journal of Public Health Dentistry 64, 209–215.

Jimenez, M., Krall, E. A., Garcia, R. I., Vokonas, P. S. & Dietrich, T. (2009) Periodontitis and incidence of cerebrovascular disease in men.

Annals of Neurology 66, 505–512.

Joshipura, K. J., Douglass, C. W., Garcia, R. I., Valachovic, R. & Willett, W. C. (1996a) Validity of a self-reported periodontal disease measure.

Journal of Public Health Dentistry 56, 205–212.

Joshipura, K. J., Hung, H. C., Rimm, E. B., Wil- lett, W. C. & Ascherio, A. (2003) Periodontal disease, tooth loss, and incidence of ischemic stroke. Stroke 34, 47–52.

Joshipura, K. J., Rimm, E. B., Douglass, C. W., Trichopoulos, D., Ascherio, A. & Willett, W. C.

(1996b) Poor oral health and coronary heart disease. Journal of Dental Research 75, 1631–1636.

Jung, Y. S., Shin, M. H., Kim, I. S., Kweon, S.

S., Lee, Y. H., Kim, O. J., Kim, Y. J., Chung, H. J. & Kim, O. S. (2014) Relationship between periodontal disease and subclinical atherosclerosis: the Dong-gu study. Journal of Clinical Periodontology 41, 262–268.

Kodovazenitis, G., Pitsavos, C., Papadimitriou, L., Vrotsos, I. A., Stefanadis, C. & Madianos, P. N.

(2013) Association between periodontitis and acute myocardial infarction: a case-control study of a nondiabetic population. Journal of Peri- odontal Research 49, 246–52. doi: 10.1111/

jre.12101. Epub 2013 May 29. PMID: 23713486 Lockhart, P. B., Bolger, A. F., Papapanou, P. N.,

Osinbowale, O., Trevisan, M., Levison, M. E., Taubert, K. A., Newburger, J. W., Gornik, H.

L., Gewitz, M. H., Wilson, W. R., Smith, S.

C. Jr & Baddour, L. M. (2012) Periodontal disease and atherosclerotic vascular disease:

does the evidence support an independent association?: a scientific statement from the American Heart Association. Circulation 125, 2520–2544.

Mucci, L. A., Hsieh, C. C., Williams, P. L., Arora, M., Adami, H. O., de Faire, U., Douglass, C. W. & Pedersen, N. L. (2009) Do genetic factors explain the association between poor oral health and cardiovascular disease? A prospective study among Swedish twins. American Journal of Epidemiology 170, 615–621.

de Oliveira, C., Watt, R. & Hamer, M. (2010) Toothbrushing, inflammation, and risk of cardiovascular disease: results from Scottish Health Survey. BMJ 340, c2451.

Ridker, P. M., Cook, N. R., Lee, I. M., Gordon, D., Gaziano, J. M., Manson, J. E., Hennekens, C. H. & Buring, J. E. (2005a) A randomized trial of low-dose aspirin in the primary prevention of cardiovascular disease in women. New England Journal of Medicine 352, 1293–1304.

Ridker, P. M., Rifai, N., Cook, N. R., Bradwin, G. & Buring, J. E. (2005b) Non-HDL cholesterol, apolipoproteins A-I and B100, standard lipid measures, lipid ratios, and CRP as risk factors for cardiovascular disease in women.

JAMA 294, 326–333.

Schaefer, A. S., Richter, G. M., Groessner- Schreiber, B., Noack, B., Nothnagel, M., El Mokhtari, N. E., Loos, B. G., Jepsen, S. & Schre- iber, S. (2009) Identification of a shared genetic susceptibility locus for coronary heart disease and periodontitis. PLoS Genetics 5, e1000378.

Schaefer, A. S., Richter, G. M., Nothnagel, M., Manke, T., Dommisch, H., Jacobs, G., Arlt, A., Rosenstiel, P., Noack, B., Groessner-Schreiber, B., Jepsen, S., Loos, B. G. & Schreiber, S. (2010) A genome-wide association study identifies GLT6D1 as a susceptibility locus for periodontitis. Human Molecular Genetics 19, 553–562.

Teumer, A., Holtfreter, B., Volker, U., Peters- mann, A., Nauck, M., Biffar, R., Volzke, H., Kroemer, H. K., Meisel, P., Homuth, G. &

Kocher, T. (2013) Genome-wide association study of chronic periodontitis in a general Ger- man population. Journal of Clinical Periodon- tology 40, 977–985.

Therneau, T. & Crowson, C. (2014) Using time dependent covariates and time dependent coeffi- cients in the cox model. Mayo clinic is in Roche- ster, Minnesota. http://cran.r-project.org/web/

packages/survival/vignettes/timedep.pdf.

Tonetti, M. S., D’Aiuto, F., Nibali, L., Donald, A., Storry, C., Parkar, M., Suvan, J., Hingora- ni, A. D., Vallance, P. & Deanfield, J. (2007) Treatment of periodontitis and endothelial function. New England Journal of Medicine 356, 911–920.

Tonetti, M. S. & Van Dyke, T. E. (2013) Peri- odontitis and atherosclerotic cardiovascular disease: consensus report of the Joint EFP/

AAP Workshop on Periodontitis and Systemic Diseases. Journal of Clinical Periodontology 40 (Suppl 14), S24–S29.

Tu, Y. K., Galobardes, B., Smith, G. D., McCar- ron, P., Jeffreys, M. & Gilthorpe, M. S. (2007) Associations between tooth loss and mortality patterns in the Glasgow Alumni Cohort. Heart 93, 1098–1103.

Tuominen, R., Reunanen, A., Paunio, M., Pau- nio, I. & Aromaa, A. (2003) Oral health indica- tors poorly predict coronary heart disease deaths. Journal of Dental Research 82, 713–718.

Wu, T., Trevisan, M., Genco, R. J., Dorn, J. P., Falkner, K. L. & Sempos, C. T. (2000) Peri- odontal disease and risk of cerebrovascular disease: the first national health and nutrition examination survey and its follow-up study.

Archives of Internal Medicine 160, 2749–2755.

Supporting Information

Additional Supporting Information may be found in the online version of this article:

Table S1. Risk of cardiovascular events among women never vs. ever having had periodontal disease.

Table S2. Baseline characteristics of women used in different statistical analyses.

Table S3. Risk of cardiovascular events among women in the reduced set as in the multivariate model 2 analysis.

Table S4. Risk of cardiovascular events among subgroups of women in the reduced set by periodontal disease status and family history of MI.

Address:

Yau-Hua Yu

Harvard School of Dental Medicine 188 Longwood Ave.

Boston MA 02115 USA

E-mails: yauhuayu@gmail.com;

yau-hua_yu@hsdm.harvard.edu

(8)

Clinical Relevance

Scientific rationale for the study:

Does incident periodontal disease, recorded during a prospective observational study, support previously known associations between cardiovascular events and prevalent periodontal disease?

Principle findings: Among middle- aged women, having either incident or baseline prevalent periodontal disease imposed greater chances of developing future myocardial infarction, ischaemic stroke and vascular procedures when compared with women free of periodontal disease. Hazard ratios of women

with incident periodontal disease were at least as high as those with baseline prevalent periodontal disease.

Practical implications: Women with either incident or prevalent periodontal disease are at elevated life- time risk for life-threatening cardiovascular events.