Periodontal disease may be associated with increased breast cancer risk, but studies have not considered invasive breast cancer and ductal carcinoma in situ (DCIS) separately in the same population. We assessed the relationship between periodontal disease and breast cancer in a large prospective cohort study. The Sister Study followed women without prior breast cancer ages 35 to 74 years from 2003 to 2017 (N = 49,968). Baseline periodontal disease was self-reported, and incident breast cancer was ascertained over a mean follow-up of 9.3 years. We estimated hazard ratios (HR) and 95% confidence intervals (CI) using Cox proportional hazards regression, adjusting for multiple potential confounders, including smoking status. Heterogeneity in risk for invasive breast cancer versus DCIS was also estimated. About 22% of participants reported a history of periodontal disease at baseline. A total of 3,339 incident breast cancers (2,607 invasive breast cancer, 732 DCIS) were identified. There was no clear association between periodontal disease and overall breast cancer risk (HR = 1.02; 95% CI, 0.94–1.11). However, we observed a nonstatistically significant suggestive increased risk of invasive breast cancer (HR = 1.07; 95% CI, 0.97–1.17) and decreased risk of DCIS (HR = 0.86; 95% CI, 0.72–1.04) associated with periodontal disease, with evidence for heterogeneity in the risk associations (relative HR for invasive breast cancer versus DCIS = 1.24; 95% CI, 1.01–1.52). A case-only analysis for etiologic heterogeneity confirmed this difference. We observed no clear association between periodontal disease and overall breast cancer risk. The heterogeneity in risk associations for invasive breast cancer versus DCIS warrants further exploration.

Characterized by chronic infection and inflammation (1), the prevalence of periodontal disease increases with age. Approximately 34% of U.S. female adults 30 years or older were affected by periodontal disease during 2009 to 2014 (2). Accumulating evidence suggests that periodontal disease is associated with increased risk of several cancers, especially oral cancer, lung cancer, and gastrointestinal cancer, with a hypothesized role of immune-inflammatory mechanisms that may be common to both periodontal disease and cancers (3–8).

Inflammation has also been implicated in the etiology of breast cancer, which is thought to be driven by immune system cells and associated inflammatory mediators in the tumor microenvironment (9). Studies also indicated that both breast tumor tissue and normal adjacent breast tissue from patients with breast cancer have different bacterial profiles from healthy controls (10, 11). Investigators have hypothesized that periodontal pathogens may directly influence breast carcinogenesis, citing evidence for the presence of oral bacterial species in breast tissues (12), whereas others have suggested that periodontal disease may influence and/or reflect systemic inflammatory processes that promote breast carcinogenesis (9, 13). However, no consensus has been reached for the relationship between periodontal disease and the risk of breast cancer, the most common cancer diagnosed among American women (14). Some studies have observed an increased risk of breast cancer among women with periodontal disease (6, 15, 16), whereas others have detected no relationship between periodontal disease and breast cancer risk (4, 17, 18). Two recent meta-analyses of studies reported that periodontal disease may increase breast cancer risk by about 1.2-fold (19, 20); however, notable limitations of prior studies included small sample sizes and inadequate adjustment for potential confounding factors. In prior studies, the sample sizes ranged from 151 (8) to 80,280 (6), with the Women's Health Initiative Observational Study (WHI OS; ref. 15) having the most breast cancer cases (N = 2,124), and all other studies each having fewer than 550 breast cancer cases. Larger prospective studies with robust multivariable adjustment, including controlling for the potential confounding effects of smoking, are needed to further examine the association between periodontal disease and breast cancer risk.

Studies of periodontal disease appear to have only considered risk of overall breast cancer or invasive breast cancer, with no prior study having assessed the risk for invasive breast cancer and ductal carcinoma in situ (DCIS) separately. Although DCIS is considered to be a nonobligate precursor of invasive breast cancer (21), invasive breast cancers with or without history of DCIS have shown differences in tumor characteristics and protein expression, suggesting potential differences in etiologic risk factors (22). Furthermore, any observed differences in risk of invasive breast cancer versus DCIS associated with periodontal disease may point to important differences in behaviors and method of detection, because DCIS is largely identified through screening mammography (23).

Given the limited and conflicting evidence on the association between periodontal disease and breast cancer risk, we aimed to assess the relationship between periodontal disease and breast cancer, separately for risk of DCIS and invasive breast cancer, controlling for potential confounding factors, in a large, prospective cohort study.

Study population

The Sister Study is a nationwide prospective cohort study, which was designed to study environmental and genetic risk factors for breast cancer (24). Briefly, 50,884 women ages 35 to 74, who had a sister diagnosed with breast cancer, but were without prior breast cancer themselves, were enrolled during 2003 to 2009. At baseline, characteristics of participants, including sociodemographic characteristics and reproductive information, were obtained via a computer-assisted telephone interview. Participants were contacted annually to complete a short questionnaire or a detailed follow-up questionnaire. The data presented in this study were obtained from Sister Study Data Release 7.0, which includes follow-up through September 15, 2017. The Sister Study was approved by the institutional review boards of the National Institute of Environmental Health Sciences and the Copernicus Group. All participants provided written informed consent.

Analytic population

Women diagnosed with breast cancer prior to completing baseline activities or with unknown timing of breast cancer diagnosis relative to baseline (N = 80) were excluded. Women who reported never having a menstrual period were also excluded (N = 7). For participants with missing menopausal status or age at menopause (N = 360), menopausal status and age at menopause were imputed using age 55 as the presumed age at menopause (24). Those who were missing values for other key variables were excluded from our analyses as follows: missing baseline periodontal disease history (N = 76), missing potential confounding factors (N = 753). After exclusions, a total of 49,968 women were included in this analysis.

Periodontal disease and breast cancer ascertainment

At baseline, participants were asked, “have you ever been told you had periodontal or gum disease?” They also were asked if they had been to a dentist within the 12 months prior to baseline.

Breast cancer diagnosis was self-reported during follow-up through 2017, and for 82.3% of the reported breast cancers medical records were acquired, which all confirmed the self-report (Sister Study Data Release 7.0, https://sisterstudy.niehs.nih.gov/English/images/SIS-BrCa-Inv-DCIS-DR7-508.pdf; ref. 25). Self-report was used when the medical record was not available given the high agreement between self-reported diagnosis and the medical records (24). The International Statistical Classification of Diseases and Related Health Problems 10th Revision (ICD-10) was used to define DCIS and invasive breast cancer. Reports of patients' first DCIS or invasive breast cancer, regardless of possible prior lobular carcinoma in situ (LCIS) diagnosis, were used in our analyses.

Assessment of potential confounders

Factors that were associated with both periodontal disease and breast cancer risk were identified as potential confounders. We obtained baseline information for the following factors: age, education level (high school or less, some college or associate degree, bachelor's degree, master's or doctoral degree), race/ethnicity (non-Hispanic White, non-Hispanic Black, Hispanic and other), smoking status (never, former, current), metabolic equivalent (MET)-hours physical activity per week that summed over self-reported sports/exercise and daily activities (≤27.06, >27.06 and ≤44.37, >44.37 and ≤67.21, >67.21), body mass index (BMI) was estimated using measured weight and height from examiner data, regular antibiotic use (never or unknown, former, current regular use, defined as ≥3 times per week for 3 months in a row or longer), menopausal hormone therapy (MHT) ever use [never, estrogen combined with progestin therapy (with or without ever use of estrogen-only therapy history), estrogen-only therapy, other therapy], tamoxifen use (never, ever) and age at first birth (nulliparity, <20, 20 to <30, ≥30). Menopausal status and age at menopause were assessed at baseline and updated on the basis of subsequent follow-up questionnaires.

Statistical analysis

Multivariable stepwise logistic regression and Cox proportional hazards regression were used to identify potential confounding factors associated with periodontal disease and breast cancer risk, respectively, and variables with an alpha level of 0.10 were retained in the model, in addition to a priori knowledge of potential confounders (race/ethnicity, BMI × menopausal status interaction).

We estimated hazard ratios (HR) and 95% confidence intervals (CI) for the association between periodontal disease and breast cancer risk using multivariable Cox proportional hazards regression, with age as the primary time scale, adjusting for age at baseline, education, race/ethnicity, smoking status, physical activity, BMI, regular antibiotic use, MHT use, tamoxifen use, age at first birth, menopausal status at baseline. A BMI × menopausal status at baseline interaction term was also included in multivariable models since the association between BMI and breast cancer changes after menopause (26). A joint Cox proportional hazards regression model was used to simultaneously estimate the HR for invasive breast cancer and DCIS, and the heterogeneity in risk for invasive breast cancer versus DCIS was estimated (27). This approach uses distinct baseline hazards for the subtypes, treating them as competing outcomes. Because DCIS and invasive breast cancer are not necessarily independent events, case-only logistic regression analysis that adjusted for age at diagnosis and menopausal status at diagnosis was also performed to test for etiologic heterogeneity in risk for invasive breast cancer versus DCIS (28).

We additionally conducted stratified analyses by menopausal status because alterations in endogenous sex hormone levels with menopause are thought to be related to periodontal disease as well as breast cancer (29, 30). When estimating the associations stratified by menopausal status, menopausal status was considered as a time-varying factor, so that the same woman could contribute person-time, for example, to both the premenopause and the postmenopause strata. Given that there is considerable heterogeneity in risk factor associations by tumor estrogen receptor (ER) status (31, 32), analyses evaluating risk by ER status were also conducted. Smoking is one of the major modifiable risk factors for periodontal disease, and may be weakly associated with risk for breast cancer (33–35). Smoking status may also be associated with other health behaviors, like cancer screening (36). Thus, we stratified analyses by smoking status at baseline.

The proportional hazards assumption was tested using an age times exposure interaction term, and the assumption did not appear to be violated. In multivariable models we additionally tested the interaction of periodontal disease with both menopausal status and smoking status with respect to breast cancer risk. A sensitivity analysis was conducted, restricting to women who reported both having seen the dentist and having had a mammogram within the 12 months prior to baseline (N = 35,362, 71%). We also conducted sensitivity analyses further adjusting for alcohol drinking status, recency of mammogram, age at menarche, and parity. All analyses were performed in SAS 9.4 software (SAS Institute Inc.).

The mean age at enrollment was 56 years (SD = 9), and 22% of participants reported having ever been diagnosed with periodontal disease (Table 1). Women with periodontal disease tended to be older than those without (mean ages 58 and 55 years, respectively). Women with periodontal disease were more likely to be postmenopausal and slightly more likely to have seen a dentist or had a mammogram within the 12 months prior to baseline. Women with periodontal disease were more likely to be heavier, smokers, and MHT users, and were less likely to be non-Hispanic white, physically active, and to have ever used tamoxifen (Table 1). Multivariable associations between baseline participant characteristics and periodontal disease are shown in Supplementary Table S1.

Table 1.

Baseline characteristics of participants in the Sister Study, United States, 2003–2009.a

All participants (n = 49,968)Women with history of periodontal disease (n = 10,830)Women without history of periodontal disease (n = 39,138)
Mean (SD)No. of womenFrequency (%)Mean (SD)No. of womenFrequency (%)Mean (SD)No. of womenFrequency (%)
Baseline age (years) 55.6 (9.0)   57.9 (8.3)   54.9 (9.0)   
Follow-up time (years) 9.3 (2.4)   9.2 (2.5)   9.3 (2.4)   
BMIb (kg/m227.8 (6.3)   28.3 (6.4)   27.7 (6.2)   
Education 
 High school or less 7,643 15.3  1,680 15.5  5963 15.2 
 Some college or associate degree 16,881 33.8  3,780 34.9  13,101 33.5 
 Bachelor's degree 13,494 27.0  2,687 24.8  10,807 27.6 
 Master's or doctoral degree 11,950 23.9  2,683 24.8  9,267 23.7 
Race/ethnicity 
 Non-Hispanic White 41,852 83.8  8,700 80.3  33,152 84.7 
 Non-Hispanic Black 4,371 8.8  1,298 12.0  3,073 7.9 
 Hispanic 2,444 4.9  527 4.9  1,917 4.9 
 Other 1,301 2.6  305 2.8  996 2.5 
Smoking status 
 Never smoker 28,057 56.2  4,685 43.3  23,372 59.7 
 Former smoker 17,817 35.7  4,786 44.2  13,031 33.3 
 Current smoker 4,094 8.2  1,359 12.6  2,735 7.0 
Physical activity (MET-hours per week) 
 ≤27.06 (Q1) 12,489 25.0  2,972 27.4  9,517 24.3 
 27.06–44.37 (Q2) 12,499 25.0  2,709 25.0  9,790 25.0 
 44.38–67.21 (Q3) 12,488 25.0  2,639 24.4  9,849 25.2 
 >67.21 (Q4) 12,492 25.0  2,510 23.2  9,982 25.5 
Had been to a dentist 12 months prior to baseline 
 No 7,164 14.3  1,472 13.6  5,692 14.6 
 Yes 42,796 85.7  9,356 86.4  33,440 85.5 
Had a mammogram 12 months prior to baseline 
 No 9,186 18.6  7,097 18.4  2,089 19.5 
 Yes 40,218 81.4  31,572 81.7  8,646 80.5 
Menopause status at baseline 
 Premenopausal 16,887 33.8  2,509 23.2  14,378 36.7 
 Postmenopausal 33,081 66.2  8,321 76.8  24,760 63.3 
Age at first birth 
 Nulliparity 9,055 18.1  2,023 18.7  7,032 18.0 
 <20 6,629 13.3  1,586 14.6  5,043 12.9 
 20–29 26,979 54.0  5,802 53.6  21,177 54.1 
 ≥30 7,305 14.6  1,419 13.1  5,886 15.0 
Take antibiotics regularly 
 Never user or missing 43,621 87.3  9,565 88.3  34,056 87.0 
 Former user 5,159 10.3  987 9.1  4,172 10.7 
 Current user 1,188 2.4  278 2.6  910 2.3 
MHT use 
 Never user 27,429 54.9  5,222 48.2  22,207 56.7 
 EPT user (with/without ET) 11,303 22.6  2,966 27.4  8,337 21.3 
 ET user 9,889 19.8  2,364 21.8  7,525 19.2 
 Other user 1,347 2.7  278 2.6  1,069 2.7 
Tamoxifen use 
 Never user 48,951 98.0  10,634 98.2  38,317 97.9 
 Ever user 1,017 2.0  196 1.8  821 2.1 
All participants (n = 49,968)Women with history of periodontal disease (n = 10,830)Women without history of periodontal disease (n = 39,138)
Mean (SD)No. of womenFrequency (%)Mean (SD)No. of womenFrequency (%)Mean (SD)No. of womenFrequency (%)
Baseline age (years) 55.6 (9.0)   57.9 (8.3)   54.9 (9.0)   
Follow-up time (years) 9.3 (2.4)   9.2 (2.5)   9.3 (2.4)   
BMIb (kg/m227.8 (6.3)   28.3 (6.4)   27.7 (6.2)   
Education 
 High school or less 7,643 15.3  1,680 15.5  5963 15.2 
 Some college or associate degree 16,881 33.8  3,780 34.9  13,101 33.5 
 Bachelor's degree 13,494 27.0  2,687 24.8  10,807 27.6 
 Master's or doctoral degree 11,950 23.9  2,683 24.8  9,267 23.7 
Race/ethnicity 
 Non-Hispanic White 41,852 83.8  8,700 80.3  33,152 84.7 
 Non-Hispanic Black 4,371 8.8  1,298 12.0  3,073 7.9 
 Hispanic 2,444 4.9  527 4.9  1,917 4.9 
 Other 1,301 2.6  305 2.8  996 2.5 
Smoking status 
 Never smoker 28,057 56.2  4,685 43.3  23,372 59.7 
 Former smoker 17,817 35.7  4,786 44.2  13,031 33.3 
 Current smoker 4,094 8.2  1,359 12.6  2,735 7.0 
Physical activity (MET-hours per week) 
 ≤27.06 (Q1) 12,489 25.0  2,972 27.4  9,517 24.3 
 27.06–44.37 (Q2) 12,499 25.0  2,709 25.0  9,790 25.0 
 44.38–67.21 (Q3) 12,488 25.0  2,639 24.4  9,849 25.2 
 >67.21 (Q4) 12,492 25.0  2,510 23.2  9,982 25.5 
Had been to a dentist 12 months prior to baseline 
 No 7,164 14.3  1,472 13.6  5,692 14.6 
 Yes 42,796 85.7  9,356 86.4  33,440 85.5 
Had a mammogram 12 months prior to baseline 
 No 9,186 18.6  7,097 18.4  2,089 19.5 
 Yes 40,218 81.4  31,572 81.7  8,646 80.5 
Menopause status at baseline 
 Premenopausal 16,887 33.8  2,509 23.2  14,378 36.7 
 Postmenopausal 33,081 66.2  8,321 76.8  24,760 63.3 
Age at first birth 
 Nulliparity 9,055 18.1  2,023 18.7  7,032 18.0 
 <20 6,629 13.3  1,586 14.6  5,043 12.9 
 20–29 26,979 54.0  5,802 53.6  21,177 54.1 
 ≥30 7,305 14.6  1,419 13.1  5,886 15.0 
Take antibiotics regularly 
 Never user or missing 43,621 87.3  9,565 88.3  34,056 87.0 
 Former user 5,159 10.3  987 9.1  4,172 10.7 
 Current user 1,188 2.4  278 2.6  910 2.3 
MHT use 
 Never user 27,429 54.9  5,222 48.2  22,207 56.7 
 EPT user (with/without ET) 11,303 22.6  2,966 27.4  8,337 21.3 
 ET user 9,889 19.8  2,364 21.8  7,525 19.2 
 Other user 1,347 2.7  278 2.6  1,069 2.7 
Tamoxifen use 
 Never user 48,951 98.0  10,634 98.2  38,317 97.9 
 Ever user 1,017 2.0  196 1.8  821 2.1 

Abbreviations: ET, estrogen therapy; EPT, estrogen plus progestin therapy; Q, quartile.

aMissing data: had been to a dentist 12 months prior to baseline, 2 women with periodontal disease, 6 women without periodontal disease; mammogram 12 months prior to baseline, 469 women with periodontal disease, 95 women without periodontal disease.

bBMI calculated as weight (kg)/height (m)2.

A total of 3,339 incident breast cancer cases, including 2,607 invasive breast cancer and 732 DCIS, were identified during a mean (SD) follow-up of 9.25 (2.41) years. In the age-adjusted model, there was little evidence of association between periodontal disease and overall breast cancer risk (HR = 1.04; 95% CI, 0.96–1.13; Table 2). The multivariable-adjusted estimate was similar (HR = 1.02; 95% CI, 0.94–1.11). In multivariable analyses, there was heterogeneity in the association of periodontal disease with risk of invasive breast cancer versus DCIS (relative HR = 1.24; P-het = 0.04), such that periodontal disease was associated with a small nonstatistically significant increased risk for invasive breast cancer (HR = 1.07; 95% CI, 0.97–1.17) and decreased risk for DCIS (HR = 0.86; 95% CI, 0.72–1.04; Table 2). In case-only analysis, the estimated odds ratio (OR) for risk of invasive breast cancer versus DCIS (OR = 1.24; 95% CI, 1.01–1.52) was the same as the relative HR from the joint model (Supplementary Table S2).

Table 2.

Association of periodontal disease with risk of breast cancer, Sister Study, 2003–2017 (N = 49,968).

IBC and DCIS combinedIBC onlyDCIS onlyHeterogeneityc
No. of casesHR (95% CI)No. of casesHR (95% CI)No. of casesHR (95% CI)Relative HR (95% CI)P-het
Model 1a 3,339 1.04 (0.96–1.13) 2,607 1.09 (0.99–1.19) 732 0.88 (0.73–1.05) 1.24 (1.01–1.52) 0.04 
Model 2b 3,339 1.02 (0.94–1.11) 2,607 1.07 (0.97–1.17) 732 0.86 (0.72–1.04) 1.24 (1.01–1.52) 0.04 
IBC and DCIS combinedIBC onlyDCIS onlyHeterogeneityc
No. of casesHR (95% CI)No. of casesHR (95% CI)No. of casesHR (95% CI)Relative HR (95% CI)P-het
Model 1a 3,339 1.04 (0.96–1.13) 2,607 1.09 (0.99–1.19) 732 0.88 (0.73–1.05) 1.24 (1.01–1.52) 0.04 
Model 2b 3,339 1.02 (0.94–1.11) 2,607 1.07 (0.97–1.17) 732 0.86 (0.72–1.04) 1.24 (1.01–1.52) 0.04 

Abbreviation: IBC, invasive breast cancer.

aModel 1 adjusted for age at baseline.

bModel 2 adjusted for age at baseline, education level, race/ethnicity, smoking status, BMI, menopausal status at baseline, BMI × menopause, antibiotics regular use, menopausal hormone therapy use, tamoxifen use, physical activity, age at first birth.

cHeterogeneity in the associations of periodontal disease with risk of IBC versus DCIS.

In analyses stratified by menopausal status, we observed similar patterns of association for premenopausal and postmenopausal person-time. Multivariable HRs for overall breast cancer, invasive breast cancer and DCIS were 1.02 (95% CI, 0.80–1.29), 1.05 (95% CI, 0.80–1.38), and 0.90 (95% CI, 0.55–1.49) among premenopausal women, and 1.02 (95% CI, 0.93–1.11), 1.00 (95% CI, 0.55–1.80), and 0.85 (95% CI, 0.70–1.04) among postmenopausal women, respectively. Statistical heterogeneity in the association of periodontal disease with invasive breast cancer versus DCIS was most apparent in the larger group of postmenopausal women (relative HR = 1.25; 95% CI, 1.00–1.55; Table 3), and the case-only analysis demonstrated similar heterogeneity in results (OR = 1.25; 95% CI, 1.01–1.56). Consistent with the patterns of association that we observed for overall breast cancer risk, statistically significant heterogeneity in the risk associations for invasive breast cancer versus DCIS was also observed for ER-positive breast cancer (relative HR = 1.35; 95% CI, 1.06–1.72; Table 4); the case-only analysis showed similar results (OR = 1.35; 95% CI, 1.05–1.73). In analyses stratified by smoking status, the estimated HRs was strongest among former smokers, with almost statistically significant heterogeneity in risk associations for invasive disease versus DCIS observed among former smokers (relative HR = 1.34; 95% CI, 0.98–1.84; Table 5); in case-only analysis, the estimated OR was 1.35 (95% CI, 0.98–1.85) in former smokers. We did not detect multiplicative interactions between periodontal disease and menopausal status or smoking status with respect to breast cancer risk by either the Cox model or the case-only analyses (all P values for interaction > 0.2).

Table 3.

Association of periodontal disease with risk of breast cancer by menopausal status, Sister Study, 2003–2017.

IBC and DCIS combinedIBC onlyDCIS onlyHeterogeneityd
No. of casescHR (95% CI)No. of casesHR (95% CI)No. of casesHR (95% CI)Relative HR (95% CI)P-het
Premenopausal women 
 Model 1a 594 0.98 (0.77–1.24) 451 1.01 (0.77–1.32) 143 0.87 (0.53–1.43) 1.17 (0.67–2.05) 0.59 
 Model 2b 594 1.02 (0.80–1.29) 451 1.05 (0.80–1.38) 143 0.90 (0.55–1.49) 1.17 (0.67–2.05) 0.59 
Postmenopausal women 
 Model 1 2,745 1.05 (0.96–1.15) 2,156 1.03 (0.57–1.86) 589 0.88 (0.73–1.07) 1.25 (1.00–1.55) 0.046 
 Model 2 2,745 1.02 (0.93–1.11) 2,156 1.00 (0.55–1.80) 589 0.85 (0.70–1.04) 1.25 (1.00–1.55) 0.046 
IBC and DCIS combinedIBC onlyDCIS onlyHeterogeneityd
No. of casescHR (95% CI)No. of casesHR (95% CI)No. of casesHR (95% CI)Relative HR (95% CI)P-het
Premenopausal women 
 Model 1a 594 0.98 (0.77–1.24) 451 1.01 (0.77–1.32) 143 0.87 (0.53–1.43) 1.17 (0.67–2.05) 0.59 
 Model 2b 594 1.02 (0.80–1.29) 451 1.05 (0.80–1.38) 143 0.90 (0.55–1.49) 1.17 (0.67–2.05) 0.59 
Postmenopausal women 
 Model 1 2,745 1.05 (0.96–1.15) 2,156 1.03 (0.57–1.86) 589 0.88 (0.73–1.07) 1.25 (1.00–1.55) 0.046 
 Model 2 2,745 1.02 (0.93–1.11) 2,156 1.00 (0.55–1.80) 589 0.85 (0.70–1.04) 1.25 (1.00–1.55) 0.046 

Abbreviation: IBC, invasive breast cancer.

aModel 1 adjusted for age at baseline.

bModel 2 adjusted for age at baseline, education level, race/ethnicity, smoking status, BMI, menopausal status as a time-varying variable, antibiotics regular use, menopausal hormone therapy use, tamoxifen use, physical activity, age at first birth.

cNumber of cases based on the menopausal status at diagnosis within 91,970 premenopausal person-years, and 370,246 postmenopausal person-years.

dHeterogeneity in the associations of periodontal disease with risk of IBC versus DCIS.

Table 4.

Association of periodontal disease with risk of breast cancer by estrogen receptor status, Sister Study, 2003–2017.

IBC and DCIS combinedIBC onlyDCIS onlyHeterogeneityc
No. of casesHR (95% CI)No. of casesHR (95% CI)No. of casesHR (95% CI)Relative HR (95% CI)P-het
ER-positive 
 Model 1a 2461 1.03 (0.94–1.13) 1937 1.09 (0.99–1.22) 524 0.80 (0.64–1.00) 1.35 (1.06–1.73) 0.02 
 Model 2b 2461 1.02 (0.93–1.23) 1937 1.09 (0.98–1.21) 524 0.80 (0.64–1.00) 1.35 (1.06–1.72) 0.02 
ER-negative 
 Model 1 436 0.95 (0.76–1.20) 334 0.94 (0.72–1.23) 102 0.99 (0.62–1.58) 0.95 (0.55–1.63) 0.84 
 Model 2 436 0.95 (0.75–1.20) 334 0.94 (0.71–1.23) 102 0.98 (0.61–1.58) 0.95 (0.55–1.63) 0.84 
IBC and DCIS combinedIBC onlyDCIS onlyHeterogeneityc
No. of casesHR (95% CI)No. of casesHR (95% CI)No. of casesHR (95% CI)Relative HR (95% CI)P-het
ER-positive 
 Model 1a 2461 1.03 (0.94–1.13) 1937 1.09 (0.99–1.22) 524 0.80 (0.64–1.00) 1.35 (1.06–1.73) 0.02 
 Model 2b 2461 1.02 (0.93–1.23) 1937 1.09 (0.98–1.21) 524 0.80 (0.64–1.00) 1.35 (1.06–1.72) 0.02 
ER-negative 
 Model 1 436 0.95 (0.76–1.20) 334 0.94 (0.72–1.23) 102 0.99 (0.62–1.58) 0.95 (0.55–1.63) 0.84 
 Model 2 436 0.95 (0.75–1.20) 334 0.94 (0.71–1.23) 102 0.98 (0.61–1.58) 0.95 (0.55–1.63) 0.84 

Abbreviation: IBC, invasive breast cancer

aModel 1 adjusted for age at baseline.

bModel 2 adjusted for age at baseline, education level, race/ethnicity, BMI, menopausal status at baseline, BMI × menopause, antibiotics regular use, menopausal hormone therapy use, tamoxifen use, physical activity, age at first birth.

cHeterogeneity in the associations of periodontal disease with risk of IBC versus DCIS.

Table 5.

Association of periodontal disease with risk of breast cancer by smoking status, Sister Study, 2003–2017.

IBC and DCIS combinedIBC onlyDCIS onlyHeterogeneityc
No. of participantsNo. of casesHR (95% CI)No. of casesHR (95% CI)No. of casesHR (95% CI)Relative HR (95% CI)P-het
Never smoker 
 Model 1a 28,057 1,794 1.02 (0.90–1.15) 1,383 1.06 (0.92–1.21) 411 0.90 (0.69–1.18) 1.17 (0.87–1.58) 0.31 
 Model 2b 28,057 1,794 1.01 (0.89–1.15) 1,383 1.05 (0.91–1.20) 411 0.90 (0.68–1.17) 1.17 (0.87–1.58) 0.31 
Ever smoker 
 Model 1 21,911 1,545 1.03 (0.92–1.15) 1,224 1.08 (0.96–1.22) 321 0.86 (0.66–1.10) 1.26 (0.95–1.67) 0.10 
 Model 2 21,911 1,545 1.03 (0.92–1.15) 1,224 1.07 (0.95–1.21) 321 0.85 (0.66–1.10) 1.26 (0.95–1.67) 0.11 
Former smokers 
 Model 1 17,817 1,302 1.05 (0.93–1.18) 1,030 1.11 (0.97–1.27) 272 0.82 (0.62–1.09) 1.34 (0.98–1.84) 0.07 
 Model 2 17,817 1,302 1.03 (0.91–1.17) 1,030 1.09 (0.95–1.25) 272 0.82 (0.61–1.08) 1.34 (0.98–1.84) 0.07 
Current smokers 
 Model 1 4,094 243 0.99 (0.76–1.30) 194 0.98 (0.73–1.32) 49 1.06 (0.58–1.91) 0.93 (0.48–1.80) 0.82 
 Model 2 4,094 243 0.98 (0.75–1.28) 194 0.96 (0.71–1.30) 49 1.04 (0.57–1.88) 0.93 (0.48–1.80) 0.82 
IBC and DCIS combinedIBC onlyDCIS onlyHeterogeneityc
No. of participantsNo. of casesHR (95% CI)No. of casesHR (95% CI)No. of casesHR (95% CI)Relative HR (95% CI)P-het
Never smoker 
 Model 1a 28,057 1,794 1.02 (0.90–1.15) 1,383 1.06 (0.92–1.21) 411 0.90 (0.69–1.18) 1.17 (0.87–1.58) 0.31 
 Model 2b 28,057 1,794 1.01 (0.89–1.15) 1,383 1.05 (0.91–1.20) 411 0.90 (0.68–1.17) 1.17 (0.87–1.58) 0.31 
Ever smoker 
 Model 1 21,911 1,545 1.03 (0.92–1.15) 1,224 1.08 (0.96–1.22) 321 0.86 (0.66–1.10) 1.26 (0.95–1.67) 0.10 
 Model 2 21,911 1,545 1.03 (0.92–1.15) 1,224 1.07 (0.95–1.21) 321 0.85 (0.66–1.10) 1.26 (0.95–1.67) 0.11 
Former smokers 
 Model 1 17,817 1,302 1.05 (0.93–1.18) 1,030 1.11 (0.97–1.27) 272 0.82 (0.62–1.09) 1.34 (0.98–1.84) 0.07 
 Model 2 17,817 1,302 1.03 (0.91–1.17) 1,030 1.09 (0.95–1.25) 272 0.82 (0.61–1.08) 1.34 (0.98–1.84) 0.07 
Current smokers 
 Model 1 4,094 243 0.99 (0.76–1.30) 194 0.98 (0.73–1.32) 49 1.06 (0.58–1.91) 0.93 (0.48–1.80) 0.82 
 Model 2 4,094 243 0.98 (0.75–1.28) 194 0.96 (0.71–1.30) 49 1.04 (0.57–1.88) 0.93 (0.48–1.80) 0.82 

Abbreviation: IBC, invasive breast cancer.

aModel 1 adjusted for age at baseline.

bModel 2 adjusted for age at baseline, education level, race/ethnicity, BMI, menopausal status at baseline, BMI × menopause, antibiotics regular use, menopausal hormone therapy use, tamoxifen use, physical activity, age at first birth.

cHeterogeneity in the associations of periodontal disease with risk of IBC versus DCIS.

In sensitivity analyses restricted to participants who reported both having seen a dentist and having had a mammogram within the 12 months before study enrollment, and in analyses further adjusting for additional potential confounding factors, findings were consistent with those observed among the overall cohort (Supplementary Tables S3 and S4).

In this large, prospective cohort study, we did not observe a clear association between self-reported periodontal disease and overall breast cancer risk. However, we observed suggestive heterogeneity in risk for invasive breast cancer versus DCIS, with periodontal disease being associated with a nonstatistically significant higher incidence of invasive disease and lower incidence of DCIS. In addition, these findings persisted under adjustment for multiple potential confounders and in sensitivity analyses restricted to those with recent dental exams and mammograms. Overall, in one of the largest studies to date to examine the association between periodontal disease and breast cancer risk, our findings suggest that periodontal disease is not strongly associated with breast cancer risk. However, the observed heterogeneity in periodontal disease risk associations for invasive breast cancer versus DCIS warrants further exploration.

Our decisively null findings for the association between periodontal disease and overall breast cancer risk are consistent with several previous studies that incorporated detailed exposure assessment. In the Atherosclerosis Risk in Communities (ARIC) cohort study, which had ascertained clinical records of periodontal disease, increasing severity of periodontal disease was associated with overall cancer risk but not specifically with breast cancer (4). In the Buffalo OsteoPerio Study, a prospective study where oral alveolar crestal bone height was measured, as a biomarker, to examine the association between periodontal disease and invasive breast cancer risk, no association was detected, but the number of breast cancer cases was small (N = 9; ref. 18). In contrast, a retrospective cohort study within the Taiwan National Health Insurance Program, where participants with chronic periodontitis were ascertained on the basis of medical insurance claims, found increased rates of overall cancer, including higher rates of breast cancer specifically, with a hazard ratio of 1.23 (95% CI, 1.11–1.36); however, the cancer cases in this study were sourced from an administrative database, with the authors noting possible inaccuracy in diagnosis confirmation and potential survival bias (6). Results from the WHI OS, which identified 2,124 postmenopausal invasive breast cancer cases after a mean follow-up of 6.7 years, found that periodontal disease was related to a borderline increased risk of invasive breast cancer among postmenopausal women, with an estimated HR of 1.11 (95% CI, 1.00–1.23; ref. 15). Differences in participants (both pre- and postmenopausal women or postmenopausal women only), methods of exposure ascertainment (self-reported with or without clinical dental measurements), severity of periodontal disease (not specified, mild, moderate or severe), as well as adjustment factors (differences in potential confounding factors) may explain part of the discrepancies observed across these observational studies. In the two recent meta-analyses (19, 20), modest positive associations between periodontal disease and breast cancer risk were observed, with estimated RRs of 1.18 (95% CI, 1.11–1.26) and 1.22 (95% CI, 1.06–1.40), and the WHI OS contributing the most to the overall RRs of any prior study (weights of 25.6% and 52.9%, respectively).

This is the first analysis to consider the association of periodontal disease with breast cancer risk separately for invasive breast cancer and DCIS, and a number of factors could contribute to our observed differences in associations for the invasive versus in situ lesions. First, prior work has suggested etiologic heterogeneity for DCIS as compared with invasive breast cancer for several breast cancer risk factors (22, 37). Although biologic differences in the oral microbiota of women diagnosed with DCIS versus invasive disease have not been previously explored, numerous taxa have been found to be significantly enriched in healthy oral samples as compared with those from breast cancer cases studied prior to any systemic therapy, and these taxonomic differences corresponded to functional differences (38). Further investigation is needed to evaluate whether periodontal disease or the oral microbiota differentially affect DCIS and invasive breast cancer. Alternatively, the suggestive association of increased risk of invasive breast cancer associated with periodontal disease could reflect confounding by surveillance behavior if women who tend to have periodontal disease (and a history of poor dental hygiene) also tend to get diagnosed at a later stage of tumor progression. In such a scenario, some of the tumors that would have been DCIS are not found until they are invasive, creating an artifactual “protection” against DCIS. We therefore conducted sensitivity analyses restricted to the participants who had reported having had both a dental appointment and a mammogram within the 12 months before baseline, and the findings were hardly altered. Although it is possible that unmeasured confounders may have influenced our findings, risk estimates were very similar for our two multivariable models as well as in sensitivity analyses, suggesting little evidence for confounding by the measured covariates.

The associations between periodontal disease and breast cancer for premenopausal and postmenopausal women were similar, consistent with the absence of an interaction between periodontal disease and menopausal status with respect to breast cancer risk. Among postmenopausal participants in the WHI OS, the highest HR for invasive breast cancer was observed among former smokers, especially for those who quit in the past 20 years (HR = 1.36; 95% CI, 1.05–1.77; ref. 15). The estimated association with invasive breast cancer was also most apparent among former smokers, although not statistically significant, in our study (HR = 1.09; 95% CI, 0.95–1.25). These results may reflect a potential role of former smoking exposure in breast cancer risk, and persisting effects on the oral microbiome (15, 39). There were too few current smokers in our analysis to disentangle the effects of ever versus former smoking. However, alternative explanations including differential screening behavior by smoking status may also contribute to the apparent heterogeneity in risk for invasive disease versus DCIS among former smokers. Indeed, in a prior study by Sanford and colleagues, compared with never smokers, former smokers were more likely to undergo mammographic screening, while current smokers were less likely to receive screening (36). In our study, we observed a similar trend for recent breast cancer screening. Although our observed findings remained robust to a series of adjustments by potential confounding factors, we cannot rule out the possibility of chance findings. Future studies with more extended follow-up periods may permit evaluation of the effects of latency on risk associated with changes in periodontal disease over the course of study follow up. On the other hand, more frequent screening per se would be expected to favor DCIS over invasive disease, and thus would not explain our findings.

Strengths of our study should be highlighted. First, this is a prospective cohort study with a large population size and long follow-up time. This population has been well-characterized, which enabled us to control for multiple potential confounders. Second, because all participants had a sister with breast cancer, the women in this study had a higher risk, on average, than that in the general population, allowing us to accrue a large number of breast cancer cases in a short amount of time. Third, women in this volunteer cohort are highly compliant, with more than 90% responding to their most recent follow-up request. Fourth, this is the first study to assess the risk association with periodontal disease separately for invasive breast cancer and DCIS, which suggests future avenues of research will need to allow for potential etiologic heterogeneity for these subtypes of breast cancer.

Limitations in our study should also be noted. First, self-reported periodontal disease history was used, and it may have been under-reported. Such misclassification may lead to the underestimation of the association between periodontal disease and breast cancer risk. However, self-reported periodontal disease has been shown to have acceptable validity for use in large epidemiologic studies (40). Second, we did not have data regarding the frequency of breast cancer screening during follow-up, and were unable to address whether periodontal disease risk associations varied by screening behavior. However, as participants in the Sister Study are at elevated risk and the majority reported having had a mammogram within the 12 months prior to baseline, presumably they continued to undergo breast cancer screening during follow-up (41, 42). Third, DCIS and invasive breast cancer are not necessarily independent events, suggesting that the assumptions of the competing risks approach we used may not apply. However, we also considered case-only analyses to test for etiologic heterogeneity, and found similar results.

In conclusion, our findings indicate no clear association between periodontal disease and breast cancer. Additional studies are warranted to explain potential heterogeneity in the risk for invasive breast cancer versus DCIS, given that we observed evidence of non-statistically significant elevated risk for invasive breast cancer and reduced risk for DCIS associated with history of periodontal disease.

No potential conflicts of interest were disclosed.

M. Jia: Formal analysis, writing-original draft, writing-review and editing. Z. Wu: Formal analysis, writing-review and editing. E. Vogtmann: Conceptualization, formal analysis, supervision, writing-original draft, writing-review and editing. K.M. O'Brien: Formal analysis, writing-original draft, writing-review and editing. C.R. Weinberg: Resources, formal analysis, writing-review and editing. D.P. Sandler: Resources, writing-original draft, writing-review and editing. G.L. Gierach: Conceptualization, formal analysis, supervision, writing-original draft, writing-review and editing.

This work was supported in part by the Intramural Research Program of the NIH, NCI (project ZIA CP010126-17 to G.L. Gierach) and National Institute of Environmental Health Sciences (project Z01-ES044005 to D.P. Sandler), and Chinese Academy of Medical Sciences Innovation Fund for Medical Sciences (CIFMS 2017-12M-B&R-03 to M. Jia).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Kinane
DF
,
Stathopoulou
PG
,
Papapanou
PN
. 
Periodontal diseases
.
Nat Rev Dis Primers
2017
;
3
:
17038
.
2.
Eke
PI
,
Thornton-Evans
GO
,
Wei
L
,
Borgnakke
WS
,
Dye
BA
,
Genco
RJ
. 
Periodontitis in US adults: National Health and Nutrition Examination Survey 2009–2014
.
J Am Dent Assoc
2018
;
149
:
576
88
.
3.
Michaud
DS
,
Fu
Z
,
Shi
J
,
Chung
M
. 
Periodontal disease, tooth loss, and cancer risk
.
Epidemiol Rev
2017
;
39
:
49
58
.
4.
Michaud
DS
,
Lu
J
,
Peacock-Villada
AY
,
Barber
JR
,
Joshu
CE
,
Prizment
AE
, et al
Periodontal disease assessed using clinical dental measurements and cancer risk in the ARIC Study
.
J Natl Cancer Inst
2018
;
110
:
843
54
.
5.
Meyer
MS
,
Joshipura
K
,
Giovannucci
E
,
Michaud
DS
. 
A review of the relationship between tooth loss, periodontal disease, and cancer
.
Cancer Causes Control
2008
;
19
:
895
907
.
6.
Chung
SD
,
Tsai
MC
,
Huang
CC
,
Kao
LT
,
Chen
CH
. 
A population-based study on the associations between chronic periodontitis and the risk of cancer
.
Int J Clin Oncol
2016
;
21
:
219
23
.
7.
Mai
X
,
LaMonte
MJ
,
Hovey
KM
,
Nwizu
N
,
Freudenheim
JL
,
Tezal
M
, et al
History of periodontal disease diagnosis and lung cancer incidence in the Women's Health Initiative Observational Study
.
Cancer Causes Control
2014
;
25
:
1045
53
.
8.
Dizdar
O
,
Hayran
M
,
Guven
DC
,
Yilmaz
TB
,
Taheri
S
,
Akman
AC
, et al
Increased cancer risk in patients with periodontitis
.
Curr Med Res Opin
2017
;
33
:
2195
200
.
9.
Jiang
X
,
Shapiro
DJ
. 
The immune system and inflammation in breast cancer
.
Mol Cell Endocrinol
2014
;
382
:
673
82
.
10.
Xuan
C
,
Shamonki
JM
,
Chung
A
,
Dinome
ML
,
Chung
M
,
Sieling
PA
, et al
Microbial dysbiosis is associated with human breast cancer
.
PLoS One
2014
;
9
:
e83744
.
11.
Thompson
KJ
,
Ingle
JN
,
Tang
X
,
Chia
N
,
Jeraldo
PR
,
Walther-Antonio
MR
, et al
A comprehensive analysis of breast cancer microbiota and host gene expression
.
PLoS One
2017
;
12
:
e0188873
.
12.
Urbaniak
C
,
Cummins
J
,
Brackstone
M
,
Macklaim
JM
,
Gloor
GB
,
Baban
CK
, et al
Microbiota of human breast tissue
.
Appl Environ Microbiol
2014
;
80
:
3007
14
.
13.
Chan
DS
,
Bandera
EV
,
Greenwood
DC
,
Norat
T
. 
Circulating C-reactive protein and breast cancer risk-systematic literature review and meta-analysis of prospective cohort studies
.
Cancer Epidemiol Biomarkers Prev
2015
;
24
:
1439
49
.
14.
DeSantis
CE
,
Ma
J
,
Goding Sauer
A
,
Newman
LA
,
Jemal
A
. 
Breast cancer statistics, 2017, racial disparity in mortality by state
.
CA Cancer J Clin
2017
;
67
:
439
48
.
15.
Freudenheim
JL
,
Genco
RJ
,
LaMonte
MJ
,
Millen
AE
,
Hovey
KM
,
Mai
X
, et al
Periodontal disease and breast cancer: prospective cohort study of postmenopausal women
.
Cancer Epidemiol Biomarkers Prev
2016
;
25
:
43
50
.
16.
Soder
B
,
Yakob
M
,
Meurman
JH
,
Andersson
LC
,
Klinge
B
,
Soder
PO
. 
Periodontal disease may associate with breast cancer
.
Breast Cancer Res Treat
2011
;
127
:
497
502
.
17.
Hujoel
PP
,
Drangsholt
M
,
Spiekerman
C
,
Weiss
NS
. 
An exploration of the periodontitis—cancer association
.
Ann Epidemiol
2003
;
13
:
312
6
.
18.
Mai
X
,
LaMonte
MJ
,
Hovey
KM
,
Freudenheim
JL
,
Andrews
CA
,
Genco
RJ
, et al
Periodontal disease severity and cancer risk in postmenopausal women: the Buffalo OsteoPerio Study
.
Cancer Causes Control
2016
;
27
:
217
28
.
19.
Shi
T
,
Min
M
,
Sun
C
,
Zhang
Y
,
Liang
M
,
Sun
Y
. 
Periodontal disease and susceptibility to breast cancer: a meta-analysis of observational studies
.
J Clin Periodontol
2018
;
45
:
1025
33
.
20.
Shao
J
,
Wu
L
,
Leng
WD
,
Fang
C
,
Zhu
YJ
,
Jin
YH
, et al
Periodontal disease and breast cancer: a meta-analysis of 173,162 participants
.
Front Oncol
2018
;
8
:
601
.
21.
Burstein
HJ
,
KP
,
Wong
JS
,
Lester
SC
,
Kaelin
CM
. 
Ductal carcinoma in situ of the breast
.
N Engl J Med
2004
;
350
:
1430
41
.
22.
Ruszczyk
M
,
Zirpoli
G
,
Kumar
S
,
Bandera
EV
,
Bovbjerg
DH
,
Jandorf
L
, et al
Breast cancer risk factor associations differ for pure versus invasive carcinoma with an in situ component in case-control and case-case analyses
.
Cancer Causes Control
2016
;
27
:
183
98
.
23.
Sprague
BL
,
Gangnon
RE
,
Hampton
JM
,
Egan
KM
,
Titus
LJ
,
Kerlikowske
K
, et al
Variation in breast cancer-risk factor associations by method of detection: results from a series of case-control studies
.
Am J Epidemiol
2015
;
181
:
956
69
.
24.
Sandler
DP
,
Hodgson
ME
,
Deming-Halverson
SL
,
Juras
PS
,
D'Aloisio
AA
,
Suarez
LM
, et al
The Sister Study Cohort: baseline methods and participant characteristics
.
Environ Health Perspect
2017
;
125
:
127003
.
25.
Sister Study First DCIS or Invasive Breast Cancer Events and Characteristics: Data Release 7.0
; April 01, 
2019.
26.
Chen
Y
,
Liu
L
,
Zhou
Q
,
Imam
MU
,
Cai
J
,
Wang
Y
, et al
Body mass index had different effects on premenopausal and postmenopausal breast cancer risks: a dose-response meta-analysis with 3,318,796 subjects from 31 cohort studies
.
BMC Public Health
2017
;
17
:
936
.
27.
Xue
X
,
Kim
MY
,
Gaudet
MM
,
Park
Y
,
Heo
M
,
Hollenbeck
AR
, et al
A comparison of the polytomous logistic regression and joint cox proportional hazards models for evaluating multiple disease subtypes in prospective cohort studies
.
Cancer Epidemiol Biomarkers Prev
2013
;
22
:
275
85
.
28.
Begg
CB
,
Zhang
ZF
. 
Statistical analysis of molecular epidemiology studies employing case-series
.
Cancer Epidemiol Biomarkers Prev
1994
;
3
:
173
5
.
29.
Paulo Mascarenhas
RG
,
Al-Shammari
K
,
Wang
H-L
. 
Influence of sex hormones on the periodontium
.
J Clin Periodontol
2003
;
30
:
671
81
.
30.
Collaborative Group on Hormonal Factors in Breast Cancer
. 
Menarche, menopause, and breast cancer risk: individual participant meta-analysis, including 118 964 women with breast cancer from 117 epidemiological studies
.
Lancet Oncol
2012
;
13
:
1141
51
.
31.
Schoemaker
MJ
,
Nichols
HB
,
Wright
LB
,
Brook
MN
,
Jones
ME
,
O'Brien
KM
, et al
Adult weight change and premenopausal breast cancer risk: a prospective pooled analysis of data from 628,463 women
.
Int J Cancer
2020
;
147
:
1306
14
.
32.
Figueroa
JD
,
Davis Lynn
BC
,
Edusei
L
,
Titiloye
N
,
Adjei
E
,
Clegg-Lamptey
J-N
, et al
Reproductive factors and risk of breast cancer by tumor subtypes among Ghanaian women: a population-based case–control study
.
Int J Cancer
2020
;
147
:
1535
47
.
33.
Jerome Haber
JW
,
Crowley
M
,
Mandell
R
,
Joshipura
K
,
Kent
RL
. 
Evidence for cigarette smoking as a major risk factor for periodontitis
.
J Periodontol
1993
;
64
:
16
23
.
34.
Scott
L
,
Tomar
SA
. 
Smoking-attributable periodontitis in the United States: findings from NHANES III
.
J Periodontol
2000
;
71
:
743
51
.
35.
Gaudet
MM
,
Gapstur
SM
,
Sun
J
,
Diver
WR
,
Hannan
LM
,
Thun
MJ
. 
Active smoking and breast cancer risk: original cohort data and meta-analysis
.
J Natl Cancer Inst
2013
;
105
:
515
25
.
36.
Sanford
NN
,
Sher
DJ
,
Butler
S
,
Xu
X
,
Ahn
C
,
D'Amico
AV
, et al
Cancer screening patterns among current, former, and never smokers in the United States, 2010–2015
.
JAMA Netw Open
2019
;
2
:
e193759
.
37.
Mullooly
M
,
Khodr
ZG
,
Dallal
CM
,
Nyante
SJ
,
Sherman
ME
,
Falk
R
, et al
Epidemiologic risk factors for in situ and invasive breast cancers among postmenopausal women in the National Institutes of Health-AARP Diet and Health Study
.
Am J Epidemiol
2017
;
186
:
1329
40
.
38.
Campbell
MJ
,
McCune
E
,
Johnson
B
,
O'Meara
T
,
Heditsian
D
,
Brain
S
, et al
Breast cancer and the human oral and gut microbiomes [abstract]
.
In:
Proceedings of the Annual Meeting of the American Association for Cancer Research
; March 29–April 3, 
2019
;
Atlanta, GA
.
Abstract nr GA2019
.
39.
Bizzarro
S
,
Loos
BG
,
Laine
ML
,
Crielaard
W
,
Zaura
E
. 
Subgingival microbiome in smokers and non-smokers in periodontitis: an exploratory study using traditional targeted techniques and a next-generation sequencing
.
J Clin Periodontol
2013
;
40
:
483
92
.
40.
Abbood
HM
,
Hinz
J
,
Cherukara
G
,
Macfarlane
TV
. 
Validity of self-reported periodontal disease: a systematic review and meta-analysis
.
J Periodontol
2016
;
87
:
1474
83
.
41.
Tsunematsu
M
,
Kawasaki
H
,
Masuoka
Y
,
Kakehashi
M
. 
Factors affecting breast cancer screening behavior in Japan–assessment using the health belief model and conjoint analysis
.
Asian Pac J Cancer Prev
2013
;
14
:
6041
8
.
42.
Campitelli
MA
,
Chiarelli
AM
,
Mirea
L
,
Stewart
L
,
Glendon
G
,
Ritvo
P
, et al
Adherence to breast and ovarian cancer screening recommendations for female relatives from the Ontario site of the Breast Cancer Family Registry
.
Eur J Cancer Prev
2011
;
20
:
492
500
.

Supplementary data