J Clin Periodontol 2004; 31: 574–580 doi: 10.1111/j.1600-051X.2004.00531.

x Copyright r Blackwell Munksgaard 2004

Printed in Denmark. All rights reserved

The relationship between tomato Nelson Wood and Roger B. Johnson

Department of Periodontics, University of
Mississippi School of Dentistry, Jackson, MS,

intake and congestive heart USA

failure risk in periodontitis

subjects
Wood N, Johnson RB: The relationship between tomato intake and congestive heart
failure risk in periodontitis subjects. J Clin Periodontol 2004; 31: 574–580. doi:
10.1111/j.1600-051X.2004.00531.x. r Blackwell Munksgaard, 2004.

Abstract
Background: The objective of this study was to investigate the relationship between
monthly tomato consumption (MTC) and serum lycopene (sLyco) levels, and a self-
reported history of congestive heart failure (CHF) in individuals with periodontitis
using data available in the Third National Health and Nutrition Examination Survey
(NHANES III).
Methods: Adult participants in NHANES III were used in this study. Zero to thirty
three percent of sites with a periodontal attachment loss (PAL) of 43 mm was
considered a healthy periodontium, while greater than 433% of sites with PAL of
43 mm as periodontitis. The outcome variable was the self-reported history of CHF.
MTC and sLyco levels were categorized into quartiles. Data was analyzed by
Kruskal–Wallis, ANOVA and multivariate analyses using SPSSs . po0.05 was used to
reject the null hypothesis.
Results: Individuals with periodontitis showed a dose–response relationship between
dietary MTC and self-reported CHF risk; moderate MTC (risk ratio (RR), 3.15; 95%
confidence interval (CI), 1.03–9.67), low MTC (RR, 3.31; 95% CI, 1.33–8.24)
(po0.05) and very low MTC (RR, 5.10; 95% CI, 1.67–15.57) (po0.01), adjusting for
confounders of both diseases (periodontitis and CHF). The moderate sLyco level-
healthy periodontium group showed a significant decrease in CHF risk (RR, 0.25; 95%
CI, 0.07–0.84) (po0.05), adjusting for confounders. Significant inverse dose–response
relationships were seen between sLyco and C-reactive protein, and MTC and white
blood cell count in periodontitis subjects, respectively (po0.05). MTC was correlated
Key words: carotenoids; congestive heart
with sLyco concentration (r 5
0.018, po0.05), adjusting for confounders abolished
failure; periodontitis; serum lycopene;
that significance. tomatoes
Conclusions: A relationship exists between periodontitis and CHF risk, and high
MTC appears to affect this relationship in a positive direction in periodontitis subjects. Accepted for publication 9 September 2003

Congestive heart failure (CHF) is a occurs as a consequence of other exercise (Forrester et al. 1996, NHLBI
major cause of death in the United cardiac diseases, with coronary artery 1996).
States, affecting 4.8 million Americans disease being the most common etiolo- Lycopene, a carotenoid without pro-
(NHLBI 1996). It is a complex clinical gic factor for left ventricular systolic vitamin-A activity, is one of the major
syndrome characterized by exertional dysfunction. Preventing plaque from carotenoids in Western diets, and is
dyspnea, fatigue and often peripheral developing in coronary arteries could found primarily in tomatoes and tomato
edema resulting from left ventricular have profound effects on the incidence products. It accounts for about 50% of
dysfunction. Specific mechanisms under- of CHF, in addition to dietary lipid carotenoids in human serum. In contrast
lying these symptoms remain poorly reduction, blood pressure control, smok- to other carotenoids, serum lycopene
understood; however, CHF always ing cessation, weight reduction and (sLyco) levels are not usually reduced
 Tomato intake, periodontitis, congestive heart failure 575

by smoking or alcohol consumption but periodontal disease (Loos et al. 2000, was used to increase the likelihood that
are reduced by increased age. Also, Slade et al. 2000, Wu et al. 2000). There attachment loss was the result of disease
reduced sLyco concentration has been are several reports suggesting that sev- and not measurement error. This group-
reported in patients with inflammatory eral periopathogenic bacteria are present ing was consistent with other studies
diseases (Gerster 1997). Epidemiologi- within atheromatous plaque and have reporting NHANES III data (Arbes et
cal studies suggest a cardioprotective direct effects on blood platelet activation al. 1999). The analysis excluded persons
role for carotenoid-rich foods (Chopra and aggregation, suggesting that they who were edentulous.
et al. 2000), with numerous observa- may be factors in the initiation and Another outcome variable was the
tional studies showing that individuals progression of atherosclerosis (Herzberg self-reported history of CHF. The admin-
who ingest more dietary carotenoids et al. 1983, Herzberg & Meyer 1996, istration of food-frequency question-
have a reduced risk of several chronic Haraszthy et al. 2000). naires and a detailed 24 h dietary recall
diseases, including cardiovascular dis- There is not extensive data concern- was used to record food consumption.
eases (Mayne 1996). Among the caro- ing the relationship between sLyco Monthly tomato consumption (MTC)
tenoids, lycopene is commonly present concentrations and the incidence of was calculated and divided into quartiles:
within the serum and other tissues. periodontal disease. The purpose of this 0–3 tomatoes were considered very low;
Serum and tissue lycopene levels were study is to determine whether relation- 43–9 was low; 49–17 was moderate;
inversely related to chronic disease risk ships exist between sLyco concentra- and 417 was high consumption.
(Rao & Agarwal 2000), and exhibited a tions, periodontal disease, and CHF and Non-fasting, venous blood was col-
protective function in the development whether a relationship exists between lected and analyzed for CRP level and
of atherosclerosis (Klipstein-Grobusch low sLyco and CHF risk in periodontal white blood cell (WBC) count (USDHHS
et al. 2000) and myocardial infarction subjects. Since lycopene is derived from 1996c). sLyco (mg/dl) was also analyzed,
(Kohlmeier et al. 1997). tomatoes, we also included tomato calculated and categorized into quartiles:
Recent reports suggest that perio- consumption, in order to study the 0–14 was considered very low; 414–21
dontitis may be a risk factor for cardio- relationship between dietary intake of was low; 421–29 was moderate; and
vascular diseases (Beck et al. 1998, tomatoes, sLyco concentrations, and the 429 was a high concentration.
Arbes et al. 1999, Loos et al. 2000). In incidence of periodontal disease and Data was analyzed using SPSS ver-
addition, acknowledged risk factors for CHF risk using the Third National sion 10.1. Group comparisons were
cardiovascular diseases are also risk Health and Nutrition Examination Sur- made using Kruskal–Wallis, ANOVA,
factors for periodontal disease (Grossi vey (NHANES III) data. multivariate general linear models using
& Genco 1998). Periodontitis is a a Bonferroni adjustment, and multi-
chronic inflammation of the supporting variate logistic regression to calculate
tissues of the teeth and affects 75% of crude odds ratios. Established risk
the adults in the United States (Genco et Materials and Methods factors for periodontal disease and
al. 2002). Bacteria within dental plaque Data for this study was obtained from CHF were selected covariables. The
are a major factor for the initiation and NHANES III, conducted from 1988 to covariables were age, race, gender,
progression of this disease. There is 1994, designed to provide estimates of body mass index (BMI), smoking his-
general agreement that various systemic the health status of the United States’ tory, a self-reported history of diabetes
diseases may be risk factors in the civilian, non-institutionalized population (self-reported by ‘‘Has the doctor ever
etiology of periodontal disease. Increas- aged 2 months and over (Ezzati et al. told you that you have diabetes?’’),
ing evidence exists for a bi-directional 1992). For this analysis, three public use hypertension, socio-economic status
relationship between the etiologies of data files – household adult (USDHHS [poverty income ratio (unimputed in-
both systemic and periodontal diseases 1996a), examination (USDHHS 1996b) come)], education level (years), serum
(Wu et al. 2000, Fowler et al. 2001, and laboratory (USDHHS 1996c) – were CRP and WBC. Furthermore, in order to
Taylor 2001). A causal association obtained from CD-ROM and merged determine if this effect was due to
between systemic diseases and perio- into one data file. Subjects, aged 18 years exogenous lycopene, or lycopene de-
dontitis has not been established; how- and greater, were used in this study. The rived from tomato consumption, we
ever, systemic factors likely modify the independent variable of interest was the controlled for lycopene in the tomato
host response to bacteria and bacterial percent of sites per subject with a multinominal logistic regression model
lipopolysaccharides (LPS), resulting in periodontal attachment loss (PAL) of as well as other serum antioxidants
specific forms and patterns of perio- 43 mm. Periodontal examinations were (serum carotene, serum vitamin E and
dontal diseases (Kinane & Marshall conducted in the mobile examination serum vitamin C). While, in the lyco-
2001). Periopathogenic bacterial LPS, centers by six calibrated examiners pene multinominal logistic regression
initiate a cascade of proinflammatory (Arbes et al. 1999). For this study, extent model, we controlled for tomato intake
cytokines, having both local and sys- scores (Carlos et al. 1986), representing and the other above-mentioned serum
temic effects, including activation of the percent of sites per subject with antioxidants, po0.05 was used to reject
monocytes/macrophages, increased neu- attachment loss of 3 mm or greater, were the null hypothesis.
trophil numbers and plasma fibrinogen calculated and categorized into two
concentrations along with other coagula- groups.
tion factors. Alterations in lipid metabo- Zero to thirty-three percent of sites Results
lism, and enhancement of the synthesis with PAL of 43 mm was considered Baseline findings
of acute phase proteins such as C- normal, while greater than 433% of
reactive protein (CRP) and interleukin- sites with PAL43 mm was defined as Subjects with periodontitis had risk
6 (IL-6) also occur in patients with periodontitis. The threshold of 3 mm factors including demographics, smok-
576 Wood & Johnson

Table 1. Baseline demographics, medical conditions, mean blood chemistry and mean antioxidants
Baseline characteristics No periodontology (n 5 1443) Yes periodontology (n 5 4087) No CHF (n 5 17054) Yes CHF (n 5 640)

demographics mean (SEM) mean (SEM)

age (years) 47.7 (0.6) 47.7 (0.4) 46.7 (0.1) 68.8 (0.5)n
male (%) 52.4 48.6 48.9 46.6
Race (%)
Caucasian 27.2 72.8 95.3 4.7
African-American 27.0 73.0 96.3 3.7
other 23.4 76.6 97.5 2.5
education level (years) 10.76 (0.12) 10.76 (0.07) 10.79 (0.03) 10.72 (0.14)
poverty index 231.9 (5.6) 239.9 (3.6) 240.1 (1.4) 186.1 (6.0)n
CHF risk factors
systolic BP (mmHg) 111.03 (0.28) 116.07 (0.21)n 118.38 (0.18) 117.30 (0.90)
diastolic BP (mmHg) 65.90 (0.29) 70.82 (0.17)n 68.71 (0.12) 67.92 (0.64)
pulse rate (beats/min) 73.46 (0.29) 73.77 (0.18) 75.61 (0.11) 75.24 (0.49)
diabetes history (% yes) 7.9 8.1 7.3 28.4n
body mass index 23.58 (0.17) 23.91 (0.10) 23.54 (0.05) 23.69 (0.25)
waist to hip ratio 0.909 (.002) 0.909 (.001) 0.908 (0.001) 0.909 (0.003)
smoking (packs/day) 1.11 (0.18) 1.23 (0.16) 1.19 (0.05) 1.56 (0.46)
current smoker (% yes) 53.6 50.5 52.0 28.1
Blood chemistry
serum CRP (mg/dl) 0.45 (0.02) 0.46 (0.02) 0.42 (0.01) 0.43 (0.03)
WBC count (  109/l) 7.27 (0.06) 7.42 (0.04) 7.37 (0.02) 7.41 (0.09)
Antioxidants
serum carotene (mg/dl) 478.4 (24.1) 457.4 (13.5) 434.9 (6.1) 422.1 (34.4)
serum tocopherol (mg/dl) 8.52 (0.22) 8.55 (0.15) 8.29 (0.07) 7.84 (0.33)n
serum vitamin C (mg/dl) 0.786 (0.014) 0.769 (0.008) 0.751 (0.004) 0.773 (0.019)
n
po0.05.

ing, medical conditions, inflammatory Table 2. Relationships between lycopene quartiles, serum C-reactive protein (CRP) and white
biomarkers and antioxidant intake. When blood cell (WBC) count (mean  SEM) using a multivariate general linear model
compared with healthy subjects, perio- Lycopene quartiles Biomarker No periodontology Yes periodontology
dontitis subjects had significantly higher
systolic and diastolic blood pressures very low Serum CRP 0.561  0.091 0.582  0.058NS
(po0.05), and had higher BMI and low 0.422  0.082 0.392  0.052n,NS
lower serum vitamin C. Individuals moderate 0.365  0.070 0.339  0.039n,NS
with a history of CHF also had risk high 0.223  0.103 0.453  0.064n,NS
very low WBC count 7.14  0.25 7.34  0.16NS
factors including demographics, smok-
low 7.78  0.29 7.24  0.18NS
ing, medical conditions, inflammatory moderate 7.03  0.33 7.42  0.18NS
markers and antioxidant intake. When high 6.90  0.27 7.34  0.17NS
compared with individuals reporting
no history of CHF, individuals with Multivariate general linear model with a Bonferroni adjustment for gender, race, age, body mass
a history of CHF were significantly index, smoking status, diabetes history, hypertension, socioeconomic status, education level, serum
carotene, vitamin E, vitamin C and tomato consumption.
older, poorer and reported a history n
po0.05. There is only a significant relationship between yes periodontology, serum lycopene
of diabetes (po0.05). CHF subjects
quartiles and CRP.
also had higher BMI and lower NS, there are no significant differences between the yes and no periodontology subjects for CRP or
serum carotene and tocopherol levels WBC at any serum lycopene quartiles.
(Table 1).

and sLyco levels (p 5 0.05). When periodontitis (po0.05). When analyses

Periodontitis and inflammatory
biomarkers
analyses were stratified for sLyco quar- were stratified for MTC quartiles in
tiles in individuals with periodontitis individuals with periodontitis versus
Various dose–response relationships be- versus individuals with healthy perio- individuals with healthy periodontium,
tween sLyco levels in individuals with dontium, no significant differences were no significant differences were seen in
periodontitis and healthy periodontium seen in either CRP or WBC count and either CRP or WBC count and MTC
were explored using a Bonferroni ad- sLyco levels (Table 2). levels (Table 3).
justment of the data (Table 2). When the Various dose–response relationships
analyses were restricted to individuals of different MTC in individuals with Periodontitis and self-reported CHF risk
with periodontitis, a significant relation- periodontitis and healthy periodontium
ship was found between serum CRP were explored using a Bonferroni ad- Multivariate logistic regression for in-
levels and periodontitis (po0.05), justment of the data (Table 3). When the dividuals with periodontitis and indivi-
When analyses were restricted to in- analyses were restricted to individuals duals with healthy periodontium,
dividuals with healthy periodontium, a with periodontitis, a significant relation- stratified for MTC quartiles are shown
strong trend was found for serum CRP ship was found in WBC count and in Table 4. Various dose–response
 Tomato intake, periodontitis, congestive heart failure 577

Table 3. Relationship between monthly tomato consumption (MTC) quartiles and serum C- Ridker et al. 1999, 2000a, b, Liuizzo &
reactive protein (CRP) and white blood cell (WBC) count (mean  SEM) using a multivariate Rizello 2001). In addition, high leuko-
general linear model cyte counts are positively correlated
MTC quartiles Biomarker No periodontology Yes periodontology with cardiovascular diseases (Kannel et
al. 1992, Phillips et al. 1992).
very low serum CRP 0.386  0.084 0.455  0.052NS In the present study, we observed a
low 0.376  0.092 0.488  0.057NS significant association between perio-
moderate 0.397  0.051 0.370  0.034NS dontitis and serum CRP, similar to those
high 0.394  0.107 0.443  0.058NS
reported in other studies (Loos et al.
very low WBC count 7.36  0.26 7.58  0.16NS
low 7.20  0.25 6.97  0.15n,NS 2000). We also observed significant
moderate 7.22  0.30 7.36  0.20NS associations between CRP and sLyco
high 7.38  0.35 7.43  0.19NS levels, and MTC and WBC in indivi-
duals with periodontitis (po0.05).
Multivariate general linear model with a Bonferroni adjustment for gender, race, age, body mass In our statistical analyses, we con-
index, smoking status, diabetes history, hypertension, socioeconomic status, education level, serum
trolled for CRP and WBC count, which
carotene, vitamin E, vitamin C and tomato consumption.
n
po0.05. There is only a significant relationship between yes periodontology, MTC quartiles and
were associated with MTC levels in
WBC. individuals with periodontitis (po0.05).
NS, there are no significant differences between the ‘Yes’ and ‘No’ periodontology subjects for CRP Finally, we adjusted for tomato intake in
or WBC at any MTC quantities. the sLyco–CHF risk model and sLyco in
the MTC–CHF risk model, to show that
these results were due to tomato con-
relationships were explored. Individuals carotenes, vitamin E and vitamin C sumption levels, rather than sLyco,
with periodontitis showed a dose– further reduced the relationship between which were not significantly correlated
response relationship between dietary sLyco and CHF risk even more; mod- after adjustments for confounders, ex-
MTC and self-reported CHF risk; mod- erate MTC (po0.05). When we further cept in the moderate sLyco-healthy
erate MTC, low MTC and very low adjusted for dietary MTC, the relation- periodontium subjects.
MTC (po0.05), adjusting for demo- ship between sLyco and CHF risk Lycopene, best known as an anti-
graphic, medical and lifestyle factors. remained the same; moderate sLyco oxidant, is mainly supplied by tomatoes,
Further adjusting for additional con- (po0.05). which contain many other beneficial
founders for both periodontitis and substances (high levels of vitamin A,
CHF, inflammatory biomarkers (CRP vitamin C, vitamin E, folate, potassium,
and WBC count), the dose–response bioflavonoids (especially quercetin)
relationship increased between dietary Discussion and phytosterols). Lycopene can be
MTC and CHF risk; moderate MTC, The results of this study provide evi- absorbed more efficiently by the body
low MTC (po0.05) and very low MTC dence of a relationship between perio- after it has been processed into juice,
(po0.01). Further adjusting for serum dontitis subjects with very low MTC sauce, paste, or ketchup (Fuhrman et al.
carotenes, vitamin E and vitamin C and a significantly elevated risk of CHF 1997). Tomato intake, in the form of
increased the significance of the rela- when compared with high MTC perio- juice, tomato oleoresin, tomato paste,
tionship between dietary MTC and dontitis subjects. MTC levels in indivi- tomato sauce and/or orange and red
CHF; moderate MTC, low MTC (po duals with healthy periodontium were tomatoes, delayed and reduced low
0.05) and very low MTC (po0.01). not significantly associated with the risk density lipoprotein (LDL) oxidation
When data were further adjusted for of CHF (all were low). Inflammatory (Fuhrman et al. 2000, Upritchard et al.
sLyco, the relationship between MTC markers, CRP and WBC count, were not 2000).
and CHF risk in periodontally involved significantly associated with an elevated In recent epidemiological studies,
individuals remained high; moderate risk for CHF. The findings of this study tissue and sLyco levels were inversely
MTC, low MTC (po0.05) and very did show evidence of significant inverse related to risk of coronary heart disease.
low MTC (po0.01). dose–response relationships between Increases in LDL oxidation have been
No dose–response relationship was sLyco levels and CRP, and MTC and hypothesized to be causally associated
found between the sLyco levels and WBC, in the periodontitis subjects; and with increasing risk of atherosclerosis
self-reported CHF risk in individuals also demonstrated a decreased risk of and coronary heart disease (Agarwal &
with periodontitis (Table 5). However, CHF with moderate sLyco levels in Rao 1998) and studies have shown
the moderate sLyco level-healthy perio- individuals with a healthy periodontium. contradictory lycopene effects on LDL
dontium group showed a significant A positive relationship was found oxidation (Carroll et al. 2000, Upritch-
decrease in CHF risk (po0.05) adjust- between high MTC and CHF risk in ard et al. 2000). As an antioxidant,
ing for age, race, gender, BMI, smoking periodontally involved subjects. lycopene has a singlet-oxygen-quench-
status, history of diabetes, hypertension, Cross-sectional and prospective stu- ing ability twice as high as that of b-
education level and socioeconomic sta- dies have established that elevated carotene (a vitamin A relative) and 10
tus. Further adjustment for additional peripheral blood levels of several sys- times higher than that of a-tocopherol (a
confounders for both periodontitis and temic inflammatory markers (including vitamin E relative) (Gerster 1997).
CHF, inflammatory biomarkers (serum CRP) are associated with the risk of In this study, the differences between
CRP and WBC count), showed a similar cardiovascular diseases and the severity sLyco levels and MTC groups seen
CHF risk reduction; moderate sLyco of atherosclerosis (Berk et al. 1990, periodontitis subjects and CHF risk,
(po0.05). Further adjustment for serum Maseri et al. 1996, Mendall et al. 1996, may be reflective of tomatoes contain-
578 Wood & Johnson

Table 4. Risk ratios (RRs) for congestive heart failure associated with periodontitis and healthy periodontium, and monthly tomato consumption
(MTC)
Periodontology MTC RR for CHF RR for CHF RR for CHF RR for CHF
status Y versus N Y versus N Y versus N Y versus N
(95% CI)n (95% CI)w (95% CI)z (95% CI)§

Y high 1.00 (reference) 1.00 (reference) 1.00 (reference) 1.00 (reference)

moderate 2.14 (0.83,5.55) 3.15 (1.03,9.67) 2.66 (0.85,8.31) 2.70 (0.86,8.44)
low 2.03 (0.93,4.43) 3.31 (1.33,8.24)nn 3.42 (1.28,9.13)nn 3.35 (1.26,8.93)nn
very low 3.35 (1.30,8.64)nn 5.10 (1.67,15.57)nnn 17.68 (2.30,136.11)nnn 17.50 (2.27,134.71)nnn
N high 1.32 (0.43,4.06) 1.31 (0.42,4.09) 1.10 (0.34,3.51) 1.47 (0.40,5.34)
moderate 0.92 (0.33,2.60) 1.78 (0.50,6.35) 1.60 (0.44,5.82) 1.57 (0.43,5.69)
low 1.46 (0.56,3.81) 2.23 (0.72,6.88) 1.99 (0.63,6.27) 1.98 (0.63,6.21)
very low 1.18 (0.47,2.92) 1.68 (0.60,4.76) 1.66 (0.53,5.24) 1.38 (0.48,3.99)
Y, yes periodontitis; N, no periodontitis.
n
Adjusted for age, race, gender, body mass index (BMI), smoking status, history of diabetes, hypertension, socioeconomic status and education level.
w
Adjusted for age, race, gender, BMI, waist to hip ratio, serum CRP, WBC count, smoking status, history of diabetes, hypertension, socioeconomic
status, and education level.
z
Adjusted for age, race, gender, BMI, waist to hip ratio, serum CRP, WBC count, serum carotene, serum vitamin E, serum vitamin C, smoking status,
history of diabetes, hypertension, socioeconomic status and education level.
§
Adjusted for age, race, gender, BMI, waist to hip ratio, serum CRP, WBC count, serum carotene, serum vitamin E, serum vitamin C, smoking status,
history of diabetes, hypertension, socioeconomic status, and education level and serum lycopene.
nn
po0.05,
nnn
po0.01.

Table 5. Risk ratios (RRs) for congestive heart failure associated with periodontitis and healthy periodontium, and serum lycopene (sLyco) levels
Periodontology SLyco quartiles RR for CHF RR for CHF RR for CHF RR for CHF
status Y versus N Y versus N Y versus N Y versus N
(95% CI)n (95% CI)w (95% CI)z (95% CI)§

Y high 1.00 (reference) 1.00 (reference) 1.00 (reference) 1.00 (reference)

moderate 0.71 (0.22,2.29) 0.72 (0.23,2.32) 0.64 (0.18,2.32) 0.64 (0.18,2.32)
low 0.42 (0.15,1.22) 0.42 (0.14,1.21) 0.39 (0.12,1.28) 0.39 (0.12,1.28)
very low 0.53 (0.18,1.59) 0.65 (0.21,2.03) 0.70 (0.19,2.52) 0.85 (0.22,3.21)
N high 0.67 (0.16,2.87) 0.66 (0.15,2.82) 0.52 (0.11,2.39) 0.51(0.11,2.36)
moderate 0.24 (0.07,0.81)nn 0.25 (0.07,0.84)nn 0.19 (0.05,0.69)nn 0.19 (0.05,0.70)nn
low 0.71 (0.17,3.06) 1.04 (0.20,5.49) 0.83 (0.15,4.65) 0.56 (0.12,2.59)
very low 0.62 (0.16,2.36) 0.59 (0.15,2.24) 0.57 (0.12,2.62) 0.59 (0.13,2.71)
Y, yes periodontitis; N, no periodontitis.
n
Adjusted for age, race, gender, body mass index (BMI), smoking status, history of diabetes, hypertension, socioeconomic status and education level.
w
Adjusted for age, race, gender, BMI, waist to hip ratio, serum CRP, WBC count, smoking status, history of diabetes, hypertension, socioeconomic
status and education level.
z
Adjusted for age, race, gender, BMI, waist to hip ratio, serum CRP, WBC count, serum carotene, serum vitamin E, serum vitamin C, smoking status,
history of diabetes, hypertension, socioeconomic status and education level.
§
Adjusted for age, race, gender, BMI, waist to hip ratio, serum CRP, WBC count, serum carotene, serum vitamin E, serum vitamin C, smoking status,
history of diabetes, hypertension, socioeconomic status, and education level and monthly tomato intake.
nn
po0.05,

ing high levels of other antioxidants. A Lycopene appears to be the most IMT) has been reported by the ARIC
sufficient supply of antioxidants from effective carotenoid in reducing both study (Beck et al. 2001). With coronary
the diet might help prevent or delay the human aortic endothelial cell adhesion artery disease being the most common
occurrence of pathological changes to monocytes and expression of adhe- etiologic factor for left ventricular
associated with oxidative stress (Giu- sion molecules on the cell surface systolic dysfunction, the major symp-
gliano 2000). (Martin et al. 2000). An inverse rela- tom of CHF, reduction or maintenance
It has also been proposed that high tionship between plasma lycopene con- of CCA-IMT could reduce CHF risk.
levels of reactive oxygen species pro- centrations and CCA-IMT in Finnish Taken together, our data confirms
mote progression of periodontal disease. males has been reported (Rissanen et al. other data that periodontitis may have
This may be a result of relatively low 2000), and it was concluded that low systemic sequelae, as serum levels of
levels of superoxide dismutase and plasma lycopene concentrations was CRP and WBC count were both ele-
catalase in the tissues (Ellis et al. associated with early atherosclerosis, vated in individuals with periodontitis.
1998). The beneficial effects of lyco- manifested as increased CCA-IMT (Ris- We theorize that a relationship exists
pene may be a result of lowering sanen et al. 2000). A direct relationship between periodontitis and elevated CHF
gingival tissue free radical concentra- between periodontitis and carotid artery risk, and that higher MTC reduced this
tions, arresting disease progression. intima-media wall thickness (CCA- risk in periodontitis subjects. CRP
 Tomato intake, periodontitis, congestive heart failure 579

levels and WBC count were reduced in Forrester, J., Merz, C., Bush, T., Cohn, J., American Journal of Epidemiology 146,
individuals with higher sLyco levels and Hunninghake, D. & Parthasarathy, S. (1996) 618–626.
higher MTC levels, respectively. The 27th Bethesda Conference: matching the Liuizzo, G. & Rizello, V. (2001) C-reactive
current observations may explain the intensity of risk factor management with the protein and primary prevention of ischemic
hazard for coronary heart disease events. heart disease. Clinical Chimica Acta 311,
epidemiologic links between tomato
Journal of the American College of Nutrition 45–48.
consumption and incidence of inflam- 27, 996–1006. Loos, B. G., Craandijk, J., Hoek, F. J.,
matory diseases and explain the hypoth- Fowler, E., Breault, L. & Cuenin, M. (2001) Wertheim-van Dillen, P. M. & van der
esis that diets rich in antioxidants Periodontal disease and its association with Velden, U. (2000) Evaluation of systemic
reduce the risk of CHF in individuals systemic disease. Military Medicine 166, markers related to cardiovascular disease in
with periodontitis. 85–89. peripheral blood of periodontitis patients.
Fuhrman, B., Elis, A. & Aviram, M. (1997) Journal of Periodontology 71, 1528–1534.
Hypercholesterolemic effect of lycopene and Martin, K., Wu, D. & Meydani, M. (2000) The
beta-carotene is related to suppression of effect of carotenoids on the expression of cell
References cholesterol synthesis and augmentation of surface adhesion molecules and binding of
LDL receptor activity in macrophages. monocytes to human aortic endothelial cells.
Agarwal, S. & Rao, A. (1998) Tomato lycopene Biochemistry and Biophysical Research Atherosclerosis 150, 265–274.
and low density lipoprotein oxidation: a Communications 233, 658–662. Maseri, A., Biasucci, L. & Liuzzo, G. (1996)
human dietary intervention study. Lipids 33, Fuhrman, B., Volkova, N., Rosenblat, M. & Elevated levels of interleukin-6 in unstable
981–984. Aviram, M. (2000) Lycopene synergistically angina. Circulation 94, 874–877.
Agarwal, S. & Rao, A. (2000) Tomato lycopene inhibits LDL oxidation in combination with Mayne, S. (1996) Beta-carotene, carotenoids,
and its role in human health and chronic vitamin E, glabridin, rosmarinic acid, carno- and disease prevention in humans. FASEB
diseases. Canadian Medical Association sic acid, or garlic. Antioxidant Redox Signal- Journal 10, 690–701.
Journal 163, 739–744. ing 2, 491–506. Mendall, M. A., Patel, P., Ballam, L., Strachan,
Arbes, S. Jr., Slade, G. & Beck, JD. (1999b) Genco, R., Offenbacher, S. & Beck, J. (2002) D. & Northfield, T. C. (1996) C reactive
Association between extent of periodontal Periodontal disease and cardiovascular dis- protein and its relation to cardiovascular risk
attachment loss and self-reported history of ease. Epidemiology and possible mechan- factors: a population based cross sectional
heart attack: an analysis of NHANES III isms. Journal of the American Dental study. BMJ 312, 1061.
data. Journal of Dental Research 78, 1777– Association 133, 14S–22S. National Heart, Lung and Blood Institute
1782. Gerster, H. (1997) The potential role of (September 1996) www.nhlbi.nih.gov/health/
Beck, J., Elter, J., Heiss, G., Couper, D., lycopene for human health. Journal of the
public/heart/other/CHF.htm.
Mauriello, S. & Offenbacher, S. (2001) American College of Nutrition 16, 109–126.
Phillips, A., Neaton, J., Cook, D., Grimm, R. &
Relationship of periodontal disease to carotid Giugliano, D. (2000) Dietary antioxidants for
Sharper, A. (1992) Leukocyte count and risk
artery intima-media wall thickness: the cardiovascular prevention. Nutrition Metabo-
of major coronary heart disease events.
atherosclerosis risk in communities (ARIC) lism and Cardiovascular Diseases 10, 38–44.
American Journal of Epidemiology 136,
study. Arteriosclerosis Thrombosis and Vas- Grossi, S. & Genco, R. (1998) Periodontal
59–70.
cular Biology 21, 1816–1822. disease and diabetes mellitus: a two way
Rao, A. & Agarwal, S. (2000) Tomato lycopene
Beck, J., Offenbacher, S., Williams, R., Gibbs, relationship. Annals of Periodontology 3,
and its role in human health and chronic
P. & Garcia, R. (1998) Periodontitis: a risk 51–61.
diseases. CMAJ 163, 739–744.
factor for coronary heart disease. Annals of Haraszthy, V., Zambon, J., Trevisan, M., Zeid,
Ridker, P., Buring, J., Shih, J., Matais, M. &
Periodontology 3, 127–141. M. & Genco, R. (2000) Identification of
Hennekens, C. (1999) Prospective study of
Berk, B., Weintraub, W. & Alexander, R. periodontal pathogens in atheromatous
plaques. Journal of Periodontology 71, C-reactive protein and the risk of future
(1990) Elevation of C-reactive protein in
1554–1560. cardiovascular events among apparently
‘‘active’’ coronary artery disease. American
Journal of Cardiology 65, 168–172. Herzberg, M., Brintzenhofe, K. & Clawson, C. healthy women. Circulation 104.
Carlos, J., Wolfe, M. & Kingman, A. (1986) (1983) Aggregation of human platelets and Ridker, P. M., Hennekens, C. H., Buring, J. E.
The extent and severity index: a simple adhesion of Streptococcus sanguis. Infection & Rafai, N. (2000a) C-reactive protein and
method for use in epidemiologic studies of and Immunity 39, 1457–1469. other markers of inflammation in the predic-
periodontal disease. Journal of Clinical Herzberg, M. & Meyer, M. (1996) Effects of tion of cardiovascular disease in women.
Periodontology 13, 500–505. oral flora on platelets: possible consequences New England Journal of Medicine 342, 836.
Carroll, Y., Corridan, B. & Morrissey, P. (2000) in cardiovascular disease. Journal of Perio- Ridker, P. M., Rifai, N., Stampfer, M. J. &
Lipoprotein carotenoid profiles and the dontology 67, 1138–1142. Hennekens, C. H. (2000b) Plasma concentra-
susceptibility of low density lipoprotein to Kannel, W., Anderson, K. & Wilson, P. (1992) tion of interleukin-6 and the risk of future
oxidative modification in healthy elderly White blood cell count and cardiovascular myocardial infarction among apparently
volunteers. European Journal of Clinical disease. Insights from the Framingham study. healthy men. Circulation 101, 1767.
Nutrition 54, 500–507. JAMA 267, 1253–1256. Rissanen, T., Voutilainen, S., Nyyssonen, K.,
Chopra, M., O’Neill, M., Keogh, N., Wortley, Kinane, D. & Marshall, G. (2001) Periodontal Salonen, R. & Salonen, J. (2000) Low plasma
G., Southon, S. & Thurnham, D. (2000) manifestations of systemic disease. Austra- lycopene concentration is associated with
Influence of increased fruit and vegetable lian Dental Journal 46, 45–48. increased intima-media thickness of the
intake on plasma and lipoprotein carotenoids Klipstein-Grobusch, K., Launer, L., Geleijnse, carotid artery wall. Arteriosclerrosis Throm-
and LDL oxidation in smokers and non- J., Boeing, H., Hofman, A. & Witteman, J. bosis and Vascular Biology 20, 2677–2681.
smokers. Clinical Chemistry 46, 1818–1829. (2000) Serum carotenoids and atherosclero- Slade, G., Offenbacher, S., Beck, J., Heiss, G. &
Ellis, S., Tucci, M., Serio, F. & Johnson, R. sis. The Rotterdam Study. Atherosclerosis Pankow, J. (2000) Acute-phase inflammatory
(1998) Factors for progression of periodontal 148, 49–56. response to periodontal disease in the U.S.
disease. Journal of Oral Pathology and Kohlmeier, L., Kark, J., Gomez-Gracia, E., population. Journal of Dental Research 79,
Medicine 27, 101–105. Martin, B., Steck, S., Kardinaal, A., Ring- 49–57.
Ezzati, T., Massey, J., Waksberg, J., Chu, A. & stad, J., Thamm, M., Masaev, V., Riemers- Taylor, G. (2001) Bidirectional interrelation-
Maurer, K. (1992) Sample Design: Third ma, R., Martin-Moreno, J., Huttunen, J. & ships between diabetes and periodontal dis-
National Health and Nutrition Examination Kok, F. (1997) Lycopene and myocardial eases: an epidemiologic perspective. Annals
Survey. Vital Health Statistics 2 113, 1–35. infarction risk in the EURAMIC Study. of Periodontology 6, 99–112.
580 Wood & Johnson

United States Department of Health and Human United States Department of Health and Human serum total and high density lipoprotein
Services (USDHHS) (1996a) Third National Services (USDHHS) (1996c) Third National cholesterol, C-reactive protein, and plasma
Health and Nutrition Examination Survey, Health and Nutrition Examination Survey, fibrinogen. American Journal of Epidemiol-
1988–1994, NHANES III Household Adult 1988–1994, NHANES III Laboratory Data ogy 151, 273–282.
Data File (CD-ROM). Public Use Data File. File Documentation. Hyattsville, MD: National
Centre for Health Statistics.
Documentation Number 76200. Hyattsville,
Upritchard, J., Sutherland, W. & Mann, J. (2000)
MD: National Centre for Health Statistics.
Effect of supplementation with tomato juice, Address:
United States Department of Health and Human
vitamin E, and vitamin C on LDL oxidation Nelson Wood
Services (USDHHS) (1996b) Third National
and products of inflammatory activity in type Department of Periodontics
Health and Nutrition Examination Survey, 2 diabetes. Diabetes Care 23, 733–738. University of Mississippi Medical Center
1988–1994, NHANES III Household Adult Wu, T., Trevisan, M., Genco, R., Falkner, K., 2500 North State Street
Data File (CD-ROM). Public Use Data File. Dorn, J. & Sempos, C. (2000a) Examination Jackson, MS 39216-4505, USA
Documentation Number 76300. Hyattsville, of the relationship between periodontal Fax: 1601 984 6120
MD: National Centre for Health Statistics. health status and cardiovascular risk factors: E-mail: [email protected]