Abstract
Few epidemiologic studies have examined the potential cardiovascular mechanisms of tomato-based food products, the primary dietary source of lycopene. We examined the cross-sectional association between tomato-based food product intake and coronary biomarkers in the Women’s Health Study. Tomato-based food products (tomatoes, tomato juice, tomato sauce, pizza) were summed from a semiquantitative FFQ and multiple risk factors ascertained. Plasma from baseline blood samples were assayed for lipids, lipoproteins, hemoglobin A1c, C-reactive protein, fibrinogen, soluble intracellular adhesion molecule-1, and creatinine. A total of 27,261 women aged ≥45 y who were free of cardiovascular disease and cancer provided relevant data for this study. Tomato-based food product intake was modest, with 84% of women consuming <1 serving/d, but those with greater intake had healthier lifestyle and dietary habits. Women consuming ≥10 compared with <1.5 servings/wk of tomato-based food products had significant but clinically modest improvements in total cholesterol (TC) (5.38 vs. 5.51 mmol/L; P = 0.029), the TC:HDL cholesterol ratio (4.08 vs. 4.22; P = 0.046), and hemoglobin A1c (5.02 vs. 5.13%; P < 0.001) in multivariable models. Considering clinical cutpoints, women consuming ≥10 compared with <1.5 servings/wk were 31% (95% CI = 6%, 50%), 40% (95% CI = 13%, 59%), and 66% (95% CI = 20%, 86%) less likely to have elevated TC (≥6.21 mmol/L), LDL cholesterol (≥4.14 mmol/L), and hemoglobin A1c (≥6%), respectively. Other coronary biomarkers were unassociated with tomato-based food products. In conclusion, women consuming ≥10 compared with <1.5 servings/wk of tomato-based food products had clinically modest but significant improvements in TC, the TC:HDL cholesterol ratio, and hemoglobin A1c but not other coronary biomarkers.
Introduction
Epidemiological studies have consistently established an association between higher levels of fruit and vegetable intake and a lower risk of CVD7 in a wide variety of populations (1, 2). Tomato-based food products, the primary dietary source of >80% of lycopene in the American diet (3), have generated interest for their potential impact on coronary risk factors and risk of CVD. Previously, in the WHS, middle-aged and older U.S. women consuming 7 to <10 and ≥10 servings/wk of tomato products had corresponding 32 and 29% reductions in the risk of CVD compared with women consuming <1.5 servings/wk (4). Yet these potential benefits for tomato-based food products were not explained by their dietary lycopene content, which is more bioavailable in processed tomato products (5). A meta-analysis of smaller lycopene intervention trials suggests modest benefits on LDL cholesterol and systolic blood pressure (6). Otherwise, results from other studies of dietary lycopene are inconsistent, noting a potential inverse association with stroke (7) and a lack of association with the risk of diabetes (8). It may therefore be more important to focus upon the whole tomato as a source of vitamin A, vitamin C, potassium, and other nutrients for its association with CVD, because food-based nutritional strategies offer a more generalizable dietary recommendation.
No epidemiologic studies to our knowledge have evaluated the association between the intake of tomato-based food products and coronary biomarkers to identify potentially relevant mechanisms through which consuming tomato products may affect the risk of CVD. We therefore examined cross-sectional data within the WHS that includes comprehensive risk factor information and previously measured, well-represented coronary biomarkers in nearly 28,000 women and dietary intake including tomato-based food products.
Materials and Methods
Study population.
The WHS is a completed, 2 × 2 factorial trial of low-dose aspirin (9) and vitamin E (10) in the primary prevention of CVD and cancer. Additional details of the WHS study design are available elsewhere (11, 12). A total of 39,876 female U.S. health professionals aged ≥45 y who were postmenopausal or not intending to become pregnant and were free from prior MI, stroke, transient ischemic attack, and cancer (except nonmelanoma skin cancer) were randomized into the WHS in 1992. Written informed consent was obtained from all participants and this research was approved by the institutional review board of Brigham and Women’s Hospital, Boston, MA.
At baseline, women were sent and asked to complete a 131-item validated semiquantitative FFQ, for which 39,310 (98.6%) women responded (13). Based upon their responses, we excluded 829 women who completed an insufficient number of food items or had a total energy intake of either <2510 or ≥14,644 kJ/d. We also excluded women if they reported angina pectoris, coronary artery by-pass graft surgery, or percutaneous transluminal coronary artery angioplasty at the time of randomization. In addition, participants must have responded to questions about tomato-based food products on the FFQ to be included.
Assessment of tomato-based food products and other covariates.
Four major tomato-based food products were measured on the FFQ, including tomatoes, tomato juice, tomato sauce, and pizza. For each food, participants selected from 9 responses ranging in the number of servings from “never or less than once per month” to “6+ per day.” We categorized tomatoes (1 serving = 1 standard portion size, 1 tomato) as none, 1–3 servings/mo, 1–4 servings/wk, and ≥5 servings/wk; for tomato juice (small glass), none, 1–3 servings/mo, 1 serving/wk, and ≥2 servings/wk; for tomato sauce (1/2 cup or 118 mL), none, 1–3 servings/mo, 1 serving/wk, and ≥2 servings/wk; and for pizza (2 slices), none, 1–3 servings/mo, 1 serving/wk, and ≥2 servings/wk. We then summed the total intake of tomato-based products consistent with previous publications (4, 8) as <1.5, 1.5 to <4, 4 to <7, 7 to <10, and ≥10 servings/wk. In addition to tomato-based food products, nutrient values were measured and energy adjusted using the residual method (14) from food tables maintained by the Department of Nutrition at the Harvard School of Public Health, Boston, MA. In our analyses, we considered intake of fruit and vegetables (servings/d), total fiber (g/d), saturated fat (g/d), vitamin C (mg/d), and the primary nutrient contained in tomato-based food products, lycopene (μg/d).
On the WHS baseline questionnaire, women also provided self-reported information on age (in years), weight and height (as BMI, in kg/m2), smoking status (never, former, current), alcohol use (rarely/never, 1–3 drinks/mo, 1–6 drinks/wk, and ≥1 drink/d), frequency of exercise (rarely/never, <1 time/wk, 1–3 times/wk, and ≥4 times/wk), parental history of MI at <60 y (no, yes), history of hypertension (no, yes), history of hypercholesterolemia (no, yes), and postmenopausal hormone use (never, former, or current).
For our analyses, we included 38,480 women with data on tomato-based food product intake and associated clinical, behavioral, and dietary risk factors.
Blood assays.
Prior to randomization into the WHS, 28,345 (71%) of 39,876 participants provided baseline blood samples sent via overnight courier that were aliquotted and stored in liquid nitrogen until analysis. Of the samples received, 27,939 could be evaluated and were transferred to a core laboratory facility certified by the National Heart, Lung, and Blood Institute/CDC Lipid Standardization Program. As previously detailed (15, 16), TC, LDL cholesterol, HDL cholesterol, and TG were directly assayed using reagents from Genzyme and Roche Diagnostics. C-reactive protein was measured using a validated, high-sensitivity immunoturbidimetric assay on a Hitachi 917 analyzer from Roche Diagnostics, using reagents and calibrators from Denka Seiken (17). Hemoglobin A1c was also assessed from packed RBC using an immunoturbidimetric assay by Roche Diagnostics on a Hitachi 911 analyzer. Fibrinogen was measured with an immunoturbidimetric assay by Kamiya Biomedical and sICAM-1 was measured using ELISA by R&D Systems. Creatinine was measured by a rate-blanked method that is based on the Jaffé reaction using Roche Diagnostics reagents. Lipoprotein(a) was measured with an immunoturbidimetric assay by Roche Diagnostics with reagents and calibrators from Denka Seiken. Finally, apoA1 and B-100 were measured with immunoturbidimetric assays by DiaSorin.
Among the 38,480 women included in the analyses of tomato-based food product intake and associated baseline risk factors, 27,261 (70.8%) also provided baseline blood samples with biomarker data available for analysis.
Data analyses.
All analyses were conducted with a 2-sided α = 0.05. Data in the text are mean ± SE unless otherwise noted. We first evaluated the associations between total tomato-based food product intake with various lifestyle, clinical, and dietary risk factors. We also calculated the Spearman correlations among total and individual tomato-based food products. Next, we examined the associations of increasing categories of total tomato-based food product intake with coronary biomarkers, including lipids and lipoproteins, C-reactive protein, hemoglobin A1c, sICAM-1, fibrinogen, and creatinine. Models were first adjusted for age, then other lifestyle factors, followed by clinical factors, and finally dietary factors. For these models, we compared the mean ± SE level of each coronary biomarker across categories of total tomato-based food product intake. The above analysis approach was repeated for each individual tomato-based food product.
Next, we considered the association between increasing categories of individual and total tomato-based food products and the likelihood of being classified at adverse U.S. National Cholesterol Education Program clinical cutpoints for lipids (TC, ≥6.21 vs. <6.21 mmol/L; LDL cholesterol, ≥4.14 vs. <4.14 mmol/L; HDL cholesterol, <1.55 vs. ≥1.55 mmol/L; TC:HDL cholesterol ratio, ≥6 vs. <6 units; and TG, <2.26 vs. ≥2.26 mmol/L) (18), the AHA/CDC cutpoints for elevated C-reactive protein (≥3 vs. <3 mg/L) (19), and hemoglobin A1c (<6 vs. ≥6%). Logistic regression models generated the OR (95% CI) of having an adverse lipid, C-reactive protein, or hemoglobin A1c concentration with increasing categories of individual and total tomato-based food product intake using a modeling strategy paralleling those for the continuous coronary biomarkers.
Lastly, we examined whether the association of individual and total tomato-based food product intake and coronary biomarkers is modified by baseline diabetes, hypertension, and high cholesterol out of concern that participants may have recently altered the composition of their diet.
Results
The 38,480 women who provided information on the FFQ reported modest consumption of tomato-based food products, with 4.3 ± 3.2 servings/wk [median (interquartile range), 3.9 (1.9–5.5)], of which 2.2, 1.1, 0.6, and 0.5 servings/wk were from tomatoes, tomato sauce, pizza, and tomato juice, respectively. Most women (84%) consumed <1 serving/d, with 17.1, 36.9, 29.6, 11.9, and 4.5% of women in the respective categories of <1.5, 1.5 to <4, 4 to <7, 7 to <10, and ≥10 servings/wk of tomato-based food products. Total tomato-based food product intake had Spearman correlations (r) of 0.81, 0.36, 0.58, and 0.27 (all P < 0.05) for tomatoes, tomato sauce, pizza, and tomato juice, respectively. In addition, intakes of total (Spearman r = 0.84; P < 0.001) and individual (tomatoes, r = 0.67, P < 0.001; tomato juice, r = 0.35, P < 0.001; tomato sauce, r = 0.56, P < 0.001) tomato-based food products were correlated with dietary lycopene concentrations. Pizza comprised a smaller proportion of tomato-based food product intake and was correlated with dietary lycopene, but at a weaker magnitude (Spearman r = 0.05; P < 0.001).
Women consuming greater amounts of total tomato-based food products were younger (P < 0.001) and had a pattern of healthier lifestyle and dietary risk factors, because they were more likely to exercise, less likely to currently smoke, and consumed greater amounts of fruits and vegetables, fiber, vitamin C, and lycopene and a smaller amount of saturated fat (all P < 0.001) (Table 1). There was a general U-shaped pattern to clinical risk factors (BMI, hypertension, hypercholesterolemia, and diabetes) across categories of tomato-based food product intake.
TABLE 1.
Baseline risk factors according to categories of total tomato-based food product intake in 38,480 women from the WHS in 19921
| <1.5 Servings/wk (n = 6577) | 1.5 to <4 Servings/wk (n = 14,218) | 4 to <7 Servings/wk (n = 11,377) | 7 to <10 Servings/wk (n = 4589) | ≥10 Servings/wk (n = 1719) | P value2 | |
| Age, y | 55.5 ± 7.5 | 54.6 ± 7.0 | 54.4 ± 6.9 | 54.2 ± 6.9 | 54.1 ± 6.8 | <0.001 |
| BMI, kg/m2 | 26.0 ± 5.1 | 25.9 ± 5.0 | 26.0 ± 4.9 | 26.1 ± 5.1 | 26.4 ± 5.5 | 0.006 |
| History of hypertension, % | 27.4 | 24.9 | 24.9 | 27.5 | 28.2 | <0.001 |
| History of hypercholesterolemia,% | 31.5 | 29.3 | 28.7 | 28.9 | 30.4 | 0.001 |
| History of diabetes mellitus, % | 2.8 | 2.4 | 2.3 | 2.5 | 3.6 | 0.01 |
| Parental history of MI < 60 y, % | 12.2 | 13.2 | 12.3 | 14.1 | 13.3 | 0.018 |
| Exercise, % | <0.001 | |||||
| Rarely/never | 44.8 | 39.2 | 35.7 | 33.8 | 32.8 | |
| <1 time/wk | 17.8 | 20.3 | 20.4 | 20.5 | 19.6 | |
| 1–3 times/wk | 27.8 | 30.4 | 32.8 | 33.8 | 34.9 | |
| ≥4 times/wk | 9.6 | 10.1 | 11.1 | 12.0 | 12.8 | |
| Smoking status, % | <0.001 | |||||
| Never | 51.2 | 51.5 | 51.0 | 49.8 | 50.9 | |
| Former | 34.0 | 35.3 | 36.6 | 39.1 | 37.8 | |
| Current | 14.8 | 13.2 | 12.4 | 11.1 | 11.3 | |
| Alcohol consumption, % | <0.001 | |||||
| Rarely/never | 54.5 | 44.8 | 41.0 | 41.4 | 41.3 | |
| 1–3 drinks/mo | 13.2 | 13.9 | 13.0 | 12.1 | 11.4 | |
| 1–6 drinks/wk | 24.8 | 31.9 | 34.4 | 33.8 | 33.2 | |
| ≥1 drink/d | 7.5 | 9.4 | 11.6 | 12.7 | 14.1 | |
| Hormone replacement therapy, % | 0.11 | |||||
| Never | 49.1 | 50.2 | 50.0 | 50.4 | 49.2 | |
| Former | 10.2 | 8.9 | 8.7 | 8.8 | 9.2 | |
| Current | 40.8 | 40.9 | 41.3 | 40.8 | 41.6 | |
| Fruit/vegetable intake, servings/d | 10.5 ± 3.1 | 13.5 ± 3.1 | 16.6 ± 3.5 | 19.5 ± 3.9 | 23.7 ± 5.3 | <0.001 |
| Total energy intake, kcal/d | 5960 ± 1940 | 6850 ± 1990 | 7690 ± 2100 | 8480 ± 2250 | 9270 ± 2430 | <0.001 |
| Total fiber intake,3g/d | 17.6 ± 6.4 | 18.2 ± 5.5 | 19.6 ± 5.6 | 20.6 ± 6.0 | 22.6 ± 6.6 | <0.001 |
| Saturated fat intake,3g/d | 20.2 ± 5.5 | 20.1 ± 4.7 | 19.4 ± 4.6 | 19.0 ± 4.6 | 18.3 ± 4.7 | <0.001 |
| Vitamin C intake,3mg/d | 131 ± 68 | 139 ± 59 | 153 ± 59 | 166 ± 53 | 189 ± 74 | <0.001 |
| Dietary lycopene intake,3mg/d | 3.59 ± 1.89 | 6.53 ± 2.66 | 11.0 ± 4.14 | 15.3 ± 5.30 | 24.3 ± 11.8 | <0.001 |
| Individual tomato-based food products,4servings/wk | ||||||
| Tomatoes | 0.47 ± 0.34 | 1.23 ± 0.91 | 2.73 ± 1.35 | 4.45 ± 1.95 | 7.19 ± 4.95 | <0.001 |
| Tomato juice | 0.06 ± 0.16 | 0.23 ± 0.35 | 0.46 ± 0.79 | 0.96 ± 1.44 | 2.55 ± 3.53 | <0.001 |
| Tomato sauce | 0.37 ± 0.29 | 0.74 ± 0.49 | 1.42 ± 1.05 | 1.94 ± 1.34 | 3.21 ± 2.47 | <0.001 |
| Pizza | 0.25 ± 0.28 | 0.53 ± 0.41 | 0.63 ± 0.65 | 0.84 ± 0.86 | 0.94 ± 1.23 | <0.001 |
Values are mean ± SD or percent. The data represent the association between categories of total tomato-based food products and various lifestyle, clinical, and dietary risk factors for CVD.
We used a global ANOVA test for continuous variables and chi-square tests (4 degrees of freedom) for categorical variables.
Energy-adjusted.
One serving is equivalent to 1 tomato, a small glass of tomato juice, 1/2 cup (118 mL) of tomato sauce, or 2 slices of pizza.
We next considered the cross-sectional association between increasing categories of total tomato-based food product intake and age- and multivariable-adjusted mean concentrations of coronary biomarkers in the subset of 27,128 women free of CVD with available biomarker assays (Table 2). The addition of lifestyle factors to age-adjusted models exhibited almost no confounding. Clinical risk factors had only a small impact on the associations. Dietary factors, including total energy intake and fruit and vegetable, total fiber, saturated fat, vitamin C, and lycopene intake, strongly confounded the association between tomato-based food product intake and each biomarker of interest and tended to drive the overall interpretation of results.
TABLE 2.
Baseline age- and multivariable-adjusted mean concentrations of lipids, hemoglobin A1c, C-reactive protein, and other coronary biomarkers among 27,261 women according to categories of total tomato-based food product intake1
| Weekly consumption of total tomato-based food products |
||||||
| Coronary biomarker | <1.5 Servings/wk (n = 4610) | 1.5 to <4 Servings/wk (n = 10,125) | 4 to <7 Servings/wk (n = 8108) | 7 to <10 Servings/wk (n = 3222) | ≥10 Servings/wk (n = 1196) | P-trend2 |
| TC, mmol/L | ||||||
| Age-adjusted | 5.49 ± 0.02 | 5.50 ± 0.01 | 5.47 ± 0.01 | 5.50 ± 0.02 | 5.47 ± 0.03 | 0.44 |
| +Lifestyle, clinical factors3 | 5.49 ± 0.02 | 5.50 ± 0.01 | 5.47 ± 0.01 | 5.50 ± 0.02 | 5.45 ± 0.03 | 0.46 |
| +Lifestyle, clinical, dietary factors4 | 5.51 ± 0.02 | 5.51 ± 0.01 | 5.46 ± 0.01 | 5.47 ± 0.02 | 5.38 ± 0.05* | 0.029 |
| LDL cholesterol, mmol/L | ||||||
| Age-adjusted | 3.23 ± 0.01 | 3.23 ± 0.01 | 3.20 ± 0.01 | 3.22 ± 0.02 | 3.19 ± 0.03 | 0.06 |
| +Lifestyle, clinical factors3 | 3.21 ± 0.01 | 3.22 ± 0.01 | 3.21 ± 0.01 | 3.23 ± 0.01 | 3.19 ± 0.02 | 0.48 |
| +Lifestyle, clinical, dietary factors4 | 3.22 ± 0.02 | 3.23 ± 0.01 | 3.20 ± 0.01 | 3.21 ± 0.02 | 3.16 ± 0.04 | 0.26 |
| HDL cholesterol, mmol/L | ||||||
| Age-adjusted | 1.39 ± 0.01 | 1.40 ± 0.01 | 1.39 ± 0.01 | 1.40 ± 0.01 | 1.39 ± 0.01 | 0.90 |
| +Lifestyle, clinical factors3 | 1.41 ± 0.01 | 1.40 ± 0.01 | 1.39 ± 0.01* | 1.39 ± 0.01* | 1.38 ± 0.01* | 0.003 |
| +Lifestyle, clinical, dietary factors4 | 1.40 ± 0.01 | 1.39 ± 0.01 | 1.39 ± 0.01 | 1.40 ± 0.01 | 1.40 ± 0.02 | 0.90 |
| TC:HDL cholesterol ratio, units | ||||||
| Age-adjusted | 4.23 ± 0.02 | 4.20 ± 0.01 | 4.19 ± 0.01 | 4.20 ± 0.02 | 4.22 ± 0.04 | 0.75 |
| +Lifestyle, clinical factors3 | 4.16 ± 0.02 | 4.20 ± 0.01* | 4.21 ± 0.01* | 4.23 ± 0.02* | 4.22 ± 0.03 | 0.020 |
| +Lifestyle, clinical, dietary factors4 | 4.22 ± 0.02 | 4.22 ± 0.01 | 4.19 ± 0.01 | 4.16 ± 0.03 | 4.08 ± 0.05* | 0.024 |
| TG,5mmol/L | ||||||
| Age-adjusted | 1.39 ± 0.01 | 1.38 ± 0.01 | 1.38 ± 0.01 | 1.40 ± 0.01 | 1.43 ± 0.01 | 0.10 |
| +Lifestyle, clinical factors3 | 1.37 ± 0.01 | 1.38 ± 0.01 | 1.39 ± 0.01 | 1.39 ± 0.01 | 1.41 ± 0.01 | 0.04 |
| +Lifestyle, clinical, dietary factors4 | 1.39 ± 0.01 | 1.40 ± 0.01 | 1.38 ± 0.01 | 1.36 ± 0.01 | 1.34 ± 0.01 | 0.07 |
| Hemoglobin A1c, % | ||||||
| Age-adjusted | 5.12 ± 0.01 | 5.10 ± 0.01 | 5.09 ± 0.01* | 5.09 ± 0.01 | 5.10 ± 0.02 | 0.27 |
| +Lifestyle, clinical factors3 | 5.10 ± 0.01 | 5.09 ± 0.01 | 5.09 ± 0.01 | 5.09 ± 0.01 | 5.08 ± 0.01 | 0.20 |
| +Lifestyle, clinical, dietary factors4 | 5.13 ± 0.01 | 5.11 ± 0.01* | 5.09 ± 0.01* | 5.06 ± 0.01* | 5.02 ± 0.02* | < 0.001 |
| C-reactive protein,5nmol/L | ||||||
| Age-adjusted | 17.1 ± 9.7 | 17.4 ± 9.6 | 17.8 ± 9.7 | 18.1 ± 9.7 | 18.1 ± 9.9 | 0.021 |
| +Lifestyle, clinical factors3 | 16.9 ± 9.7 | 17.4 ± 9.6 | 18.0 ± 9.6* | 18.0 ± 9.7* | 17.3 ± 9.8 | 0.018 |
| +Lifestyle, clinical, dietary factors4 | 17.4 ± 9.7 | 17.6 ± 9.6 | 17.7 ± 9.6 | 17.3 ± 9.8 | 16.0 ± 10.0 | 0.20 |
| Fibrinogen,5μmol/L | ||||||
| Age-adjusted | 10.4 ± 0.03 | 10.3 ± 0.03* | 10.3 ± 0.03* | 10.3 ± 0.03* | 10.3 ± 0.03* | 0.003 |
| +Lifestyle, clinical factors3 | 10.4 ± 0.03 | 10.3 ± 0.03 | 10.4 ± 0.03 | 10.3 ± 0.03 | 10.3 ± 0.03 | 0.49 |
| +Lifestyle, clinical, dietary factors4 | 10.4 ± 0.03 | 10.4 ± 0.03 | 10.3 ± 0.03 | 10.3 ± 0.03 | 10.2 ± 0.03* | 0.08 |
| sICAM-1,5μg/L | ||||||
| Age-adjusted | 346 ± 1.00 | 345 ± 1.00 | 345 ± 1.00 | 347 ± 1.00 | 348 ± 1.01 | 0.52 |
| +Lifestyle, clinical factors3 | 344 ± 1.00 | 345 ± 1.00 | 346 ± 1.00 | 348 ± 1.00* | 349 ± 1.01* | 0.001 |
| +Lifestyle, clinical, dietary factors4 | 343 ± 1.00 | 345 ± 1.00 | 346 ± 1.00 | 349 ± 1.01 | 350 ± 1.01 | 0.07 |
| Creatinine,5μmol/L | ||||||
| Age-adjusted | 62.9 ± 89 | 63.0 ± 89 | 62.7 ± 89 | 62.1 ± 89* | 62.6 ± 89 | 0.005 |
| +Lifestyle, clinical factors3 | 63.0 ± 89 | 63.0 ± 89 | 62.6 ± 89 | 62.1 ± 89* | 62.6 ± 89 | 0.001 |
| +Lifestyle, clinical, dietary factors4 | 62.8 ± 89 | 62.9 ± 89 | 62.8 ± 89 | 62.3 ± 89 | 63.3 ± 89 | 0.76 |
| ApoA-1,5g/L | ||||||
| Age-adjusted | 1.48 ± 0.01 | 1.49 ± 0.01 | 1.49 ± 0.01 | 1.50 ± 0.01* | 1.49 ± 0.01 | 0.017 |
| +Lifestyle, clinical factors3 | 1.50 ± 0.01 | 1.49 ± 0.01 | 1.49 ± 0.01* | 1.49 ± 0.01 | 1.49 ± 0.01 | 0.41 |
| +Lifestyle, clinical, dietary factors4 | 1.49 ± 0.01 | 1.49 ± 0.01 | 1.49 ± 0.01 | 1.49 ± 0.01 | 1.49 ± 0.01 | 0.67 |
| ApoB100,5g/L | ||||||
| Age-adjusted | 1.002 ± 0.01 | 0.999 ± 0.01 | 0.998 ± 0.01 | 1.005 ± 0.01 | 1.011 ± 0.01 | 0.23 |
| +Lifestyle, clinical factors3 | 0.994 ± 0.01 | 0.999 ± 0.01 | 1.001 ± 0.01 | 1.007 ± 0.01* | 1.006 ± 0.01 | 0.023 |
| +Lifestyle, clinical, dietary factors4 | 1.003 ± 0.01 | 1.003 ± 0.01 | 0.998 ± 0.01 | 0.998 ± 0.01 | 0.986 ± 0.01 | 0.26 |
| Lp(a),5μmol/L | ||||||
| Age-adjusted | 0.42 ± 0.04 | 0.39 ± 0.04* | 0.40 ± 0.04 | 0.38 ± 0.04* | 0.38 ± 0.04 | 0.07 |
| +Lifestyle, clinical factors3 | 0.41 ± 0.04 | 0.39 ± 0.04 | 0.40 ± 0.04 | 0.38 ± 0.04* | 0.39 ± 0.04 | 0.07 |
| +Lifestyle, clinical, dietary factors4 | 0.40 ± 0.04 | 0.39 ± 0.04 | 0.40 ± 0.04 | 0.39 ± 0.04 | 0.40 ± 0.04 | > 0.99 |
Values are mean ± SE. *Different from lowest category of tomato-based food product intake (<1.5 servings/wk), < 0.05. TC, total cholesterol.
-trend for the association between median concentrations of each category of total tomato-based food products and each coronary biomarker.
Adjusted for age (in y), smoking status (never, past, current <15 cigarettes/d, current ≥15 cigarettes/d), exercise (rarely/never, <1 time/wk, 1–3 times/wk, ≥4 times/wk), alcohol intake (rarely/never, 1–3 drinks/mo, 1–6 drinks/wk, and ≥1 drink/d), postmenopausal status (premenopausal, postmenopausal, dubious, and unknown), postmenopausal hormone use (never, former, current), high cholesterol (no, yes), diabetes (no, yes), hypertension (no, yes), and BMI (in kg/m2).
Adjusted for the above covariates plus total energy intake (kJ/d), fruit and vegetable intake (servings/d), fiber intake (g/d), saturated fat intake (g/d), nonsupplemental vitamin C intake (mg/d), and lycopene intake ( g/d).
Geometric means presented, for which we back-transformed the natural logarithms of each biomarker.
Only women consuming ≥10 compared with <1.5 servings/wk of tomato-based food products had significant but clinically modest improvements in TC and the TC:HDL cholesterol ratio after adjustment for dietary factors. Neither LDL nor HDL cholesterol were associated with intake of total tomato-based food products. There was an inverse association between increasing tomato-based food product intake and lower hemoglobin A1c concentrations that emerged in multivariable models adjusting for dietary factors. Compared with women consuming <1.5 servings/wk, those consuming 1.5 to <4, 4 to <7, 7 to <10, and ≥10 servings/wk of tomato-based food products had significantly lower hemoglobin A1c concentrations. For C-reactive protein, an initial positive association for models with lifestyle and clinical factors was reversed upon adjustment for dietary factors, with geometric means of C-reactive protein no longer significantly differing in women consuming ≥10 compared with <1.5 servings/wk. In fully adjusted models, fibrinogen, sICAM-1, creatinine, Lp(a), and apolipoproteins did not differ (all P-trend > 0.05) across categories of total tomato-based food product intake.
Dietary lycopene, strongly correlated with total tomato-based food product intake, confounded the association with each coronary biomarker. In multivariable models without dietary lycopene, the association comparing ≥10 to <1.5 servings/wk of total tomato-based food products was attenuated for the TC:HDL cholesterol ratio (P = 0.71; P-trend = 0.54). For hemoglobin A1c, the association was modestly attenuated comparing ≥10 to <1.5 servings/wk of total tomato-based food products (P = 0.09; P-trend = 0.13).
We also examined the association between increasing tomato-based food product intake and the proportion of women with adverse lipid concentrations, hemoglobin A1c, or C-reactive protein according to established clinical guidelines (Table 3). Additional adjustments for lifestyle and clinical factors did not greatly affect the OR, whereas adjustment for dietary factors tended to result in an OR indicating a lower likelihood of having adverse concentrations of biomarkers, particularly comparing ≥10 to <1.5 servings/wk of tomato-based food product intake. Women consuming ≥10 servings/wk of tomato-based food products were 31 and 40% less likely to have clinically relevant elevations in total and LDL cholesterol, respectively. There was a stronger inverse association between tomato-based food product intake and a lower likelihood of having a hemoglobin A1c ≥ 6%, because women consuming 7 to <10 and ≥10 compared with <1.5 servings/wk of tomato-based food products were 48 and 66% less likely, respectively, to have an elevated hemoglobin A1c ≥ 6%. Had we not included dietary lycopene in our model, women consuming 7 to <10 and ≥10 servings/wk of tomato-based food products were less, but not significantly, likely to have an elevated hemoglobin A1c ≥ 6%.
TABLE 3.
OR (95% CI) of having elevated lipid, hemoglobin A1c, and C-reactive protein concentrations according to categories of total tomato-based food products among 27,261 middle-aged and older women1
| Weekly consumption of total tomato-based food products |
||||||
| <1.5 Servings/wk (n = 4610) | 1.5 to <4 Servings/wk (n = 10,125) | 4 to <7 Servings/wk (n = 8108) | 7 to <10 Servings/wk (n = 3222) | ≥10 Servings/wk (n = 1196) | P-trend2 | |
| TC ≥ 6.21 mmol/L, % | 22.9 | 22.3 | 21.6 | 22.0 | 20.3 | |
| Age-adjusted | 1.00 (ref) | 1.01 (0.93, 1.10) | 0.97 (0.89, 1.06) | 1.01 (0.90, 1.12) | 0.92 (0.78, 1.07) | 0.30 |
| +Lifestyle, clinical factors3 | 1.00 (ref) | 1.04 (0.95, 1.14) | 1.01 (0.92, 1.11) | 1.04 (0.93, 1.17) | 0.89 (0.75, 1.06) | 0.29 |
| +Lifestyle, clinical, dietary factors4 | 1.00 (ref) | 0.99 (0.89, 1.10) | 0.91 (0.79, 1.05) | 0.89 (0.73, 1.08) | 0.69 (0.50, 0.94) | 0.026 |
| LDL cholesterol ≥ 4.14 mmol/L, % | 14.6 | 13.9 | 13.5 | 14.4 | 11.8 | |
| Age-adjusted | 1.00 (ref) | 0.98 (0.89, 1.09) | 0.95 (0.86, 1.06) | 1.04 (0.91, 1.18) | 0.83 (0.69, 1.01) | 0.34 |
| +Lifestyle, clinical factors3 | 1.00 (ref) | 1.03 (0.92, 1.14) | 1.00 (0.89, 1.12) | 1.12 (0.98, 1.29) | 0.84 (0.68, 1.04) | 0.70 |
| +Lifestyle, clinical, dietary factors4 | 1.00 (ref) | 0.96 (0.85, 1.08) | 0.87 (0.73, 1.02) | 0.90 (0.72, 1.14) | 0.60 (0.41, 0.87) | 0.06 |
| HDL cholesterol ≥ 1.55 mmol/L, % | 29.0 | 30.4 | 29.8 | 30.4 | 29.1 | |
| Age-adjusted | 1.00 (ref) | 1.08 (1.00, 1.17) | 1.05 (0.97, 1.14) | 1.09 (0.99, 1.20) | 1.02 (0.89, 1.18) | 0.75 |
| +Lifestyle, clinical factors3 | 1.00 (ref) | 1.02 (0.94, 1.11) | 0.95 (0.87, 1.04) | 0.96 (0.86, 1.07) | 0.91 (0.78, 1.07) | 0.051 |
| +Lifestyle, clinical, dietary factors4 | 1.00 (ref) | 1.07 (0.97, 1.17) | 1.07 (0.93, 1.21) | 1.15 (0.96, 1.38) | 1.20 (0.90, 1.59) | 0.27 |
| TC:HDL cholesterol ratio ≥ 6, % | 10.4 | 9.6 | 9.6 | 9.9 | 10.0 | |
| Age-adjusted | 1.00 (ref) | 0.93 (0.83, 1.05) | 0.93 (0.82, 1.05) | 0.97 (0.83, 1.12) | 0.98 (0.80, 1.22) | 0.99 |
| +Lifestyle, clinical factors3 | 1.00 (ref) | 1.04 (0.92, 1.19) | 1.08 (0.94, 1.23) | 1.12 (0.95, 1.32) | 1.10 (0.87, 1.40) | 0.17 |
| +Lifestyle, clinical, dietary factors4 | 1.00 (ref) | 0.98 (0.85, 1.13) | 0.93 (0.77, 1.13) | 0.90 (0.69, 1.18) | 0.78 (0.51, 1.20) | 0.28 |
| TG ≥ 2.26 mmol/L, % | 18.5 | 18.4 | 18.1 | 18.6 | 18.3 | |
| Age-adjusted | 1.00 (ref) | 1.02 (0.93, 1.12) | 1.00 (0.91, 1.10) | 1.04 (0.93, 1.17) | 1.03 (0.88, 1.22) | 0.66 |
| +Lifestyle, clinical factors3 | 1.00 (ref) | 1.10 (0.99, 1.21) | 1.09 (0.99, 1.21) | 1.09 (0.96, 1.24) | 1.05 (0.88, 1.26) | 0.55 |
| +Lifestyle, clinical, dietary factors4 | 1.00 (ref) | 1.05 (0.94, 1.17) | 0.97 (0.84, 1.13) | 0.91 (0.74, 1.12) | 0.77 (0.56, 1.07) | 0.041 |
| Hemoglobin A1c ≥ 6%, % | 3.5 | 3.1 | 2.9 | 3.2 | 3.4 | |
| Age-adjusted | 1.00 (ref) | 0.91 (0.75, 1.11) | 0.88 (0.71, 1.08) | 0.95 (0.73, 1.23) | 1.04 (0.73, 1.49) | 0.90 |
| +Lifestyle, clinical factors3 | 1.00 (ref) | 1.05 (0.79, 1.39) | 1.08 (0.81, 1.46) | 0.98 (0.67, 1.42) | 0.96 (0.58, 1.61) | 0.84 |
| +Lifestyle, clinical, dietary factors4 | 1.00 (ref) | 0.87 (0.63, 1.19) | 0.72 (0.48, 1.07) | 0.52 (0.30, 0.90) | 0.34 (0.14, 0.80) | 0.007 |
| C-reactive protein ≥ 3 mg/L, % | 36.8 | 36.7 | 37.1 | 37.8 | 37.9 | |
| Age-adjusted | 1.00 (ref) | 1.02 (0.95, 1.10) | 1.04 (0.96, 1.12) | 1.08 (0.98, 1.18) | 1.09 (0.95, 1.24) | 0.06 |
| +Lifestyle, clinical factors3 | 1.00 (ref) | 1.05 (0.97, 1.15) | 1.10 (1.01, 1.20) | 1.13 (1.02, 1.26) | 1.08 (0.93, 1.26) | 0.046 |
| +Lifestyle, clinical, dietary factors4 | 1.00 (ref) | 1.01 (0.92, 1.10) | 1.00 (0.88, 1.14) | 0.98 (0.82, 1.18) | 0.86 (0.65, 1.14) | 0.39 |
The data represent the association between tomato-based food product intake and the OR of women having adverse concentrations of lipids, hemoglobin A1c, or C-reactive protein as defined by established clinical guidelines. TC, total cholesterol.
Test for trend based on ordinal variable containing median value for each category.
Adjusted for age (in y), smoking status (never, past, current <15 cigarettes/d, current ≥15 cigarettes/d), exercise (rarely/never, <1 time/wk, 1–3 times/wk, ≥4 times/wk), alcohol intake (rarely/never, 1–3 drinks/mo, 1–6 drinks/wk, and ≥1 drink/d), postmenopausal status (premenopausal, postmenopausal, dubious, and unknown), postmenopausal hormone use (never, former, current), high cholesterol (no, yes), diabetes (no, yes), hypertension (no, yes), and BMI (in kg/m2).
Adjusted for the above covariates plus total energy intake (kJ/d), fruit and vegetable intake (servings/d), fiber intake (g/d), saturated fat intake (g/d), nonsupplemental vitamin C intake (mg/d), and lycopene intake ( g/d).
Although no individual tomato-based food products had a distinct pattern of association compared with overall tomato-based food product intake, the potential associations of interest noted for TC and hemoglobin A1c appeared to be driven by tomato sauce and pizza intakes, which provide more bioavailable lycopene (data not shown). Further, we found no evidence of effect modification by the presence of baseline diabetes, hypertension, high cholesterol, or obesity, of which any were present in 50.8% of women.
Discussion
After comprehensive adjustments for lifestyle, clinical, and dietary factors, women consuming ≥10 compared with <1.5 servings/wk of tomato-based food products had significant but clinically modest improvements in TC, the TC:HDL cholesterol ratio, and hemoglobin A1c. Significance was driven in part by our large sample size, because the small magnitudes of effect for some biomarkers may be of limited clinical relevance. Further, tomato-based food product intake was unassociated with other lipids and lipoproteins, C-reactive protein, creatinine, and other novel coronary biomarkers. Finally, only the highest level of intake of tomato-based food products was associated with a reduced likelihood of having an elevated concentration of TC, LDL, or hemoglobin A1c, but not other lipids or C-reactive protein.
Tomato-based food products have been examined in the context of broader dietary patterns in relation to various biomarkers and vascular endpoints. Among 5089 nondiabetic, middle-aged men and women in the Multi-Ethnic Study of Atherosclerosis, Nettleton et al. (20) derived a pattern of beans, tomatoes (excluding pizza), and refined grains that was positively associated with sICAM-1 concentrations but not other biomarkers of inflammation and endothelial activation. Alternatively, this dietary pattern was associated with an increased risk of incident diabetes in the Multi-Ethnic Study of Atherosclerosis (21), but it is likely due to the inclusion of refined grains. Considering tomato products in the context of heterogeneous mixtures of food patterns inherently complicates food-specific interpretations of these data. Further, tomato products are often split between disparate categories such as vegetables (e.g., tomatoes) and carbohydrates or pastas (e.g., tomato sauce) (22).
Previously in the WHS, despite a lack of association between dietary lycopene and CVD, women consuming 7–10 or ≥10 servings/wk of tomato-based products had a 30% reduction in risk of CVD (4) but no reduction in risk of diabetes (8). This contrasts with the possible associations noted between tomato products and plasma hemoglobin A1c concentrations or the likelihood of having an elevated hemoglobin A1c concentration in the present study. Additional studies are needed to confirm or refute this association given the lack of accompanying clinical data and putative biologic mechanism.
This study sought to generate new hypotheses related to the potential cardiovascular effects of tomato product intake plus complement the limited but inconsistent trial evidence evaluating various tomato products, often as a tomato extract focused upon its lycopene content, and biomarkers of oxidative stress, lipids, and coronary risk factors. Among 103 healthy volunteers fed either 300 g tomatoes daily for 1 mo compared with a usual diet with tomatoes, concentrations of high sensitivity C-reactive protein, E-selectin, and sICAM-1 were unchanged from baseline in both groups (23). In a crossover study of 77 healthy men and women aged ≥40 y who received 0, 6.5, 15, or 30 mg synthetic lycopene daily for 8 wk, plasma lycopene concentrations rose, whereas biomarkers of oxidative stress, including LDL lag time, LDL oxidation rate, lipid peroxidation, and urinary F2-isoprostanes, decreased nonsignificantly (24). When 8 obese and 8 normal-weight men and women took 30 mg/d of Lyc-o-Mato, a tomato extract that includes lycopene and other nutrients, for 4 wk, plasma lycopene and other carotenoids increased, but C-reactive protein, IL-6, and TNFα did not (25). In contrast, another small Italian crossover study of a tomato-based drink with Lyc-o-Mato reported reductions in TNFα production (26).
To our knowledge, this is the first large-scale study investigating the cross-sectional association between tomato-based food product intake and coronary biomarkers, which have otherwise been examined in smaller clinical studies. However, some limitations temper our conclusions. First, tomato-based food product intake was low in this population of middle-aged and older women, with only 16% of women consuming ≥7 servings/wk of tomato-based food products in contrast to the higher daily doses of tomato products or supplements tested in small clinical studies. Our comparison of “extreme” differences in tomato-based food product intake was actually quite narrow and may have hindered our power to detect more meaningful magnitudes of effect. Differences in TC, the TC:HDL cholesterol ratio, and hemoglobin A1c concentrations between extreme categories of tomato-based food product intake in the WHS were modest, suggesting that subsequent studies should consider a wider spectrum of intake to better evaluate potential dose-response effects. We also noted a U-shaped pattern across intake categories for BMI, hypertension, hypercholesterolemia, and diabetes, suggesting that women may consume more tomato-based food products either in response to or in advance of developing these risk factors.
Second, residual confounding cannot be overlooked, because women consuming greater amounts of tomato products tend to follow healthier dietary patterns that may favorably affect coronary biomarkers. Included dietary factors were particularly strong confounders of the association between tomato-based food products and both absolute concentrations and clinical categories of many biomarkers evaluated; residual confounding by other dietary factors may easily explain the remaining small magnitude of the associations. Third, some misclassification of our food and dietary variables, including tomato-based food products, is possible given our use of a single baseline FFQ assessment, because multiple dietary assessments typically reduce error and generate stronger magnitudes of association (27). Fourth, because this was a cross-sectional study, we cannot draw conclusions about temporality. Finally, generalizability may be limited in this study among predominantly white female health professionals. Although we see no a priori reason for differences, it remains important to frame these results stratified by gender, race/ethnicity, age, and other relevant factors.
This large, cross-sectional study provides important, preliminary insights on tomato-based food products and coronary biomarkers. Tomato-based food products are commonly consumed foods that represent the primary source of dietary lycopene. Women consuming ≥10 compared with <1.5 servings/wk of tomato-based food products had significant improvements in TC, the TC:HDL cholesterol ratio, and hemoglobin A1c after comprehensive multivariable adjustment; however, the clinical relevance of these small magnitudes of effect is questionable given the potential for residual confounding. Consuming ≥10 servings/wk of tomato-based food products reduced the likelihood of having clinically relevant elevations in TC, LDL cholesterol, and hemoglobin A1c. However, no associations were found for other lipids, lipoproteins, C-reactive protein, and other coronary biomarkers. Dietary lycopene was an important confounder of many observed associations, highlighting the need for additional studies that better distinguish any effects of tomato-based food products compared with lycopene on coronary biomarkers and the risk of CVD in a variety of populations.
Acknowledgments
H.D.S. designed the research and analyzed the data; L.W. provided scientific oversight; and H.D.S., L.W., P.M.R., and J.E.B. contributed to the writing of the manuscript. All authors read and approved the final manuscript.
Footnotes
Supported by research grants CA-047988, HL-043851, HL-080467, and HL-099355 from the NIH, Bethesda, MD, plus an investigator-initiated grant from the Tomato Product Wellness Council, Sacramento, CA.
Abbreviations used: CVD, cardiovascular disease; MI, myocardial infarction; sICAM-1, soluble intracellular adhesion molecule-1; TC, total cholesterol; WHS, Women’s Health Study.
Literature Cited
- 1.He FJ, Nowson CA, Lucas M, MacGregor GA. Increased consumption of fruit and vegetables is related to a reduced risk of coronary heart disease: meta-analysis of cohort studies. J Hum Hypertens. 2007;21:717–28 [DOI] [PubMed] [Google Scholar]
- 2.He FJ, Nowson CA, MacGregor GA. Fruit and vegetable consumption and stroke: meta-analysis of cohort studies. Lancet. 2006;367:320–6 [DOI] [PubMed] [Google Scholar]
- 3.Clinton SK. Lycopene: chemistry, biology, and implications for human health and disease. Nutr Rev. 1998;56:35–51 [DOI] [PubMed] [Google Scholar]
- 4.Sesso HD, Liu S, Gaziano JM, Buring JE. Dietary lycopene, tomato-based food products and cardiovascular disease in women. J Nutr. 2003;133:2336–41 [DOI] [PubMed] [Google Scholar]
- 5.Shi J, Le Maguer M. Lycopene in tomatoes: chemical and physical properties affected by food processing. Crit Rev Food Sci Nutr. 2000;40:1–42 [DOI] [PubMed] [Google Scholar]
- 6.Ried K, Fakler P. Protective effect of lycopene on serum cholesterol and blood pressure: meta-analyses of intervention trials. Maturitas. 2011;68:299–310 [DOI] [PubMed] [Google Scholar]
- 7.Hirvonen T, Virtamo J, Korhonen P, Albanes D, Pietinen P. Intake of flavonoids, carotenoids, vitamins C and E, and risk of stroke in male smokers. Stroke. 2000;31:2301–6 [DOI] [PubMed] [Google Scholar]
- 8.Wang L, Liu S, Manson JE, Gaziano JM, Buring JE, Sesso HD. The consumption of lycopene and tomato-based food products is not associated with the risk of type 2 diabetes in women. J Nutr. 2006;136:620–5 [DOI] [PubMed] [Google Scholar]
- 9.Ridker PM, Cook NR, Lee IM, Gordon D, Gaziano JM, Manson JE, Hennekens CH, Buring JE. A randomized trial of low-dose aspirin in the primary prevention of cardiovascular disease in women. N Engl J Med. 2005;352:1293–304 [DOI] [PubMed] [Google Scholar]
- 10.Lee IM, Cook NR, Gaziano JM, Gordon D, Ridker PM, Manson JE, Hennekens CH, Buring JE. Vitamin E in the primary prevention of cardiovascular disease and cancer: the Women's Health Study: a randomized controlled trial. JAMA. 2005;294:56–65 [DOI] [PubMed] [Google Scholar]
- 11.Buring JE, Hennekens CH, for the Women's Health Study Research Group The Women's Health Study: summary of the study design. J Myocardial Ischemia. 1992;4:27–9 [Google Scholar]
- 12.Buring JE, Hennekens CH, for the Women's Health Study Research Group The Women's Health Study: rationale and background. J Myocardial Ischemia. 1992;4:30–40 [Google Scholar]
- 13.Liu S, Manson JE, Lee IM, Cole SR, Hennekens CH, Willett WC, Buring JE. Fruit and vegetable intake and risk of cardiovascular disease: the Women's Health Study. Am J Clin Nutr. 2000;72:922–8 [DOI] [PubMed] [Google Scholar]
- 14.Hennekens CH, Mayrent SL, Willett W. Vitamin A, carotenoids, and retinoids. Cancer. 1986;58:1837–41 [DOI] [PubMed] [Google Scholar]
- 15.Mora S, Cook N, Buring JE, Ridker PM, Lee IM. Physical activity and reduced risk of cardiovascular events: potential mediating mechanisms. Circulation. 2007;116:2110–8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Mora S, Lee IM, Buring JE, Ridker PM. Association of physical activity and body mass index with novel and traditional cardiovascular biomarkers in women. JAMA. 2006;295:1412–9 [DOI] [PubMed] [Google Scholar]
- 17.Ridker PM, Rifai N, Rose L, Buring JE, Cook NR. Comparison of C-reactive protein and low-density lipoprotein cholesterol levels in the prediction of first cardiovascular events. N Engl J Med. 2002;347:1557–65 [DOI] [PubMed] [Google Scholar]
- 18.Executive Summary of The Third Report of The National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) JAMA. 2001;285:2486–97 [DOI] [PubMed] [Google Scholar]
- 19.Pearson TA, Bazzarre TL, Daniels SR, Fair JM, Fortmann SP, Franklin BA, Goldstein LB, Hong Y, Mensah GA, Sallis JF, Jr, et al. American Heart Association guide for improving cardiovascular health at the community level: a statement for public health practitioners, healthcare providers, and health policy makers from the American Heart Association Expert Panel on Population and Prevention Science. Circulation. 2003;107:645–51 [DOI] [PubMed] [Google Scholar]
- 20.Nettleton JA, Steffen LM, Mayer-Davis EJ, Jenny NS, Jiang R, Herrington DM, Jacobs DR., Jr Dietary patterns are associated with biochemical markers of inflammation and endothelial activation in the Multi-Ethnic Study of Atherosclerosis (MESA). Am J Clin Nutr. 2006;83:1369–79 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Nettleton JA, Steffen LM, Ni H, Liu K, Jacobs DR. Dietary patterns and risk of incident type 2 diabetes in the Multi-Ethnic Study of Atherosclerosis (MESA). Diabetes Care. 2008;31:1777–82 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Centritto F, Iacoviello L, di Giuseppe R, De Curtis A, Costanzo S, Zito F, Grioni S, Sieri S, Donati MB, de Gaetano G, et al. Dietary patterns, cardiovascular risk factors and C-reactive protein in a healthy Italian population. Nutr Metab Cardiovasc Dis. 2009;19:697–706 [DOI] [PubMed] [Google Scholar]
- 23.Blum A, Monir M, Khazim K, Peleg A, Blum N. Tomato-rich (Mediterranean) diet does not modify inflammatory markers. Clin Invest Med. 2007;30:E70–4 [DOI] [PubMed] [Google Scholar]
- 24.Devaraj S, Mathur S, Basu A, Aung HH, Vasu VT, Meyers S, Jialal I. A dose-response study on the effects of purified lycopene supplementation on biomarkers of oxidative stress. J Am Coll Nutr. 2008;27:267–73 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Markovits N, Ben Amotz A, Levy Y. The effect of tomato-derived lycopene on low carotenoids and enhanced systemic inflammation and oxidation in severe obesity. Isr Med Assoc J. 2009;11:598–601 [PubMed] [Google Scholar]
- 26.Riso P, Visioli F, Grande S, Guarnieri S, Gardana C, Simonetti P, Porrini M. Effect of a tomato-based drink on markers of inflammation, immunomodulation, and oxidative stress. J Agric Food Chem. 2006;54:2563–6 [DOI] [PubMed] [Google Scholar]
- 27.Giovannucci E, Rimm EB, Liu Y, Stampfer MJ, Willett WC. A prospective study of tomato products, lycopene, and prostate cancer risk. J Natl Cancer Inst. 2002;94:391–8 [DOI] [PubMed] [Google Scholar]