Association of Vegetable Nitrate Intake With Carotid Atherosclerosis and Ischemic Cerebrovascular Disease in Older Women
Background and Purpose—A short-term increase in dietary nitrate (NO3−) improves markers of vascular health via formation of nitric oxide and other bioactive nitrogen oxides. Whether this translates into long-term vascular disease risk reduction has yet to be examined. We investigated the association of vegetable-derived nitrate intake with common carotid artery intima-media thickness (CCA-IMT), plaque severity, and ischemic cerebrovascular disease events in elderly women (n=1226).
Methods—Vegetable nitrate intake, lifestyle factors, and cardiovascular disease risk factors were determined at baseline (1998). CCA-IMT and plaque severity were measured using B-mode carotid ultrasound (2001). Complete ischemic cerebrovascular disease hospitalizations or deaths (events) over 14.5 years (15 032 person-years of follow-up) were obtained from the West Australian Data Linkage System.
Results—Higher vegetable nitrate intake was associated with a lower maximum CCA-IMT (B=−0.015, P=0.002) and lower mean CCA-IMT (B=−0.012, P=0.006). This relationship remained significant after adjustment for lifestyle and cardiovascular risk factors (P≤0.01). Vegetable nitrate intake was not a predictor of plaque severity. In total 186 (15%) women experienced an ischemic cerebrovascular disease event. For every 1 SD (29 mg/d) higher intake of vegetable nitrate, there was an associated 17% lower risk of 14.5-year ischemic cerebrovascular disease events in both unadjusted and fully adjusted models (P=0.02).
Conclusions—Independent of other risk factors, higher vegetable nitrate was associated with a lower CCA-IMT and a lower risk of an ischemic cerebrovascular disease event.
Our understanding of the health impact of dietary nitrate has recently undergone a radical shift. Originally, nitrate was linked with detrimental health outcomes such as cancer, a theory unsupported despite extensive epidemiological research spanning more than 50 years.1 Currently, the potential vascular benefits of dietary nitrate is a major research focus. Dietary nitrate, found predominantly in green leafy vegetables and beetroot, dose-dependently enhances the circulating nitric oxide (NO) pool by increasing the levels of circulating nitrite, NO, and related nitroso compounds.2–5 Dietary nitrate, through the enterosalivary nitrate–nitrite–NO pathway, is now recognized as an important alternate source of NO. NO was originally assumed to be solely produced through the oxygen-dependent l-arginine–NO synthase pathway. NO is a key regulator of vascular homeostasis and integrity, with decreased production and bioavailability of NO implicated in several cardiovascular disorders.6,7 More than 30 clinical trials have now demonstrated, via effects on NO, a reduction in blood pressure or improvement in measures of vascular function with a short-term increase in nitrate intake.8–10 However, epidemiological studies exploring the relationship between nitrate intake and vascular disease risk are scarce.
Common carotid artery intima-media thickness (CCA-IMT) is a surrogate measure of atherosclerosis11 and is associated with increased vascular risk in general and cerebrovascular disease in particular, independent of conventional risk factors.12 Changes in CCA-IMT are used to assess the success of intervention studies in preventing or reducing the progress of atherosclerotic disease. Although specific dietary factors have been associated with CCA-IMT,13 the association of nitrate intake with CCA-IMT has not been studied. Furthermore, the association of nitrate intake with risk of an ischemic cerebrovascular event has yet to be explored.
Thus, the primary aim of this study was to investigate the association between intake of nitrate from vegetables and CCA-IMT, plaque severity and risk of an ischemic cerebrovascular disease event in a population-based study of older women rather than total nitrate intake. Our focus on vegetable-derived nitrate is because most dietary nitrate derives from vegetables.14 In addition, processed meat, another source of nitrate in the diet, is strongly linked to detrimental health effects.15 We hypothesized an inverse association between vegetable-derived nitrate intake and CCA-IMT, plaque severity and risk of an ischemic cerebrovascular disease.
Subjects and Methods
The Human Ethics Committee of the University of Western Australia approved the study, and written informed consents were obtained from all participants. The study complies with the Declaration of Helsinki.
In 1998, 1500 Western Australian women over the age of 70 years were recruited to a 5-year randomized, controlled trial of oral calcium supplements to prevent osteoporotic fractures (ACTRN12615000750583).16 Of these, 99% (n=1485) had complete food frequency questionnaires (Figure I in the online-only Data Supplement). After excluding implausible energy intakes <2100 kJ (500 kcal) or >14 700 kJ (3500 kcal) per day, 1468 (98%) were included. Participants with preexisting atherosclerotic vascular disease and diabetes mellitus were excluded (n=242), leaving n=1226 (82%) for the ischemic cerebrovascular disease analysis. Of these participants, 65% (n=968) had plaque data and 64% (n=954) had CCA-IMT data (Figure I in the online-only Data Supplement). A study of the demographic variables of the recruited women revealed disease burden, and pharmaceutical utilization was similar to whole populations of this age group.17
A baseline assessment was performed in 1998. This included weight, height, medical history, and lifestyle factors (online-only Data Supplement) and completion of a validated Food Frequency Questionnaire (FFQ).
Assessment of Nitrate Consumption
An estimate of nitrate concentration (mg/g) in each of the vegetable items listed in the Cancer Council of Victoria FFQ was obtained using a recently developed comprehensive nitrate content of vegetables database. This database, compiled using a systematic approach, contains 4254 records sourced from 256 references and includes data on 178 vegetables as well as 22 herbs and spices from 56 countries. The median nitrate value (mg/g) for each vegetable in the FFQ was obtained from the database and multiplied with g/d vegetable consumption to determine nitrate intake. Total nitrate intake from vegetables per day was calculated by totalling the nitrate intake from individual vegetables.
An estimate of nitrate concentration (mg/g) in each of the nonvegetable items listed in the Cancer Council of Victoria FFQ was derived using estimates from 3 published sources18–20 (online-only Data Supplement).
Assessment of CCA-IMT and Plaque Severity
Carotid high-resolution B-mode ultrasonography was used to assess common carotid intima-media thickness and focal carotid plaques at year 3 (2001) of the study (online-only Data Supplement).
Assessment of Ischemic Cerebrovascular Disease Event
The first episode of ischemic cerebrovascular hospitalizations or death was retrieved from the Western Australian Data Linkage System for each study participant from their baseline clinic visit date in 1998 until 14.5 years after their baseline visit (online-only Data Supplement).
A protocol for the statistical analysis of the data was established before analysis began. Data were analyzed using IBM SPSS Statistics (version 21; IBM Corp, Armonk, NY) and SAS (version 9.4; SAS Institute Inc, Cary, NC). The relationship between nitrate intake (vegetable total and nonvegetable nitrate) and CCA-IMT (maximum and mean) was examined using unadjusted, age- and energy-adjusted and baseline risk factor–adjusted linear regression. The baseline risk factor adjusted model included baseline age, body mass index, energy intake, alcohol intake, energy expended in physical activity, antihypertensive medication, statin medication, low-dose aspirin medication, organic nitrate medication, history of smoking, and supplementation group. Results are presented as unstandardized B±SE. Vegetable nitrate consumption was then categorized into tertiles for further analysis by ANCOVA with Bonferroni adjustment for multiple comparisons. The relationship between tertiles of nitrate intake and CCA-IMT (maximum and mean) was examined using unadjusted, age- and energy-adjusted and baseline risk factor–adjusted models as before. The relationship between atherosclerotic plaque severity and nitrate intake was examined using binary logistic regression using unadjusted, age- and energy-adjusted and baseline risk factor–adjusted models as before. Cox proportional hazards models were used for ischemic cerebrovascular disease events in unadjusted, age- and energy-adjusted and baseline risk factor–adjusted models as before. We tested for evidence of a linear trend for vegetable nitrate intakes as continuous variables by using the median value for each tertile of vegetable nitrate intake in separate Cox proportional hazards models. Cox proportional hazards assumptions were tested using log–log plots, which were shown to be parallel. Thus, the proportional hazards assumptions were not violated. P values of <0.05 in 2-tailed testing were considered statistically significant.
In a sensitivity analysis, we assessed the potential impact of other possible dietary confounders on CCA-IMT and ischemic cerebrovascular disease events. Daily intakes of total fat (g/d), protein (g/d), carbohydrate (g/d), and fiber (g/d) were investigated in multivariable adjusted Cox proportional hazards models on a variable-by-variable basis. We further explored the relationship of CCA-IMT and plaque severity with ischemic cerebrovascular disease events using Cox regression in unadjusted, age- and energy-adjusted and baseline risk factor–adjusted models as described above. As this current cohort excluded older women with prevalent atherosclerotic vascular disease and diabetes mellitus, we repeated the analysis including older women with prevalent atherosclerotic vascular disease and diabetes mellitus.
Baseline characteristics of the study participants are presented in Table 1. Classification of vegetables in the Cancer Council of Victoria FFQ according to nitrate content is presented in Table I in the online-only Data Supplement. Total nitrate intake from all vegetables was 67±29 mg/d (range: 6–224 mg/d). Total nitrate intake from all foods was 79±31 mg/d (range: 12–231 mg/d). Vegetables accounted for 84% of total nitrate intake (Figure II in the online-only Data Supplement). After vegetables, fruit, and meat were the greatest contributors to total nitrate intake contributing 6% and 3%, respectively. Participants with higher nitrate intake also had higher intakes of energy, total fat, protein, carbohydrate, and fiber (Table 1).
Nitrate Intake, CCA-IMT, and Plaque Severity
Vegetable Nitrate Intake
Vegetable nitrate consumption was linearly inversely associated with maximum CCA-IMT and mean CCA-IMT (Table 2). A 29-mg (≈1 SD) higher vegetable nitrate intake was associated with a 0.015-mm lower maximum CCA-IMT and a 0.012-mm lower mean CCA-IMT. These effects remained significant after adjustment for age and energy intake as well as after adjustment for baseline risk factors (Table 2).
The effect of vegetable nitrate intake was explored further by dividing the participants into tertiles: <53, 53–76, and >76 mg/d vegetable nitrate consumption (Table 2). Participants who consumed >76 mg/d nitrate (top tertile) had a significantly lower CCA-IMT than participants in the bottom tertile with consumption <53 mg/d. This relationship remained significant after adjustment for age and energy intake for maximum CCA-IMT but was slightly attenuated for mean CCA-IMT. A similar relationship was observed after adjustment for baseline risk factors.
Vegetable nitrate intake (odds ratio, 0.994; 95% confidence interval, 0.806–1.227; P=1.0) was not a predictor of plaque severity.
Total and Nonvegetable Nitrate Intake
Total nitrate consumption was linearly inversely associated with maximum CCA-IMT and mean CCA-IMT (Table II in the online-only Data Supplement). This relationship was not observed for nonvegetable nitrate intake (Table II in the online-only Data Supplement).
Neither total nitrate intake (odds ratio, 1.006; 95% confidence interval, 0.810–1.250; P=0.95) nor nonvegetable nitrate intake (odds ratio, 1.148; 95% confidence interval, 0.865; 1.524; P=0.34) were predictors of plaque severity.
Vegetable Nitrate Intake and Ischemic Cerebrovascular Disease Hospitalizations and Deaths
During 15 032 person-years of follow-up, 186 of 1226 (15%) participants had an ischemic cerebrovascular disease event (hospitalization or death). For every 1 SD (29 mg/d) higher intake of vegetable nitrate, there was an associated 17% lower risk of 14.5-year ischemic cerebrovascular disease event in both unadjusted and baseline risk factor adjusted models (Table 3). A similar relationship was observed for total nitrate intake (P<0.05) but not for nonvegetable nitrate intake (P>0.05).
Across tertiles of vegetable nitrate intake, compared with the lowest tertile (<53 mg/d), intake of the highest tertile (>76 mg/d) nitrate from vegetables was associated with a lower risk of an ischemic cerebrovascular disease event (P for trend <0.05 for all models). The multivariable-adjusted cumulative event rate for ischemic cerebrovascular disease according to tertiles of vegetable nitrate intake is presented in the Figure.
In separate multivariable-adjusted analyses that adjusted for individual dietary factors, total fat, protein, carbohydrates, and fiber did not change the interpretation of the association of vegetable nitrate intake with maximum and mean CCA-IMT as well as ischemic cerebrovascular disease events (P<0.05).
In this study of older women, we report an inverse association of vegetable nitrate intake with CCA-IMT and risk of an ischemic cerebrovascular event over 15 years. The relationship remained consistent for total nitrate intake, but not nonvegetable nitrate intake, with CCA-IMT and risk of an ischemic cerebrovascular disease event. These results add weight to the accumulating body of evidence that the vascular benefits associated with a vegetable-rich diet are due, in part, to dietary nitrate.
The observed differences in CCA-IMT, although small, may be clinically important. An increase in CCA-IMT is indicative of arterial wall thickening, a form of atherosclerosis11 and CCA-IMT is a predictor of future cardiovascular events across all age groups.12 Vegetable nitrate intake was associated with a 17% lower risk of an ischemic cerebrovascular disease event per 29 mg/d (1 SD) higher vegetable nitrate intake, found in approximately half a serve (30 g) of leafy green vegetable intake. This association was significant before and after adjustment for lifestyle and other cardiovascular risk factors. Our data suggest a nonlinear relationship between nitrate intake and ischemic cerebrovascular events because no additional risk reduction was observed in those consuming >76 mg/d nitrate (median 100 mg/d) compared with 53 to 76 mg/d nitrate (median 64 mg/d). As little as one serve of nitrate-rich green leafy vegetable per day may provide adequate nitrate intake for ischemic cerebrovascular disease risk reduction.
A possible mechanism for the inverse association of vegetable nitrate intake with CCA-IMT and risk of an ischemic cerebrovascular event is the augmentation of NO status. Through the endogenous nitrate–nitrite–NO pathway, nitrate has the potential to be converted into NO and to form a large and abundant storage pool for this molecule in blood and tissues.21 Recent studies demonstrating beneficial effects on blood pressure, endothelial function, platelet aggregation, ischemia reperfusion injury, and exercise performance after intake of dietary nitrate8,22 are consistent with the proposal that nitrate intake contributes to cardiovascular health.
Approximately 80% of dietary nitrate intake is from vegetables.14 In our study population, vegetables accounted for 84% of total dietary nitrate intake. Eighty five percent of vegetable nitrate intake was derived from 10 vegetables: lettuce, spinach, celery, beetroot, potatoes, cabbage, pumpkin, green beans, broccoli, and carrots. In contrast to other populations (Far Eastern, African, Latin American, and European) where potatoes often contribute well over 10% of total nitrate intake,23 potatoes only provided 8% of total nitrate intake in our population. Nitrate intake varies greatly between individuals and populations with mean intakes for populations estimated to be between 0.4 to 2.6 mg/kg or 31 to 185 mg24 and actual individual daily nitrate intakes ranging from <20 mg to >400 mg.25,26 Our mean intake of nitrate from all food (79 mg/adult per day) fell into this range. It was slightly higher than the World Health organization (WHO) estimate of nitrate intake from food in Australia of 67 mg/adult per day23.
Study Strengths and Limitations
Our study had several strengths. These included the use of a validated diet assessment tool, detailed information on lifestyle, and cardiovascular risk factors as well as the validated measure of CCA-IMT with repeated measurements leading to a high level of precision in measurement of the main outcome. Several potential limitations could be considered. First, a cross-sectional design for the relationship of dietary nitrate with CCA-IMT provides only weak evidence for causality. Second, although the possibility of potential reverse causation (changes in diet because of a disease diagnosis) exists, it is unlikely that individuals would knowingly alter their nitrate (or vegetable) intake on the basis of their IMT, which is an asymptomatic and preclinical marker of atherosclerosis, and its value would, therefore, most likely be unknown to them. Third, high nitrate intake may simply coincide with other lifestyle or dietary patterns that are associated with cardiovascular health. Although we adjusted for multiple lifestyle and cardiovascular risk factors as well as dietary factors in our analysis, residual or unmeasured confounders cannot be ruled out. A causal relationship of nitrate intake with CCA-IMT and ischemic cerebrovascular disease events cannot be established because of the observational nature of the study. Fourth, NO cannot be measured directly and endogenous levels have multifactorial influences.27 There is also no reliable biomarker of nitrate intake. Nitrate levels in plasma, saliva, and urine are influenced by factors including dietary nitrate intake; metabolites of the l-arginine–NOS pathway; bacterial synthesis of nitrate within the gastrointestinal tract; denitrifying liver enzymes; and renal function. Fifth, in this study, carotid ultrasound measures were assessed in 2001. Since then, there have been numerous advances in the evaluation of sonographic characteristics of carotid plaques such as surface irregularity, ulceration and echogenicity that were not available in for this study. Although these newer measures may have provided further insight into how dietary nitrate may affect plaque stability and cerebrovascular events, this would not change our overall findings from the study that dietary nitrate is associated with both established measures of carotid atherosclerosis and long-term cerebrovascular events. Similarly, we did not assess resistive index and we only assessed the common CCA-IMT and not the bifurcation and internal CCA-IMT, which have been suggested as better predictors of future cardiovascular events.28 Given this further studies investigating the association of dietary nitrate on newer measures of carotid atherosclerosis with IMT of the 3 sites are warranted. Finally, the current data are limited to elderly women and needs to be confirmed in men and younger women.
Our study found an association of nitrate intake from vegetables with lower maximum and mean CCA-IMT as well as ischemic cerebrovascular disease events in a cohort of older women. The results are consistent with the proposal that increased nitrate consumption, primarily from vegetables, prevents thickening of the common carotid artery-intima-media, and may play a role in stroke and atherosclerosis prevention.
We wish to thank the staff at the WA Data Linkage Branch, Hospital Morbidity Data Collection and Registry of Births, Deaths, and Marriages for their work on providing the data for this study.
Sources of Funding
This study was supported by Healthway Health Promotion Foundation of Western Australia and National Health and Medical Research Council of Australia. Salary of Dr Lewis is supported by a National Health and Medical Research Council of Australia Career Development Fellowship. Salary of Dr Ivey is supported by a National Health and Medical Research Council of Australia Early Career Fellowship. Salary of Dr Hodgson is supported by a National Health and Medical Research Council of Australia Senior Research Fellowship.
Dr Lundberg is a named coinventor on patent applications relating to medical uses of nitrate- and nitrite salts. The other authors report no conflicts.
↵The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.117.016844/-/DC1.
- Received January 29, 2017.
- Revision received April 18, 2017.
- Accepted May 8, 2017.
- © 2017 American Heart Association, Inc.
- Speijers G,
- Brandt PAVD
- Petersson J,
- Carlström M,
- Schreiber O,
- Phillipson M,
- Christoffersson G,
- Jägare A,
- et al
- Bondonno CP,
- Croft KD,
- Ward N,
- Considine MJ,
- Hodgson JM
- Lorenz MW,
- von Kegler S,
- Steinmetz H,
- Markus HS,
- Sitzer M
- Ivey KL,
- Lewis JR,
- Hodgson JM,
- Zhu K,
- Dhaliwal SS,
- Thompson PL,
- et al
- Hord NG,
- Tang Y,
- Bryan NS
- Larsson SC,
- Orsini N
- 18.↵Food Standards Australia New Zealand. Survey of nitrates and nitrites in food and beverages in australia. Canberra ACT, 2013. https://www.foodstandards.gov.au/science/surveillance/documents/Survey%20of%20nitrates%20and%20nitrites%20Attach%201.pdf. Accessed November 12, 2015.
- Inoue-Choi M,
- Virk-Baker MK,
- Aschebrook-Kilfoy B,
- Cross AJ,
- Subar AF,
- Thompson FE,
- et al
- Lundberg JO,
- Carlström M,
- Larsen FJ,
- Weitzberg E
- Hambridge T
- Naqvi TZ,
- Lee MS