Nitrate INFORMER Water Study

Overview

Nitrate is a controversial component of vegetables, meat, and drinking water. The now well-established benefits of nitrate, through the enterosalivary nitrate-nitrite-nitric oxide (NO) pathway, on cardiovascular risk factors and long-term cardiovascular disease risk are tarnished by a continuing concern about a link between nitrate ingestion and cancer. This can result in misguided advice to avoid consumption of high-nitrate leafy green vegetables by both the media and the scientific literature. A recent media headline stated, "Cancer alert over rocket: trendy salad leaves exceed safe levels of carcinogenic nitrates in one in every ten samples". One scientific review stated, "the presence of nitrate in vegetables, as in water and generally in other foods, is a serious threat to man's health". Controversy in the literature, and gaps in the knowledge are leading to confusing messages around vegetables that may play a critical role in cardiovascular health. The major dietary sources of nitrate are vegetables, meat, and drinking water. Source of nitrate could be a crucial factor determining whether the consumption of nitrate is linked with beneficial (such as improving cardiovascular health) versus harmful (N-nitrosamine formation) effects. For example, unlike meat and water-derived nitrate, vegetables contain high levels of vitamin C and/or polyphenols that may inhibit the production of N-nitrosamines. So far, no study has investigated the formation of N-nitrosamines after consumption of these different sources in humans. This study will compare N-nitrosamine formation after intake of water with and without added nitrate.

Study design A crossover study design will be used with a 1-week washout period between interventions. Participants diet for the day before and the day of the intervention will be standardised (food choices on visit 1 will be matched for visit 2) as follows: For the day before the study visit: * Breakfast: white bread toast with eggs, cheese, peanut butter or jam. * Lunch will be provided comprising pumpkin soup and white bread * Dinner: a low nitrate vegetarian meal * All water will be provided * Participants will be asked to refrain from drinking any coffee or caffeine containing beverages 17 and any alcoholic beverage 18 and do any exercise 24 hours prior to their study visit. For the day of the study visit: * Participants will arrive at the study unit fasting * Breakfast will comprise white bread toast with cheese, peanut butter or jam together with the first water intervention * Lunch will comprise pumpkin soup and white bread together with the second water intervention * Participants will be provided with a standardised dinner: a low nitrate vegetarian meal * All water will be provided * Participants will be asked to refrain from drinking any coffee or caffeine containing beverages 17 and any alcoholic beverage 18 and do any exercise 24 hours prior to their study visit. Interventions • Water with added nitrate: 137 mg sodium nitrate or 164 mg potassium nitrate (100mg of nitrate) in a 500 mL water bottle (200 mg/L) provided twice, one with breakfast and one with lunch. This is equivalent to 200mg nitrate in total. In Australia, guidelines published in 2011 recommend that levels less than 50 mg/L are safe to drink, levels of 50 - 100 mg/L are safe to drink for all adults including pregnant women but not infants less than 3 months of age, drinking water with nitrate concentrations greater than 100 mg/L is not recommended. • Water without added nitrate: 137 mg sodium chloride or 164 mg potassium chloride in a 500 mL water bottle provided twice, one with breakfast and one with lunch. Sequence generation. The sequence of intervention allocation will be generated via block randomisation using computer-generated random numbers. Random block sizes of 2 and 4 will be used. Concealment mechanism. Fifty randomly generated sequences of the interventions for each of the study participants will be printed on separate pieces of paper and sealed in opaque envelopes, numbered 1-50, by a study investigator not involved in performing the intervention, the data collection or the data analysis. Implementation. Once a participant is deemed eligible and enrolled in the study, the study coordinator will contact the study investigator responsible for randomisation and intervention allocation to obtain the next available envelope and randomly generated intervention sequence. Blinding. Given the nature of the interventions, participants, and the investigators responsible for delivering the interventions will be unblinded throughout the trial. However, all researchers performing the laboratory analyses and data analyses will be blinded to the interventions that the participants received until after the data analysis has been performed. Participant timeline. Prior to the first clinic visit, each participant will complete a food frequency questionnaire (FFQ) to assess background habitual diet. At each clinic visit (Figure 1), a baseline urine and faecal stool sample will be collected for measurement of N-nitrosamines. Prior to the intervention participants will be instructed to void their bladder into baseline urine collection container. After the first intervention all urine for the subsequent 24-hour period will be collected for measurement of N-nitrosamines, nitrate, and nitrite. A faecal stool will be collected for measurement of N-nitrosamines, nitrate, and nitrite. Assessments Urine and faecal stool collection. A urine sample will be collected at baseline, and from intervention up until 24 hours post-intervention. For the baseline sample, participants will be provided with a sterilized container and instructions to discard the first urine sample of the day and then collect all urine until the intervention which will be brought into the clinic. Participants will be instructed to drink 250 ml water on waking. For the 24-hour urine samples, participants will be provided with sterilized containers and instructions to collect all urine until 24 hours post intervention. Urine aliquots will be frozen at -80°C until analysis. A stool sample will be collected at baseline and for the 24 hour period post the first intervention. Participants will be provided with instructions and a stool sample collection pack (collection bags, cable ties, large zip lock bags, freezer ice blocks and a designated cooler bag for transport). Collected stool samples will be weighed and frozen at -80°C until analysis. N-nitrosamines. Nitrosamines in the interventions, urine and stool samples will be measured by gas chromatography mass spectrometry (GCMS). Volatile nitrosamines will be extracted with dichloromethane and are well separated by gas chromatography. Identification will be by retention time and unique molecular ions for each of five common nitrosamines. Quantitation will be made by using deuterium labelled internal standards, N-nirtrosodimethylamine-D6; N-nitrosodiethylamine D10; N-nitrosomorpholine D8 (Cambridge Isotopes) and N-nitrosopiperidine D10 (Toronto Research Chemicals). The most commonly detected nitrosamines are NDMA and N-Nitrosopiperidine (NPIP). Control experiments in which water stored in the same containers as those used for sample collection and storage will be performed to test for any leaching of volatile nitrosamines. Sample size and power Sample size is based on a crossover design and the primary outcome of N-nitrosamines, specifically N-Nitrosodimethylamine (NDMA) in urine. At α=0.05, 2 measures per subject (baseline and post) and a correlation of ρ=0.6 between measures, 25 participants will provide 80% power to detect a 0.6 SD increase in NDMA concentration. To allow for a 10% withdrawal rate, we plan to recruit 25 participants into the study. A rolling recruitment of participants will be performed until the sample size is reached. Statistical methods Statistical analyses will be performed using IBM SPSS Statistics for Windows, version 25 (IBM) and STATA/IC 17.0 (StataCorp LLC). Descriptive statistics of normally distributed continuous variables will be expressed as mean (± standard deviation, SD), non-normally distributed continuous variables as median (interquartile range, IQR) and categorical variables as number (proportion, %). Data will be assessed for outliers and normality prior to analysis. Non-normally distributed data will be log transformed if necessary. Treatment effects for outcomes will be obtained using linear mixed models including baseline measurements, treatment order, period, and time (as a categorical variable) as fixed effects. We will also include a treatment X period interaction term to assess for possible treatment-period interactions and will separately assess for carryover effects with dummy variables to indicate the previous treatment. Treatment effects for outcomes with a single post-intervention measurement will be obtained using linear mixed models including baseline measurements, treatment order, and period as predictors. The subject ID number will be included as a random intercept in each model. An overall 2-sided type-1 error rate of P\<0.05 will be used to assess statistical significance for all hypothesis testing.