Dietary protein, blood pressure and mortality : the value of repeated measurements

Cardiovascular diseases (CVD) are the main cause of death worldwide. In 2012, about 17.5 million people died from CVD, accounting for 30% of all deaths. High blood pressure (BP) is a major cardiovascular risk factor, which was responsible for 10.4 million deaths in 2013. Diet and lifestyle play an important role in the etiology of hypertension. Maintenance of a desirable body weight, physical activity, and low intake of alcohol and salt are well-known measures to avoid high BP. Whether dietary protein, or more specifically plant and animal protein, could contribute to maintaining a healthy BP is less clear. The association between BP and CVD mortality has been extensively investigated. BP in prospective studies can be analyzed using different approaches, such as single BP (measured at one moment in time), single BP adjusted for regression dilution, average BP, and trajectories of BP. It is not yet clear which of these approaches is to be preferred for CVD risk prediction. This thesis is centered on BP as a major cardiovascular risk factor. In the first part (Chapter 2, 3 and 4), the relation of dietary protein intake with BP level and change was examined. In the second part (Chapter 5 and 6), various approaches for analyzing repeated BP measurements were compared in relation to CVD and all‑cause mortality risk. The final chapter discusses the main findings and their implications. Chapter 2 describes the association of 24-h urinary urea excretion, as a biomarker of total protein intake, with 9-year incidence of hypertension. We analyzed data of ~4000 men and women aged 28–75 years, who participated in the Prevention of Renal and Vascular Endstage Disease (PREVEND) study, a prospective cohort study. BP was measured four times during 1997–2009 and participants were followed for hypertension incidence, defined as BP ≥140/90mmHg or use of antihypertensive medication. Urea excretion was assessed in two consecutive 24-h urine collections at baseline and approximately 4 years later, from which total protein intake was estimated. Protein intake based on 24-h urinary urea excretion was not associated with incident hypertension. Chapter 3 presents findings for long-term total, animal and plant protein intake in relation to 5‑year BP change. Analyses were based on 702 observations of 272 men who participated in the Zutphen Elderly Study. Participants did not use antihypertensive medication and were initially free of CVD. Physical and dietary examinations were performed in 1985, 1990, 1995, and 2000. BP was measured twice at each examination and protein intake was assessed using the cross-check dietary history method. The upper tertiles of plant protein intake were associated with a mean 5‑year change in systolic BP of ‑2.9 mmHg (95% CI: ‑5.6, ‑0.2), compared with the bottom tertile. Total and animal protein intake was not associated with BP. Chapter 4 describes a meta‑analysis of 12 observational studies and 17 randomized controlled trials (RCTs) of dietary protein, including animal and plant protein, in relation to BP. Protein intake in prospective cohort studies was not associated with incident hypertension. For RCTs that used carbohydrate as a control treatment, the pooled BP effect was ‑2.1 mmHg systolic (95% CI: ‑2.9, ‑1.4) for a weighted mean contrast in protein intake of 41 grams per day. There was no differential effect of animal and plant protein on BP. Chapter 5 describes repeated BP measures and their association with CVD and all‑cause mortality and life years lost in two prospective and nearly extinct cohorts of middle-aged men, the Minnesota Business and Professional Men Study (n=261) and the Zutphen Study (n=632). BP was measured annually during 1947–1957 in Minnesota and 1960–1970 in Zutphen. After 10 years of BP measurements, men were followed until death on average 20 years later. Each 25-mmHg increase in average SBP was associated with a 49% to 72% greater CVD mortality risk, 34% to 46% greater all-cause mortality risk and 3 to 4 life years lost. Four systolic BP trajectories were identified, in which mean systolic BP increased by 5 to 49 mmHg in Minnesota and 5 to 20 mmHg in Zutphen between age 50 and 60. In Zutphen, a 2-times greater CVD and all-cause mortality risk and 4 life years lost were observed when comparing trajectories. In Minnesota, associations were twice as strong. BP trajectories were the strongest predictors of CVD mortality and life years lost in Minnesota men, whereas in Zutphen men, the average BP was superior to other measures. Chapter 6 presents findings for average BP and BP trajectories in relation to CVD and all-cause mortality, taking into account antihypertensive medication. A total of 762 participants aged ≥50 years of the Rancho Bernardo Study were examined five times from 1984 to 2002 and monitored for cause‑specific mortality from 2002 to 2013. Each 20‑mmHg increment in average systolic BP was associated with 35% greater CVD mortality and 25% greater all-cause mortality risk. We identified four trajectories for systolic BP for which BP increases ranged from 5 to 12 mmHg between age 60 and 70. In individuals who belonged to the higher trajectories, 2‑3 times greater CVD mortality and 1.5-times greater all-cause mortality risks were observed, compared to those who belonged to the lowest trajectory. Long-term systolic BP trajectories and average systolic BP were both significant predictors of CVD and all-cause mortality. The associations were not modified by antihypertensive medication. As described in Chapter 7, various approaches were used to study the relation between protein intake and BP. Findings from individual studies and a meta-analysis suggest that dietary protein per se does not affect BP within the range of intake generally consumed in the Netherlands. Replacing carbohydrates by protein, however, has a beneficial effect on BP. Moreover, this thesis showed that BP trajectories are not superior to average BP in predicting CVD and all-cause mortality. A few repeated BP measurements, e.g. three or four, are likely to be sufficient for obtaining a reliable average BP and had a similar predictive value for mortality compared to BP trajectories. Therefore, average BP can be considered the most practical tool for estimating mortality risk.