Arterial stiffness

Strokes and coronary artery disease are the leading causes of death world wide.1 Atherosclerosis is the cause of much of this epidemic, but there is an additional form of arterial disease that contributes greatly to currently high levels of cardiovascular mortality and morbidity. Arteriosclerosis is a disease of the arterial media, rather than the intima, that increases the stiffness of the arterial system and thus impairs one of the important functions of the aorta and conduit arteries—namely, the conversion of pulsatile to continuous blood flow.2 In health, the arterial system is highly distensible ensuring that most tissues receive near-steady flow with no exposure to peak systolic pressures; this mechanism is so efficient that there is almost no drop in peripheral mean arterial pressure compared with ascending aortic pressure. The importance of this is most apparent when it is lost. Increased arterial stiffness results in higher ascending aortic systolic and pulse pressure as a result of loss of the “cushioning effect” of the aorta. This effect may be compounded by the theoretical earlier return of reflected arterial pressure waves from the peripheral vasculature, although the clinical importance of wave reflection to flow is contentious.3 4 There is no doubt, however, that the result of arterial stiffening is systolic hypertension, increased left ventricular after load and reduced diastolic perfusion of the coronary arteries. These factors expose the cerebral micro- and macrocirculations to high-pressure fluctuations and increased cyclic stress, leading to an increased risk of stroke, and causing left ventricular hypertrophy (LVH) and fibrosis predisposing to heart failure and arrhythmias. This principle also applies to other vascular beds, including the renal circulation. Arterial stiffness is independently associated with cardiovascular mortality in patients with hypertension, diabetes and chronic kidney disease (CKD) as well as in previously healthy subjects.5

The increase in arterial stiffness that occurs with age in Western populations appears to be a result of prolonged exposure to environmental factors rather than an inevitable age-related effect. Many studies of populations in Africa and China living hunter-gatherer or subsistence-farming lifestyles have shown that in such communities blood pressure is low and systolic blood pressure does not rise with age. Systolic pressure only increases with advancing age when such populations migrate to urban areas and assume “Western” lifestyles.6

We suggest that an important mechanism for the development of arterial stiffness in societies that have become affluent is exposure of the vascular system to high levels of phosphate. This is caused by the consumption of a phosphate-rich diet and exacerbated by a high prevalence of CKD, which impairs phosphate excretion. Animal protein is rich in bioavailable phosphate and the use of phosphate-containing additives such as phosphoric acid (E338), used as an acidulate and preservative in many foods, including beer, cheese, jam and especially cola soft drinks, is widespread.7 Although “primitive” diets based on grains and legumes often have a high phosphate content they provide little bioavailable phosphate as most of this occurs in the form of non-absorbable phytates. Effectively, these diets result in prolonged eating of low-phosphate foods rather than the “binge eating” of high-phosphate foods associated with large increases in post-prandial serum phosphate levels, common in today’s fast-food society. This mechanism in no way underestimates the contribution of a high sodium intake to the development of hypertension. Indeed, by reducing arterial distensibility with age, phosphate exposure would exacerbate the effect of sodium loading.

Lessons from chronic kidney disease

Almost 15% of the US population have evidence of CKD and the extreme risk of cardiovascular disease associated with this condition has been well described, with a graded inverse relationship to renal function.8 Arterial stiffness is already raised in early CKD9 and is a very powerful independent predictor of mortality in end-stage CKD, suggesting that it has a direct role in the causation of adverse cardiovascular events.5 Arterial stiffness and calcification in CKD are highly correlated and appear to be driven by the characteristic renal mineral bone disorder (MBD-CKD) comprising hyperphosphataemia, hypocalcaemia and hyperparathyroidism together with reduced bone mineral density.10 The presence of MBD-CKD is strongly associated with the extent of arterial calcification, a powerful predictor of mortality.10 Several studies have shown that MBD-CKD is related to adverse outcomes in patients undergoing dialysis.10 In the largest of these involving over 40 000 patients, the population-attributable risk for disorders of mineral metabolism was 17.5%, owing largely to the high prevalence of hyperphosphataemia.11 Thus it would appear that serum phosphate is a significant contributor to the very high cardiovascular mortality seen in patients receiving regular dialysis.

Phosphate as a cardiovascular risk factor in the general population

Serum phosphate (at values within the normal reference range) has also been shown to be associated with cardiovascular mortality in the general population in two recent independent studies.12 13 In a prospective analysis of over 3000 Framingham offspring study participants free of overt cardiovascular disease and with normal renal function, higher serum phosphate levels were associated with increased cardiovascular risk in a continuous graded relationship.12 A similar association between higher levels of serum phosphate and the risk of death and cardiovascular events has also been demonstrated in survivors of myocardial infarction.13

Many cross-sectional studies have reported an association between low bone mineral density and cardiovascular morbidity and mortality in the general population.14 There is an inverse relationship between bone mineral content and vascular calcification, and several longitudinal studies have shown that people with the greatest rate of bone loss demonstrate the fastest progression of vascular calcification.14 In simple terms, it would appear that affected subjects, like patients with CKD, are characterised by a redistribution of calcium and phosphate from the skeleton into the vasculature with an associated increase in arterial stiffness.15

The determinants of serum phosphate concentration within apparently healthy populations are unclear but both diet and minor reductions in renal function may be important. In people with normal renal function, normal phosphate and calcium levels are maintained despite wide variation in dietary intake by modulation of vitamin D and parathyroid hormone (PTH).16 Close control of calcium and phosphate metabolism is an important aspect of renal function that is disturbed even with small reductions in glomerular filtration rate,15 initially with lower levels of vitamin D and higher levels of PTH keeping both calcium and phosphate levels within the normal range. This suggests that the concepts that humans are born with “surplus” renal function and that the decline in glomerular filtration rate with age is harmless may well be incorrect. Raised serum phosphate, calcium and PTH and vitamin D deficiency are all associated with increased all-cause and cardiovascular mortality.17 The mortality risk for phosphate appears greater than with these other parameters, however,17 suggesting a primary role for phosphate in the development of this pathophysiology. This is supported by the observation that phosphate, even within the normal range, remains a cardiovascular risk factor in patients with normal renal function.12 13

By what mechanism might phosphate cause increased arterial stiffness?

The cell biology of increasing vascular stiffness and calcification is complex and phosphate may have multiple influences. Vascular smooth muscle cells (VSMCs) and osteoblasts derive from a common mesenchymal precursor cell and it appears that VSMCs retain their ability to mineralise. Core binding factor α-1 (cbfa-1) is likely to be the switch that turns mesenchymal cells into osteoblasts and this protein is upregulated by exposure of VSMCs to phosphate.15 In the media of calcified arteries, cbfa-1 is highly expressed in association with osteopontin and collagen type 1, suggesting that cbfa-1 may lead to de-differentiation of VSMCs to osteoblast-like cells.15 Secondly, when VSMCs are exposed to phosphate a calcium phosphate precipitate forms in association with the extracellular matrix.18 This process is completely inhibited by antagonism of the sodium phosphate co-transporter Pit-1, suggesting that calcification is an active cellular process dependent on phosphate uptake.18 Thirdly, calcium and phosphate induce VSMC death and apoptotic body release (associated with inflammation and calcification), as well as matrix vesicle release from living cells.18 Vesicles released by VSMCs after prolonged exposure to calcium and phosphate contain preformed calcium phosphate-apatite and calcify intensively. Finally, increasing phosphate levels suppress vitamin D synthesis and increase PTH secretion.10 Lower levels of vitamin D are powerfully associated with increased cardiovascular risk.10 Hypovitaminosis D is also associated with increased arterial calcification,19 decreased cardiac contractility,20 dilated cardiomyopathy,21 upregulation of the renin–angiotensin axis, hypertension and LVH.21 22 23 PTH receptors are found on VSMCs and hyperparathyroidism is also strongly associated with hypertension,24 increased arterial stiffness,24 LVH, cardiac fibrosis,25 impaired cardiac contractility26 and impaired endothelial function.27

Lowering serum phosphate by either dietary or pharmacological means may well improve arterial stiffness by any one of the mechanisms mentioned, or by a combination of these, or indeed by other mechanisms not yet discovered.

By lowering phosphate exposure, can we reduce arterial stiffness and cardiovascular mortality?

We hypothesise that the control of phosphate intake will slow down the progression of arterial stiffness that occurs in affluent populations, reducing the development of systolic hypertension and LVH with major beneficial effects upon cardiovascular mortality and morbidity. There is encouraging information from patients with CKD to support this. Hyperphosphataemia in CKD is commonly treated with oral phosphate-binding agents. Until recently the standard agents have been calcium-based (calcium carbonate/acetate), which have exerted effects related not only to phosphate lowering but also to calcium loading. Lowering serum phosphate with new non-calcium-containing agents such as sevelamer hydrochloride reduces the progression of arterial calcification in patients with CKD undergoing dialysis.14 It also appears to attenuate the progression of arterial stiffness compared with calcium-based phosphate binders.28 No effect on cardiovascular mortality has been shown but trials to date have been underpowered for this end point.

How can this hypothesis be tested?

Attempting to modify dietary behaviour as part of a large randomised trial is a notoriously difficult undertaking and will invariably have multiple effects on both micro- and macronutrients. Observational studies of diet, such as vegetarian versus omnivorous diets, particularly between two different populations, are full of potential confounders, perhaps most notably fat and salt intake, and any associations between diet and outcomes do not necessarily reflect a common pathway.29 Currently available phosphate binders have to be taken with meals and all have side effects. Possible future treatments to reduce phosphate exposure include the use of niacin and related compounds,30 which inhibit active gut phosphate absorption. There are some concerns, however, about long-term efficacy, tolerability and safety.30 Thus like the cholesterol story, which was only accepted after demonstration of the prognostic benefits of statins, the proof of our hypothesis will have to wait until an effective and tolerable means of lowering dietary phosphate consumption or absorption is available. In the meantime, studies examining the effect of existing non-calcium-based phosphate binders on surrogate end points such as arterial stiffness should yield important information. Patients with early-stage CKD, an important public health problem because of their vulnerability to cardiovascular disease, might be a useful group in which to test such a hypothesis.