The left atrium in hypertension

As the proportion of elderly in our communities around the world increases, atrial fibrillation, with its consequent risk of stroke and heart failure, becomes an increasingly important complication of hypertension. The ability of the left atrium to compensate for increasing left ventricular filling pressure is important in terms of both the development of atrial fibrillation and its complications. Hypertensives carry approximately one and a half times the risk of developing atrial fibrillation compared to their normotensive peers [1]. As a highly prevalent risk factor in the general community, hypertension accounts for more cases of atrial fibrillation than any other risk factor, taking into account age [2]. In a 16-year follow-up of hypertensive patients, those who developed paroxysmal atrial fibrillation carried a 2.7% annual risk of stroke. When atrial fibrillation became chronic, 4.6% of patients per year had stroke [3]. In the light of this increasingly important community and personal burden, it is important to understand the factors that determine why individual patients with hypertension progress from sinus rhythm to atrial fibrillation. For the most part, a relatively straightforward haemodynamic explanation appears to make sense. Patients with hypertension are likely to develop left ventricular hypertrophy (LVH). This is a means of normalizing wall stress in the ventricle. For genetic, environmental and probably chronological reasons, LVH sometimes manifests as concentric hypertrophy, and in other patients as eccentric. The former occurs when increased left ventricular mass is associated with high relative wall thickness. The latter occurs when it is not and the ventricle is dilated. LVH is generally associated with stiffening of the ventricle, increased left ventricular end-diastolic pressure and reduced diastolic filling properties. These are both active processes related to cardiomyocyte hypertrophy and passive due to changing properties of the intracellular matrix. The left atrium is faced with the need to cope with higher filling pressures increasing the forces opposing diastolic filling. We tend to think of the Franck–Starling mechanism as being a ventricular phenomenon but it also occurs in the atrium. Consequently, the atrium dilates to increase left atrial size and to improve its ejection performance. At some point, which varies between individual patients, the electromechanical consequences of atrial dilatation become adverse and atrial fibrillation occurs. Although this straightforward haemodynamic explanation provides a plausible account of the observed relationships between changes in the left ventricle and left atrium in hypertension, it does not explain everything. For example, the relationship between blood pressure and LVH and its consequences is poor, although improved when pressure is measured by ambulatory recording rather than in the clinic. The reason why the changes in the ventricle or the atrium remain compensated throughout life in some patients but progress to clinical sequelae in others is also unclear. It also remains to be defined why, at the same level of ventricular overload, some patients develop specific geometric patterns and others do not, and what the pathophysiological consequence of this evolution is. In the study by Verdecchia et al. [3] involving almost 2500 hypertensives, left atrial dimension predicted future atrial fibrillation during follow-up. Other univariate predictors included age, blood pressure (especially when measured with ambulatory recording) and left ventricular mass determined echocardiographically. The independent predictors of a single episode of atrial fibrillation were age and left ventricular mass. However, in those patients who progressed to chronic atrial fibrillation, left atrial dimensions became an independent predictor. The conclusion was that, as in atrial fibrillation of other aetiologies, left atrial dimension is a major determinant of whether atrial fibrillation returns and becomes chronic after the initial bout. However, signal-to-noise ratio is the bane of any quantitative M-mode or two-dimensional echocardiographic technique, and another explanation may be that there was more variance in the measurement of left atrial dimension than in the other statistically independent predictors. A number of studies have examined the relationship between LVH and left atrial dimensions in hypertensive populations. An interesting study by Cioffi et al. [4] is published in this issue of the journal. The authors confirm the results obtained in an earlier, smaller study [5], carried out in patients with newly diagnosed untreated hypertension, and show that left atrial size is, amongst other things, related to left ventricular mass. Body mass index and left ventricular mass were the best predictors of left atrial size. Other univariate associations with left atrial size included weight, left ventricular wall thickness, 24-h pulse pressure in hypertensives and a similar range of variables in normotensives. Left atrial volume was related to left ventricular mass independent of left ventricular geometry. It should be noted that left ventricular geometry was categorized as concentric (one-third of patients) and eccentric (two-thirds of patients) irrespective of whether hypertrophy was present or not. Conventionally, the terms eccentric and concentric have been applied to hypertrophied left ventricles only. On the whole, concentric hypertrophy carried worse consequences not only in terms of left atrial size, but also with respect to left ventricular mass and reduced mid-wall fractional shortening compared to patients with an eccentric pattern. This was a very specific patient group with hypertension, although the same authors have previously shown similar relationships between left atrial size and left ventricular mass in patients with aortic valve disease [6]. Other reports have drawn slightly different conclusions. In the LIFE Study [7], increased left atrial dimensions tended to be more associated with eccentric than concentric hypertrophy. This patient group was different in that all of the patients in the LIFE Study not only had hypertension, but also LVH by electrocardiographic criteria. Some of the exclusion criteria also differed. However, another study in children and adolescents drew similar conclusions to those made in the LIFE Study [8]. In these hypertensive subjects with an average age of 14 years, left atrial size was strongly determined by body mass. There was an independent relationship between left atrial size and body mass index, height, systolic blood pressure and, in this case, left ventricular geometry, with subjects with eccentric hypertrophy more likely to have increased atrial dimensions. An important factor linking eccentric hypertrophy with hypertension may be obesity, particularly in younger subjects. In comparing studies in this area, there is a general problem resulting from differences in methodologies. There are many definitions and cut-offs of LVH. In the study by Cioffi et al. [4], left ventricular mass to the power of 2.7 ≥ 51 g/m2.7 was used. Left atrial volume was determined by the area length method and Mannings formula was used to measure left atrial ejection fraction. These methods are generally sound. However, the most appropriate index for right atrial volume is unclear. Should it be adjusted for body size and other confounding factors, and what is the appropriate definition of high atrial performance? In this case the 75th percentile of the present population was used. It would be preferable to use a different population to define a ‘normal’ range. What conclusions can be drawn from the above? These results, together with those obtained in other similar studies, emphasize the importance of long-term control of blood pressure not only in reducing LVH and its consequences for the ischaemic myocardium, but also in preventing a contribution to later development of atrial fibrillation from a progressive increase in left atrial size. It may be concluded that the differentiation of LV geometry into concentric and eccentric patterns was not useful in the study by Cioffi et al. [4], and that their data, together with those obtained from cardiac imaging in hypertensive patients, are best treated as continuous variables across normal and hypertensive populations. The strong relationship between cardiac chamber size and body size raises an interesting hypothesis that control of obesity might be an important strategy for reducing left atrial dimensions and its consequences, particularly atrial fibrillation and stroke [8–11].