Left atrial function and mortality in the oldest old

Abstract

Background

Previous studies demonstrated that left atrium (LA) size is associated with mortality in an elderly population. It remains unclear whether indices of LA function including reservoir, conduit, or booster elements of LA function provide incremental prognostic information.

Hypothesis

Echocardiographic measures of the various parameters of LA function would predict 5‐year mortality in a community‐dwelling population of 85 to 86 year olds independently of LA volume.

Methods

Subjects ages 85 to 86 years old underwent home echocardiography. LA volumes were assessed by the biplane Simpson's method from apical views using measurements of phasic volumes and functions of the LA, including LA expansion index. LA passive and active emptying fractions were assessed. Survival status at 5‐year follow‐up was assessed.

Results

Two hundred eighty‐two subjects were included, of whom 87 (31%) had died at follow‐up. Survival of the subjects in the lowest quartile of the LA expansion index as well as LA active filling index was significantly lower. When measurements of LA volume index were added to the model, the relationship between survival and indices of LA function remained significant.

Conclusions

This study demonstrated that elderly subjects aged 85 to 86 years with significantly impaired LA function had increased 5‐year mortality independently of indices of LA volume.

1. INTRODUCTION

People over the age of 85 years are the world's most rapidly growing age group and provide an increasing challenge for cardiovascular care given the relatively high frequency of cardiac death in this population.1, 2 Previous studies that have utilized echocardiography in elderly patients to examine prognosis including a broad range of ages with relatively few patients over the age of 80 years.3, 4, 5 In addition, most existing studies have been performed in the hospital or clinic setting, possibly contributing to a biased study population in this elderly age group. The introduction of portable echocardiography machines has made it possible to study patients in the home, and therefore a more representative population of the oldest old.

Left atrial (LA) size increases with age, and previous studies have demonstrated that LA diameter and volume are associated with mortality in an elderly population.6, 7 Indices of LA function may provide incremental prognostic information to that obtained by LA size; however, data regarding prognostic value of LA function in a community‐dwelling elderly population are limited.8, 9 In addition, it remains unclear whether the reservoir, conduit, or booster elements of LA function are the most important prognostically. These distinct atrial functions can be assessed separately using echocardiography. Therefore, the aim of this study was to examine the association between echocardiographic measures of the various parameters of LA function performed at the subject's home and 5‐year mortality in an age‐homogenous, community‐dwelling population born in 1920 and 1921.

2. METHODS

Subjects were recruited from the Jerusalem Longitudinal Cohort Study that was initiated in 1990 and has followed an age homogenous representative cohort of West Jerusalem residents born between June 1920 and May 1921. The methodology has been described elsewhere in detail.10, 11 The present study examines data from the third most recent phase of data collection, which took place during 2005 and 2006. The institutional ethics committee of the Hadassah Hebrew University Medical Center approved the study design, and written informed consent was obtained from all participants.

Subjects identified from the electoral register were randomly chosen from the total sample of people born in 1920 and 1921 and living in Jerusalem in 2005. As reported previously, we performed an examination of death certificates and hospital admission records 3 years following the initiation of the study, and compared the study group to other subjects of the sample frame in Jerusalem.10, 11 Subjects of the study group, those who declined to participate, and those baseline cohort members not enrolled had near identical mortality and disease‐specific hospital morbidity, demonstrating the representative nature of the initial study group. Echocardiography was performed in 502 randomly selected subjects, evenly distributed between new recruits and subjects participating from previous phases. Survival status at the 5‐year follow‐up was assessed via the centralized Ministry of Interior database. Follow‐up was available for all study subjects.

Diagnosis of ischemic heart disease (IHD) was based on a history of hospitalization for acute coronary syndrome, coronary catheterization with evidence of significant coronary artery disease, myocardial infarction on electrocardiogram, a history typical for angina pectoris on exertion, or previous coronary artery bypass grafting surgery. Subjects were studied at home, and therefore by definition, patients with unstable angina were not included. Hypertension was assessed by the examining study physician and generally was defined as treatment with antihypertensive medications or subjects' self‐reporting. Hyperlipidemia was defined as use of cholesterol‐lowering medications. Diabetes mellitus was a composite of hypoglycemic medications, personal history, or a medical record diagnosis. Congestive heart failure (CHF) was based on hospital discharge diagnosis and according to examining research physician diagnosis at the time of examination at home. As we have previously demonstrated the influence of marital status on prognosis, this variable was included in the analysis.12

Subjects had standard 2‐dimensional (2‐D) and Doppler echocardiography at their place of residence with a portable echocardiograph (Vivid I; GE Healthcare, Haifa, Israel). Patients with atrial fibrillation were excluded from the study. All subjects underwent 2‐D and Doppler echocardiography with M‐mode measurements of the interventricular septum, posterior wall, and left ventricular (LV) end‐systolic and end‐diastolic diameters according to the recommendations of the European Association of Echocardiography/American Society of Echocardiography.13 Measurements were performed for 3 consecutive cardiac cycles and averaged. Subjects' height and weight at the time of the study were recorded and body surface area calculated. LV mass was calculated according to a necropsy validated formula of LV mass (g) = 0.8 × (1.04 × ((septal thickness + LV internal diameter + posterior wall thickness)3 − (LV internal diameter)3)) + 0.6 and indexed to body surface area.14

Ejection fraction (EF) was calculated by averaging measurements of end‐diastolic and end‐systolic volumes from apical views using the biplane Simpson's method for 3 consecutive beats. Diastolic parameters were measured from the apical 4‐chamber view using pulsed‐wave Doppler at the level of the mitral annulus and tissue Doppler imaging of the septal and lateral myocardial walls and included early (E) and late (A) transmitral flow velocities, the ratio of early to late velocities (E/A), deceleration time of E velocity and isovolumic relaxation time. Early (e’) and late (a’) diastolic mitral annular tissue velocities at both the septum and lateral walls were obtained, and the ratio of E/e’ using the average of septal and lateral tissue velocities obtained was calculated as an index of diastolic function.15

LA volumes were assessed by the biplane Simpson's method from apical views. Subjects with inadequate visualization of the atria from this view or electrocardiogram Tracings inadequate for assessment of p wave were excluded. Maximum volume (Vmax) was measured just before opening of the mitral valve, minimal volume (Vmin) measured at mitral valve closure, and volume prior to atrial contraction (Vp) was measured at the onset of the p wave. From measurements of these phasic volumes, phasic functions of the LA were measured as follows: LA expansion index (reflecting reservoir function) (Vmax − Vmin)/Vmin × 100; LA passive emptying fraction (reflecting conduit function) (Vmax − Vp)/Vmax × 100; and LA active emptying fraction (reflecting booster function) (Vp − Vmin)/Vp × 100.16 In addition, LA total emptying fraction was assessed (Vmax − Vmin)/Vmax × 100.

Descriptive statistics were performed, and as the cardiac parameter data were normally distributed, results are described as means and standard deviations. Percentages were calculated as appropriate. For continuous variables, differences between means were calculated using t tests, and cumulative survival was assessed by Kaplan–Meier analysis and log‐rank test for statistical significance. Adjusted and unadjusted Cox proportional hazard models were performed. As it was hypothesized that only significantly impaired atrial function would affect mortality, comparisons were made between the quartile with the most impaired atrial function and the upper 3 quartiles. All models were adjusted for and cardiac measurements of EF, E:e’, LV mass index, and LA volume index. All P values were 2‐tailed, and P < 0.05 was considered significant. Data storage and analysis were performed using SAS version 9.1e (SAS Institute, Inc., Cary, NC).

3. RESULTS

Two hundred eighty‐two subjects were included in the study, of whom 87 (31%) had died at the time of 5‐year follow‐up. Clinical characteristics and echocardiographic measurements of the study population as a whole are depicted in Supporting Tables 1 and 2 in the online version of this article. Clinical parameters of the lowest quartile as compared to the upper 3 quartiles of the 3 parameters of atrial function examined are depicted in Table 1. There were no significant differences between the groups except for more diabetics in the lowest quartile of LA expansion index and total LA emptying fraction, and more subjects post‐cardiovascular accident in the lowest quartile of LA active filling. Echocardiographic parameters in the different subgroups of atrial function are shown in Table 2. A significantly larger E:e’ was noted in subjects with reduced LA active filling. There were no significant differences in LA volume in any of the parameters of LA function examined.

Table 1.

Clinical characteristics of study population by quartiles of LA function

	LA Expansion					LA Passive Filling					LA Active Filling
	Upper Quartiles (n = 221)		Lowest Quartile (n = 61)		P	Upper Quartiles (n = 208)		Lowest Quartile (n = 74)		P	Upper Quartiles(n = 214)		Lowest Quartile (n = 68)		P
	N	%	N	%		N	%	N	%		N	%	N	%
Female	119	53.8	32	52.5	0.85	115	55.3	36	48.1	0.33	116	54.2	35	51.5	0.69
Education >12 years	113	51.1	30	49.2	0.79	100	48.1	43	58.1	0.14	111	51.9	32	47.1	0.49
Married	100	48.1	31	51.7	0.62	96	49.0	35	48.6	0.96	95	47.3	36	53.7	0.36
Diabetes	39	18.6	21	35.0	0.01	41	20.7	19	26.4	0.32	37	18.2	23	34.3	0.01
IHD	83	39.5	20	33.3	0.38	76	38.4	27	37.5	0.89	78	38.4	25	37.3	0.87
HTN	149	71.0	46	76.7	0.39	148	74.7	47	65.3	0.12	141	69.5	54	80.6	0.08
CHF	26	12.4	6	10.0	0.68	21	10.6	11	15.3	0.29	23	11.3	9	13.4	0.64
CVA	30	14.3	11	18.3	0.44	28	14.1	13	18.1	0.43	25	12.3	16	23.9	0.02
Never smoked	126	58.9	33	55.9	0.07	120	59.7	39	54.2	0.15	121	59.0	38	55.9	0.52
Frailty	29	16.8	9	19.1	0.70	29	18.0	9	15.3	0.63	26	15.6	12	22.6	0.24
	LA Total Emptying Fraction

	Upper Quartiles (n = 203)		Lowest Quartile (n = 79)		P
	N	%	N	%
Female	110	54.2%	41	51.9%	0.7293
Education >12 years	107	52.7%	36	45.6%	0.2815
Married	91	47.4%	40	52.6%	0.4396
Diabetes	33	17.1%	27	35.1%	0.0013
IHD	76	39.4%	27	35.1%	0.51
HTN	134	69.4%	61	79.2%	0.1049
CHF	24	12.4%	8	10.4%	0.6387
CVA	27	14.0%	14	18.2%	0.3861
Never smoked	117	59.7%	42	54.5%	0.2042
Frailty	24	15.2%	14	22.6%	0.192

Abbreviations: CHF, congestive heart failure; CVA, cerebrovascular accident; HTN, hypertension; IHD, ischemic heart disease; LA, left atrial.

Table 2.

Echocardiographic parameters in the subgroups of LA function

	LA Expansion Index			LA Passive Filling			LA Active Filling
	Upper Quartiles (n = 221)	Lowest Quartile(n = 61)	P Value	Upper Quartiles(n = 208)	Lowest Quartile(n = 74)	P Value	Upper Quartiles(n = 214)	Lowest Quartile(n = 68)	P Value
LV mass index (g/m2)	124.6 ± 38.8	115.1 ± 34.1	0.11	122.9 ± 37.9	121.3 ± 38.6	0.79	124 ± 35.7	118.3 ± 44.2	0.34
LVEF (%)	60 ± 10	60 ± 10	0.67	60 ± 10	50 ± 10	0.17	60 ± 10	60 ± 10	0.68
E/E’ ratio	12.1 ± 4.5	12.8 ± 4.8	0.34	12.3 ± 4.6	12 ± 4.5	0.72	11.9 ± 4.4	13.5 ± 5	0.041
LA volume index (mL/m2)	39 ± 16.6	38.6 ± 15.9	0.86	39.5 ± 16.5	37.3 ± 16.1	0.36	38.5 ± 16.6	40.2 ± 15.8	0.46

	LA Total Emptying Fraction		P Value
	Upper Quartiles (n = 203)	Lowest Quartile (n = 79)
LV mass index (g/m2)	124.4 ± 39.4	118.1 ± 34.3	0.2711
LVEF (%)	60 ± 10	60 ± 10	0.6353
E/E’ ratio	11.9 ± 4.3	13.0 ± 5.2	0.1087
LA volume index (mL/m2)	39 ± 17.1	38.6 ± 14.7	0.8353

Abbreviations: LA, left atrial; LV, left ventricular; LVEF, left ventricular ejection fraction.

Kaplan–Meier curves depicting cumulative 5‐year survival for subjects stratified into the lowest quartile and the 3 upper quartiles of LA function are demonstrated in Figures 1, 2, 3. Survival of subjects in the lowest quartile of LA expansion as well as LA active filling was significantly lower than survival in the upper quartiles. There was a trend toward significantly lower survival in the lowest quartile of total LA emptying fraction (see Supporting Figure 1 in the online version of this article) with no significant difference in survival related to LA passive filling. When measurements of LA volume index were added to the model, the relationship between survival and indices of LA function remained significant (P = 0.016 for LA expansion index, P = 0.003 for LA active filling). However, when other indices of cardiac function, including left ventricular ejection fraction, E:e’, and LV mass, were added to the model, the relationship was attenuated and no longer significant.

Figure 1.

Atrial filling and mortality. The blue line indicates the upper quartiles, and the red line indicates the lowest quartile (P < 0.016)

Figure 2.

Active atrial emptying and mortality. The red line indicates the lowest quartile, and the blue line indicates the upper quartiles (P < 0.0013)

Figure 3.

Passive emptying and mortality. The blue line indicates the upper quartiles, and the red line indicates the lowest quartile (P < 0.12)

A sensitivity analysis testing the relationship between the different components of LA function and mortality was performed in subgroups of subjects with IHD and CHF, and the results were similar to those demonstrated in the total population.

4. DISCUSSION

This study utilizing home‐based echocardiography demonstrated that subjects ages 85 to 86 years old with significantly impaired LA function had increased 5‐year mortality independently of indices of LA volume. To our knowledge, this is the first community‐based study to use home echocardiography to address the effects of LA function on prognosis in the elderly.

Many studies have examined the prognostic importance of changes in LA volume with aging; however, limited data are available regarding changes in, and the prognostic value of, different parameters of LA function in an elderly population. Using 3‐dimensional (3‐D) echocardiography in a cross‐sectional study in an elderly population, Goncalves et al. demonstrated impaired LA emptying fraction in subjects with CHF, but not in subjects with hypertension (HTN) or who were free of cardiovascular disease, suggesting that unlike increasing LA volumes, impaired LA active contraction is not a simple consequence of ageing.17 Therefore, changes in LA function in this age group may represent pathology and influence prognosis. In a study utilizing cardiac magnetic resonance imaging in a younger population (mean age 43 years), Gupta et al. showed that LA emptying fraction was weakly associated with LA volume and had a greater impact on prognosis.9 LA emptying fraction does not distinguish between active and passive LA filling. Our study extends these findings to an elderly population using more detailed parameters of LA function.

The left atrium contributes to cardiac performance through 3 functions: as a reservoir during LV systole represented by the LA expansion index, as a passive conduit during early diastole represented by LA passive filling, and as a pump during late diastole represented by LA active filling. The atrial endocardium tends to become thicker and more dense with aging.18 These changes in physical properties would be expected to have a greater influence on active parameters of LA function, such as reservoir and pump, supporting the findings of our study. Our findings that the incidence of pathologies such as HTN and IHD were similar across quartiles of LA function suggests that these changes are related to intrinsic changes in atrial function with aging and are not secondary to other pathologies. On the other hand, the conduit function is strongly influenced by LV relaxation and compliance, properties universally decreased with aging, and changes in this LA parameter may be less associated with prognosis in this elderly population.19 The findings in our study may have clinical importance in the elderly population. For example, left atrial dysfunction may contribute to the pathophysiology of heart failure with preserved ejection fraction (HFPEF), a common clinical problem in the elderly.20 In patients with HFPEF, previous studies have shown the prognostic importance of LA function even in subgroups with normal LA size.21 Our findings also suggest that preserving LA pump function, for example, by trying to maintain sinus rhythm in patients with atrial arrhythmias may be important in this subset of patients. Future clinical trials are necessary to examine these questions.

The major strengths of our study are the use of an age‐homogenous cohort to minimize variability of the clinical findings and the use of home echocardiography to study a more representative sample of this age group. The limitations include the fact that only a subset of subjects had adequate imaging of the atria as well as the relatively small number of subjects, which may have limited the power of our findings in the statistical models. Newer echocardiographic techniques such as 3‐D imaging and atrial strain imaging were not performed; however, a recent study has shown a good correlation between LA strain and LA emptying fraction.22

5. CONCLUSION

Left atrial function is predictive of increased 5‐year mortality independently of indices of LA volume in the oldest old. Clinical strategies aimed at preserving LA function may improve outcomes in this population.

Conflicts of interest

The authors declare no potential conflicts of interest.

Supporting information

Supplementary figure 1 Total emptying fraction and mortality

Supplemental Table 1 Clinical characteristics of the whole study population

Supplemental Table 2 Echocardiographic measurements in the whole study population

Leibowitz D, Koslowsky J, Gilon D, Jacobs JM, Stessman‐Lande I, Stessman J. Left atrial function and mortality in the oldest old. Clin Cardiol. 2017;40:1323–1327. 10.1002/clc.22831