• heart failure

Many studies have reported declining rates of heart failure hospitalisations since the 1990s.1 However, there is increasing evidence that these trends differ by age. Typically, heart failure is a common and insidious disease with substantial associated morbidity and mortality burden worldwide. This has led to a significant focus on improving treatments and medical therapy for this condition, particularly in older populations. Although heart failure is seen as a disease of the elderly, new research suggests that younger patients are being hospitalised for heart failure at an increasing and alarming rate.

Chan and colleagues2 report on trends in incident heart failure hospitalisations for the population of New Zealand between 2006 and 2018. Importantly, they suggest a differential trend by age, with incidence in patients aged <50 years increasing on average by 1.5% per year during the study period. In people aged between 50 and 79 years, trend analyses showed generally declining rates until 2013, when rates in men and women in these middle age groups started rising. These trends contrast with those in people aged over 80 years, where a consistent linear decline over the 13-year study period was noted. Comparable trends in younger adults have been reported from a number of whole population studies, including from Denmark,3 Sweden4 and Western Australia,5 yet it is unclear as to why these trends are occurring. The increases in younger age groups reported from these studies were around 50% over study periods between 16 and 20 years, a likely greater magnitude of increase compared with that reported by Chan. The timing of rising incidence rates between populations varies somewhat, with increases in Denmark3 and Sweden4 occurring since the 1990s and through the 2000s, whereas a more recent increase (since 2006) is reported from Australia5 and by Chan and colleagues.2

Given the pervasive thinking that heart failure is a disease of the elderly, why is the incidence of heart failure rising in young adults? Despite the known association of ischaemic heart disease (IHD) with onset of heart failure, increasing incidence is occurring in younger adults without a history of hospitalisation for IHD. The hypothesis from Chan and colleagues, and other authors, is that the rising prevalence of obesity and diabetes may be driving the adverse trends. Although total cardiovascular risk burden has declined, primarily driven by reductions in smoking prevalence, hypertension and dyslipidaemia, most high-income countries report increasing prevalence of obesity and diabetes. This hypothesis is therefore plausible, particularly given the strong contribution of these risk factors to the risk of developing heart failure and the greater impact of obesity on cardiac function and development of heart failure in younger adults.6

Increasing rates of cardiomyopathy have also been reported in younger populations. Whether these trends are contributing to population-level heart failure trends, particularly in younger adults, is not yet clear. The small proportion of heart failure patients with cardiomyopathy in Chan’s study (2% in 2018) differs from other studies. In Sweden, 20% of adults with incident heart failure aged under 55 years were designated as having a history of varying types of cardiomyopathy.4 Weber et al 5 recently reported that in under 55-year-old patients hospitalised with incident heart failure, the proportion of cases with new onset cardiomyopathy increased from 24% to 37% during the 16-year study period. Although this does not address the underlying prevalence of cardiomyopathy, it does signify either a background rise in hospitalisations for cardiomyopathy and/or increased detection in this age group. The differing definitions of cardiomyopathy used in these studies, with some using a broad definition for cardiomyopathy (hypertrophic, dilated alcoholic and non-cardiovascular aetiology cardiomyopathies),2–4 and others excluding hypertrophic but including ischaemic cardiomyopathies,5 may explain some between-study differences. A greater understanding of the association between the onset of cardiomyopathy and development of subsequent heart failure on population trends in younger adults is necessary, as administrative data may not capture all presentations for these conditions and coding of cardiomyopathy does not necessarily match updated clinical classifications. Concurrently, an increasing awareness via screening and genetic testing may impose an upward trend in cardiomyopathy prevalence in such populations.

These worrying trends in heart failure incidence in younger adults should be viewed in the context of other major cardiovascular disease trends in this age group. There is consistent reporting of stable or increasing rates of myocardial infarction and stroke in people of comparable age. This lends credibility to the hypothesis that risk factor changes may be impacting heart failure incidence in younger adults, as obesity and diabetes are also known to contribute significantly to the population risk of atherosclerotic-driven conditions. It highlights one of the difficulties with many population-based studies—the inability to directly attribute disease trends to specific causal factors such as risk factor prevalence changes or changes in treatment for a disease. While there has been substantial modelling to determine the impact of risk factor prevalence changes on coronary heart disease, this focus has not extended to incidence of the heart failure.

An advantage of such administrative data studies, particularly where linkage of historical records is available, is the ability to define first-ever heart failure hospitalisation. While this may be a proxy for true disease onset (as patients may have initially accessed the health system via primary care or outpatient settings), the study of Chan and colleagues excludes patients with prior heart failure hospitalisation history and ensures they are not counted multiple times. While measuring total hospitalisations for heart failure is a valid indicator of acute heart failure episodes and potentially worsening outcomes, measuring the first hospitalisation reduces the impact of changes in recurrence, admission practices or treatment approaches. However, such studies lack clinical confirmation of heart failure phenotype, therefore we do not know what phenotype is driving trends across different age groups. While declining incidence of heart failure appears to be driven by greater reductions in patients with reduced versus preserved ejection fraction,7 whether this trend extends to younger adults is unclear.

The importance of large population studies in this context is that very small changes in rates can be detected, even in groups like younger adults where the number of events is often low. This is usually not possible in smaller registry or clinical studies. Whether small but statistically significant changes found in low incidence rates are clinically meaningful is often debated. When plateauing of coronary heart disease mortality in younger adults was first reported in the 2000s, there was some conjecture that this was a ‘natural’ result of low rates in this age group.8 Rates of incident heart failure hospitalisation reported by Chan and colleagues are in the order of 20–30/100 000 person-years, 50 times lower than rates in the oldest age groups. However in absolute terms, this equates to nearly 600 people annually in the New Zealand population aged <50 years with onset of heart failure, with ongoing substantial risk of poor cardiovascular outcomes and increasing heart failure burden with ageing.