‘Derivation of new reference tables for human heart weights in light of increasing body mass index’
- Correspondence to Professor Sebastian Lucas, Department of Histopathology, KCL School of Medicine, St Thomas' Hospital, London SE1 7EH, UK;
- Accepted 9 February 2011
In this issue, a group from Oxford pathology and medical statistics present the 21st century detailed comparisons of adults' heart weights with other body statistics, as measured at autopsy.1 Their results are markedly different from the standard tables that pathologists have been using through most of the last 100 years, which were derived from North America in 1928.2 They also reinforce the need to include a comparative total body measurement (surface area and weight) when considering whether or not a heart is enlarged—a common question at autopsy—rather than just noting the heart weight and sex of the patient.
Using numerous statistical calculations on hearts and bodies of persons who died without obvious heart or lung disease, heart weight correlated best with body surface area and nearly as well with body weight. The calculus for body surface area is a complicated transformation of body weight and height, so pathologists will be grateful that body weight alone is satisfactory for most purposes. UK mortuaries increasingly measure height and weight of bodies—a recommendation that came from the National Confidential Enquiry into Patient Outcome and Death (NCEPOD) 2006 report, where 45% of a large sample of coronial autopsy reports did not include a body weight.3 That said, the calibration and consistency of many mortuary body and organ scales are, anecdotally, suboptimal.
A significant statistic from this study is that the normal heart/body weight ratio for both men and women, aged 14 years and above, who are not obese or malnourished, is 0.51%. According to the 1928 data, it is 0.43% and 0.4% for men and women, respectively. This historical difference may not seem much, but for a 70 kg female the difference between the ‘normal’ heart weight being 357 g vs 280 g; and for a 85 kg man, the difference being 434 g vs 366 g. This higher (0.51%) ratio is not actually new for our times: a normal ratio of 0.5% was published from Brazil, in 2002.4
Why do we need tables and standards?
Case definitions matter for precision of investigating diseases and their epidemiology. In the area of sudden cardiac deaths when coronary atherosclerosis, hypertension and valvular disease are not the cause, several conditions have the exclusion of an enlarged heart in their descriptions. The UK Cardiac Pathology Network (UKCPN) classifies such sudden cardiac deaths as ‘explained’ versus ‘unexplained’.5 In the ‘explained’ list (which means ‘a diagnosis made’ rather than ‘pathogenesis necessarily understood’), there are ‘idiopathic left ventricular hypertrophy’ and ‘idiopathic left ventricular fibrosis’, where the heart is enlarged. Conversely, in the ‘unexplained’ list are ‘sudden unexplained cardiac death in alcohol abuse (SUDAM)’6 and ‘sudden arrhythmic death syndrome (SADS)’,7 where the definitions specifically exclude an enlarged heart.
One of the many reasons, I suspect, why the SADS group of diseases has taken time to be appreciated by pathologists, coroners and epidemiologists has been lack of close attention to measurements of heart and body at autopsy—plus a natural reluctance to admit that one cannot find a standard accepted disease at gross examination. And we have all been leaned on by coroners to state that “left ventricular hypertrophy is a reasonable cause of natural sudden death” without encouragement to consider the pathogenesis and potential hereditary implications more deeply.
Have heart and body weights changed over time?
Yes markedly. Smith did his landmark study on selected cadavers (without primary or secondary heart disease) in the Mayo Clinic, Minnesota, USA, in 1920s. For men, the average heart and body weights were 294 g and 65 kg, respectively; for women, they were 250 g and 59 kg, respectively. A happier and more innocent time—or just less well fed? In a report from Oxford, UK, in the period 2003–2006, these weights were greater across the board. This tells us that such old tables are now not fit for the purpose, and thus the new tables and measures from Oxford are most welcome.
What are the drivers of these changes?
The fact that body mass index (BMI—body weight in kg/ height in m2) has increased progressively, leading to the obesity epidemic in rich and many poorer countries, is not contested. The impact on heart weight comes from the need for a larger myocardial mass to perfuse a bigger body. But note that hypertension and sleep apnoea are both associated with obesity, so these conditions which also induce hypertrophy of heart muscle need to be excluded when preparing baseline tables and standards.
Ethnic difference is a potential confounder in all these studies. In the Oxford study, only 4% of autopsies were non-Caucasian and were not stratified out. But in the data from Arkansas, USA, in the period 1964–1999, there were no differences in heart and body weight relationships between Caucasian and African-American subjects.8 One can also see there the time shift for heart/body ratios in operation: taking 75 kg weight males, whereas the 1920s found the average heart weight to be 325 g,2 about 50 years later the average was 380 g.
Longevity is increasing along with BMI. The Oxford data noted that normal heart weight relative to body surface area increased with age. This small increase is unexplained.
What new insights can come from more accurate comparison of heart/body weight?
As well as more careful consideration of diagnostic categories in cardiac work, there are two specific scenarios where modern biometric data are significant for our understanding: obesity and pregnancy.
Smith and colleague, when they studied obese subjects at the Mayo Clinic, noted that very obese people had a subnormal heart/body weight ratio: beyond 104.5 kg body weight, the heart weight did not increase at the same rate.9 Unfortunately, the Oxford data did not include sufficient numbers of seriously obese subjects to provide more information here. The concept of sudden cardiac death in, and directly due to, morbid obesity has gained credence since 1995.10 Obviously, this is a pathogenetic minefield, which is why pathologists can help our understanding by carefully measuring bodies and hearts and categorising cardiovascular death in the obese patient at autopsy.
The UKCPN has in its database the category of ‘sudden unexplained cardiac death in obesity’.5 So, it is not encouraging to note a subset analysis from the NCEPOD data.3 When describing, in the autopsy reports, subjects with BMI over 30, and where trauma, accidents and alcohol toxicity were excluded, pathologists used the terms ‘overweight’ or ‘obese’ in only 34% of cases; and in the cause of death, obesity was indicated in only 27% of those with a BMI of >40 (morbidly obese).11 We can do better.
Concerning pregnancy, the latest triennial confidential enquiry into maternal death in the UK, for the period 2006–2008, shows cardiac deaths to be the leading causes of death during and shortly after pregnancy and are increasing in rate.12 As well as standard entities such as peripartum cardiomyopathy and ischaemic heart disease, SADS stands out, and some of these patients were also obese. As a diagnosis of exclusion, SADS is critically dependent on a thorough autopsy with histopathology and toxicology etc, and accurate comparison of body and heart size is necessary. With the ‘improved’ ratios indicated in the Oxford study data, it is also possible that some more cases might have been included in this diagnosis: case definitions matter.
What should be done now?
The Oxford study ends with a recommendation to collect many more data points from many centres, partly to characterise what happens at the extremes of body weight but also to develop even more user-friendly reference charts. Also, it will take more observational autopsy data to finally condemn to the dustbin the early 20th century heart/body weight tables that have served us for too long. So, we look forward to a renewed input of such data for our modern age, with the encouragement to think about heart and body sizes more critically, and what they mean for the individual patient case as well as for epidemiology.
Linked article 084574.
Competing interests SBL is a member of the Steering Group of the UK Cardiac Pathology Network; the pathology clinical coordinator for the National Confidential Enquiry into Patient Outcome and Death (NCEPOD); the central pathology assessor for the Centres for Maternal and Child Enquiries (CMACE).
Provenance and peer review Commissioned; not externally peer reviewed.