"Are you dying to know? Scientists develop death test to predict if you'll make it to 2020," The Daily Telegraph reports. The test is based on analysis of data collected from the UK Biobank.
This is essentially a huge ongoing cohort study that collected data from almost 500,000 middle- to older-age adults in the UK over an average of five years. This data was then used to create an online death risk calculator.
The researchers looked at around 650 different measurements, including blood tests, family history, health and medical history to work out which were most strongly associated with risk of death over the next five years.
This was then used to create an online risk calculator of death. For this, the researchers focused on factors that were easy for people to self-report. For example, you probably have no clue what size your red blood cells are or what your cholesterol level is, but you do know how many children you have.
The factors included in the tool do not necessarily cause death, but are associated with an increase in risk. Many of these factors cannot be changed, such as already having a longstanding illness, but some can be, such as smoking – the strongest predictor of death in people with no medical illness.
Researchers hope the calculator may motivate people to make improvements to their health, or help doctors identify people who could be targeted for interventions to reduce their risk. Studies will be needed to assess whether the tool does lead to these effects.
Where did the story come from?
The study was carried out by researchers from the Karolinska Institut and Uppsala University in Sweden, and was funded by the Knut and Alice Wallenberg Foundation and the Swedish Research Council.
The Knut and Alice Wallenberg Foundation is the largest private financier of research in Sweden, and their goal is to "promote scientific research, teaching and education beneficial to the Kingdom of Sweden".
The study used data from the UK Biobank, a registered charity set up in the UK by The Wellcome Trust, the Medical Research Council, the Department of Health, the Scottish Government, and the Northwest Regional Development Agency, with additional funding from the Welsh Assembly Government, the British Heart Foundation and Diabetes UK.
Most of the UK media described the questions included in the online prediction tool UbbLE (UK Longevity Explorer). They also provided expert opinion that highlighted hopes the tool may help people make healthier lifestyle choices, but also pointing out that most of the predictive factors used in it do not directly cause disease.
But some of the reporting of this study by some news sources suggests the online calculator predicts if you'll die within five years – this is not the case. The test does not categorically tell you whether you will die or not, it only gives you a percentage chance based on your characteristics.
What kind of research was this?
This research used data from a large cohort study of middle- and older-age people from the UK. The researchers wanted to work out the association between multiple measurements of health and socioeconomic status and risk of death over the next five years.
They also planned to create an online tool using the strongest predictors that could be self-reported to allow people to assess their individual risk.
As the analysis is based on a cohort study, the results do not prove cause and effect, and this was not the aim of this particular study. It wanted to identify predictors of risk of death, not necessarily things that cause death directly.
What did the research involve?
The researchers used information from the large prospective UK Biobank cohort study. By collecting data on the cohort over many years and allowing scientists access to this information, the Biobank aims to improve the prevention, diagnosis and treatment of a wide range of serious and life-threatening illnesses, including cancer, heart disease, stroke, diabetes, arthritis, osteoporosis, eye disorders, depression, and forms of dementia.
UK Biobank recruited 500,000 people aged 40 to 69 into the study between 2006 and 2010. The participants filled out questionnaires and had numerous baseline measurements taken at one of 21 assessment centres across Scotland, England and Wales. In all, there were 655 measurements, which were grouped into 10 categories:
- blood tests
- cognitive function
- early-life factors
- family history
- health and medical history
- lifestyle and environment
- physical measures
- psychosocial factors
- sex-specific factors
The researchers used information for 498,103 people and identified any of those people who died up to February 2014, or December 2012 for Scottish participants. Central NHS registers were used to obtain cause of death.
The researchers then analysed the 655 measurements separately for women and men to determine their association with risk of death over the five years of follow-up.
They also calculated the association between each measurement and risk of death for three different age groups:
- 40 to 53 years old
- 53 to 62 years old
- over 62 years old
The risks associated with all 655 measurements were then displayed for women and men on the UbbLE website on pages called the association explorer.
The researchers used the measurements showing the strongest association with risk of death that could be self-reported to create an online mortality risk calculator. This meant excluding any blood tests or physical measurements that a person could not easily, quickly and reliably take themselves.
What were the basic results?
A total of 8,532 people died during the study period, about 1.7% of participants. This was lower than that of the general UK population, which might suggest the people who took part were generally healthier than the population as a whole.
The most common causes of death were:
- cancer (53% of male and 69% of female deaths)
- cardiovascular diseases, such as a heart attack or coronary heart disease (26% of male and 13% of female deaths)
- lung cancer in men (10% of male deaths, 546 cases)
- breast cancer in women (15% of female deaths, 489 cases)
The strongest predictor of death over five years was:
- self-reported health in men
- a previous history of cancer in women
Other examples of strong predictors of death from various categories included:
- self-reported walking pace – for example, men aged 40 to 52 reporting a slow pace had a 3.7 times higher risk of dying within five years than those who reported walking at a steady average pace
- red blood cell size
- pulse rate
- forced expiratory volume in one second (as a measure of lung function)
When the researchers excluded people with serious illnesses, smoking was the strongest predictor of death.
From these results, the researchers created a prediction tool based on 13 questions for men and 11 questions for women. Based on a person's responses and death rates for the general UK population, the tool estimates how likely a person is to die in the next five years.
How did the researchers interpret the results?
The authors concluded that, "The prediction score we have developed accurately predicts five-year all-cause mortality and can be used by individuals to improve health awareness, and by health professionals and organisations to identify high-risk individuals and guide public policy."
This large study has identified numerous risk factors associated with a person's risk of death within five years. Researchers used this information to develop an online tool that predicts someone's risk of death within the next five years. The study's strengths include its large sample size and the prospective nature of the study design.
But there are some limitations. There may be some bias in the type of people who volunteered to take part. The death rate was lower than that of the average population in this age group, which may indicate that the participants were more interested in their health and so had healthier lifestyles. This may limit whether the results apply to the population as a whole.
The study only included people from the UK between the ages of 37 and 73, and results may not apply to people outside that age range or from other countries. For example, the data is reliant on self-reporting, and people from other age groups or countries might interpret some of these concepts differently. This may not be an issue for some factors, but it might be an issue for others, such as estimating walking pace or level of health.
This study did not aim to assess whether the factors directly increased risk of death – rather, it set out to identify factors that are associated with and can predict death risk when combined. Also, several factors in the calculator cannot be changed, such as current and past health. But smoking – a factor that can be altered – was the factor most strongly predictive of death over the next five years.
As the media pointed out, other aspects of an unhealthy lifestyle, such as poor diet, excess alcohol intake and being overweight, are not included in the online risk calculator.
This is most probably because other factors had stronger associations and were chosen because it made completing the risk questionnaire easier, so overall this would give a better indication of risk. Questions in the risk calculator, such as, "In general, how would you rate your overall health?" are affected by multiple factors, such as obesity and alcohol intake.
The authors say they hope the online tool may help people make positive lifestyle changes. However, they also acknowledge that online health information may increase overdiagnosis and anxiety. Follow-up studies are needed to determine the effects of the calculator – for example, to what extent it motivates people to change their lives and what impact this has.
Overall, one message the study highlights is the importance of stopping smoking. Find out how you can quit smoking.
Analysis by Bazian
Edited by NHS Website
Links to the headlines
The Daily Telegraph, 4 June 2015
The Guardian, 4 June 2015
Daily Mirror, 4 June 2015
Daily Mail, 4 June 2015
Sky News, 4 June 2015
Links to the science
udy. The Lancet. Published online June 3 2015