Archive for the ‘Health’ Category

COVID-19: The risk to men

February 17, 2021

Click for a larger version. Output from the QCOVID calculator showing the risk arising from ethnicity and biological sex compared to a 61-year old woman.

Friends, you may have read recently (e.g. The Guardian) of a new tools for analysing the risks of dying from COVID-19 including many different risk factors, including ethnicity.

The tool – called QCOVID – is available for you to use here.

Media coverage generally highlighted the relative risks appropriate to different ethnic minorities.

For some reason, the media did not mention the most prevalent risk factor for dying from COVID-19: being male.


As one does, I immediately typed in my own basic statistics: I am a white male aged 61 with a BMI of 24.8. The result was this:

Click for a larger version. Output from the QCOVID calculator for someone with my vital statistics.

I took note of the absolute risk of a COVID associated death: my risk was 0.0173% over a 90-day period i.e. a risk 1 in 5,780.

I then changed the ethnicity and biological sexuality entries to the calculator across a number of categories.

Changing ‘my’ biological sexuality to female I saw that ‘my’ risk went down to 0.0077% over a 90-day period i.e. a risk 1 in 12,987.

So the additional risk factor associated with being a biological male was 2.2.

Click for a larger version. Output from the QCOVID calculator showing the risk arising from ethnicity and biological sex compared to a 61-year old woman.

I was curious as to how this compared with differences arising from ethnicity and the results relative to the ‘Me-female’ are shown in the table above.

I found the results striking.

  • For females, the additional ethnic risk factors ranged from 1.2 to 2.0. In comparison the additional risk factor of being a white male was 2.2
  • In every ethnic category, being male carried an additional risk factor varying from 2.2 to 4.6 compared to the equivalent female.


I did not investigate all the various categories in the QCOVID calculator, so I cannot confirm the complete generality of this result.

But from my simple investigation, I conclude that:

  • Men of any ethnicity are at least twice as likely to die from COVID-19 as women from the same ethnic background.
  • The additional risk faced by women associated with their ethnicity is less than the additional risk faced by white males associated with their sexuality.

My question is this: why has this dramatic result affecting millions of people not been reported more widely?

Update 19th February

A correspondent pointed out in the comments this excellent data source

This site shows that excess mortality amongst men occurs worldwide, but the extent of it is highly variable from one country to another.

I don’t have an explanation of this phenomenon, but it does appear to be very real, whatever its cause.



Did 28,000 people die from air pollution in 2012? Part 3

May 28, 2015


COMEAP needed to estimate how many percent excess mortality would be caused by 10 micrograms per cubic metre of PM2.5 air pollution. To do this they just asked themselves what they thought the answer was. The details of this 'elicitation' process are described below.

The Committee on the Medical Effects of Air Pollutants (COMEAP) needed to estimate how many percent excess mortality would be caused by 10 micrograms per cubic metre of PM2.5 air pollution. To do this they just asked themselves what they thought the answer was. Their answers range from -1% (i.e. air pollution prolonged life) to 17% excess mortality. The details of this ‘elicitation’ process and the meaning of excess mortality are described in the text.  This graph represents their consensus view: individual estimates are shown at the end of the article. (SOURCE COMEAP 2009 Page 160.)

Continued from PART 2

In Part 1 we saw that is untrue that air pollution annually causes 28,000* deaths of otherwise healthy people.

In Part 2 we described the qualitative effects of air pollution.

We now look at how the Committee on the Medical Effects of Air Pollutants  (COMEAP) estimate the effects of air pollutants, and how they have chosen to communicate their estimates.

What I have written below is the result of staring at reading two extremely technical documents, an activity which I invite you to share:

The factual basis…

The COMEAP studied hundreds of scientific papers but singled out one for special mention: the American Cancer Study (ACS) [COMEAP 2009 Page 2: Point vii]

The ACS uses measurements of soot particles called PM2.5 as a ‘proxy’ for all the other forms of air pollution.

It does this because PM2.5 s are easy to count and the concentration of the other air pollutants is more difficult to quantify, but they are typically correlated with PM2.5s.

So we don’t know which part of the pollution is actually the active cause of any effect – a shortcoming the experts consider exhaustively in their 2009 report.

The ACS finds a correlation such that there is a 6% rise in all-cause mortality for a 10 microgram per cubic metre increase in PM2.5 concentration.

I haven’t been able to read the ACS but extraordinarily, COMEAP report no uncertainty in this 6% figure.

Perhaps it is because of this that they decide to make their own estimate. And to do this they used a procedure called elicitation.

What is elicitation?

Wikipedia defines elicitation as:

the synthesis of opinions of authorities of a subject where there is uncertainty due to insufficient data or when such data is unattainable because of physical constraints or lack of resources. Expert elicitation is essentially a scientific consensus methodology… [it] allows for … an “educated guess”.

Each expert on the committee was asked:

  • “What do you think is the chance that the correlation is not a 6% increase in mortality for a 10 microgram per cubic metre increase in PM2.5 concentration, but some other figure?

The experts were asked in turn:

  • What do you think is the chance that the true figure is greater than 0%?
  • What do you think is the chance that the true figure is greater than 1%?
  • What do you think is the chance that the true figure is greater than 2%?
  • etc. until they arrived at the figure of 17%, beyond which possible correlations were discounted as highly improbable.

The results are shown in the chart at the head of this article and in the chart at the end of the article.

And the results…

The results of this elicitation form the basis of all the mortality estimates that you have read in the previous two articles.

The results are literally, an educated guess. None of the fancy maths changes this.

The average estimate was 6% and experts considered that the true answer was ‘95% likely’ to be in the range between 0% and 15%

Since the level of PM2.5s in the UK is typically just under 10  microgram per cubic metre, it is this figure of 6% of the roughly 500,000 deaths per year that directly links to the estimate of 28,000* deaths per year.

The range of expert opinion is that the true figure could lie between 0 and 55,000: every one of the seven experts considered it marginally possible that there was no effect of pollution i.e. that the excess mortality was 0% or less.

Although the experts considered a plethora of studies – and their report is exhaustive and exhausting – in the end, this is just their opinion.

How to express their results?

COMEAP consider at length how to express the results of their elicitations. Mostly they considered that the effect on life-expectancy (a few months for most people in the UK) is clearest.

However they do specifically endorse the use of mortality to express the effect of air pollution. But they note:

… the result expressed in terms of attributable deaths or additional deaths may easily be misunderstood or misrepresented. This calculation is not [their emphasis] an estimate of the number of people whose untimely death is caused entirely by air pollution, but a way of representing the effect across the whole population when considered as a contributory factor to many more additional deaths.

...consequently we consider it unrealistic to view air pollution as the sole cause of death in a number of cases equal to the population attributable deaths.” [COMEAP 2010 Page 3]

Regarding the actual estimates COMEAP explicitly recommend that the excess mortality estimates should always include the full uncertainty estimate – i.e. including the possibility that the excess mortality may be zero.[COMEAP 2009 Page 3 Point xiii]

And my conclusion…

I agree with COMEAP that the expression of the effects of air pollution as excess deaths per year can be “easily misunderstood”. In fact I think it is nearly universally misunderstood.

And given that their recommendation that the full limits of uncertainty – which include the possibility of no effect – are generally not quoted – I think this gives quite the wrong impression.

The possibility that air pollution – mainly from traffic – might be killing 28,000* people a year – 17 times more than are killed in car accidents – is horrific.

But in fact it appears that air pollution of itself kills nobody.

Rather, when we are near the end of lives – which are now longer than they have ever been in human history – our lungs do not function as well as they might otherwise have done.

As a result, our demise – from whatever cause – is hastened.

Somehow this reality is not quite as scary as COMEAPs vision.

Other Figure

The graph as the top of the page was compiled from the responses of 7 experts. These graphs show the cumulative probabilities ascribed to a particular sensitivity coefficient by the each of the 7 experts.

The graph as the top of the page was compiled from the responses of 7 experts ‘A to G’. These graphs show the cumulative probabilities ascribed to a particular sensitivity coefficient by each of the 7 experts. Looking at any particular line, the graph tells you, in the opinion of that expert, the likelihood that the coefficient exceeds any particular % value. (SOURCE COMEAP 2009 Page 159)


* COMEAP 2010: Page 5: Point 18 estimate 29,000 deaths per year, but this is sometimes reported as 28,000. Given the large uncertainty in the figure, I have taken the lower estimate throughout.



Did 28,000 people die from air pollution in 2012? Part 2

May 27, 2015


Complicated graphic showing the correlation between 'excess mortality' and a pollution event in New York in 1962. The excess mortality is shaded in red and reduced mortality in following days is shaded in blue. Click for larger image - see text for details.

Complicated graphic showing the correlation between ‘excess mortality’ and a pollution event in New York in 1962. The excess mortality is shaded in red and reduced mortality in following days is shaded in blue. Click for larger image – see text for details.

 Continued from PART 1.

In Part 1 we saw that it is untrue that air pollution causes 28,000 deaths each year in otherwise healthy people.

But could these deaths be ‘linked to’ air pollution? Let’s look at the adverse effects of air pollution.

DEFRA list the effects of very high levels of some key pollutants (link):

  • NO2, SO2, O3 (Ozone): These gases irritate the airways of the lungs, increasing the symptoms of those suffering from lung diseases
  • Particles: Fine particles can be carried deep into the lungs where they can cause inflammation and a worsening of heart and lung diseases
  • CO (Carbon Monoxide):This gas prevents the uptake of oxygen by the blood. This can lead to a significant reduction in the supply of oxygen to the heart, particularly in people suffering from heart disease

So all these pollutants can be expected to harm us. But how much harm they do? This question – the quantitative expression of the harm caused by air pollution – is at the heart of this issue.

Broadly speaking, we can expect short-term and long-term effects. However, it is possible for air pollution to have both kinds of effects in the same person.

Short Term Effects

The US National Institute of Health (NIH) host this 1966 paper by McCarroll and Bradley looking for excess mortality coincident with air pollution events in New York City in 1962.The paper is notable for its readability.

McCarroll & Bradley analysed 3 pollution ‘events’ and the figure at the top of the page summarises the general nature of their analysis: they note excess mortality coincident with air pollution events, and a reduction in mortality for subsequent days. They note specifically that the drop is “…never of sufficient degree to compensate for the excess of deaths on the preceding day.”

So the general mechanism in action here is that air pollution has ‘brought forward’ people’s deaths: some by just a day or so, but others by unknown periods of time – probably weeks to months. These would be people with an underlying medical condition which itself may possibly be a long term effect of air pollution

Long Term Effects

Long-term effects are more difficult to assess since we do not have identical healthy populations that differ only in their long-term exposure to pollutants. However such effects can be estimated by so-called cohort studies.

So by analyses similar to McCarroll & Bradley and by cohort studies we can begin to estimate how much air pollution causes how much excess mortality.

Combined Effects

Short-term exposure to air pollution appears to bring forward some deaths – the actual number is unknown.

Long-term exposure to air pollution also appears to also shorten life.

For the UK population, a committee (COMEAP) estimates this to amount to 350,000 person years per year. What?

  • For a population of 60 million this amounts to about 2.1 days per person, per year of life.
  • So if the ‘natural lifetime’ of a person born now is 80 years, then exposure to typical air pollution for their lifetime is estimated to shorten their lives.
  • The shortening is calculated to be 80 years x 2..1 days ≈ 6 months less than they might otherwise have lived.

So now I think I understand the nature of this excess mortality. There are three components

  • Air pollution has adverse physiological effects,
  • People susceptible to heart attacks or with respiratory diseases (from other causes) can be brought into a cycle of distress which brings forward their death.
  • Additionally chronic prolonged exposure to air pollution can induce respiratory complaints that may make people susceptible to acute air pollution events.

Estimating and Communicating these effects

In PART 3 we will see how that estimate is made, and you can judge for yourself whether to be alarmed.






My body is a machine

July 12, 2014

This song is the best song ever written about glycolosis: the basic mechanism by which the ‘engine’ of the human body takes in ’fuel’ and enables ‘it’ to do ‘work’. 

Although I never studied it at school, I have known for a long time about the basic mechanism by which the ‘engine’ of the human body takes in ’fuel’ and enables ‘it’ to do ‘work’.

But only recently I have become aware of how the ‘engine’ of my body works, and what it feels like when it doesn’t.

As a slightly overweight (86 kg, 1.75 m) 54-year-old male I am aware that I am entering my prime heart-attack decade. Exercise is apparently the key to reducing my risk, but I have not been doing much of that lately: you know how it is.

And whenever I had tried to exercise by jogging at what felt a comfortable pace, I would find that after maybe half a mile I would find my heart racing, I would be out of breath and would need to slow down dramatically – even stop. I assumed that this was a symptom of early-onset death.

However, recently I decided to try this running lark again and popped into the local ‘Sweatshop’ where a very enthusiastic young man took pressure patterns of my feet as I stood on a glass plate, and then made videos of me running on a treadmill. All very high tech. He then sold me a pair of embarrassingly expensive running shoes.

And as an afterthought I bought a £50 heart monitor. It consists of band that goes around my chest – which detects the electrical signals associated with each beat of my heart – which is wirelessly linked to a watch which displays my heart rate.

Suitably equipped, I began to investigate how the engine of my own body was behaving.

The first number I looked for was my resting heart rate. I tried the monitor out throughout one whole day and found that – perhaps surprisingly – my lowest heart rate did not occur lying in bed in the morning (67 beats per minute or bpm) but instead at a planning  meeting at work (60 bpm). This is a healthy number and I was relieved. But maybe I need to contribute more to meetings.

All the web sites frame heart rates for exercise in terms of maximum heart rate. This number varies from person to person, and declines with age. A little reading told me that the typical value for a 54 year old male is around 170 beats per minute (bpm) And based on this I should be exercising at around 140 bpm.

And so the second  number I had to find was my maximum heart rate. This turned out be closer to 195 bpm which I think is basically a good thing. And based on my maximum heart rate, the web sites say I should be exercising at a much faster heart rate.

The sites predict that I should experience a transition from aerobic to anaerobic exercise at around 165 to 170 bpm. And when I ran I could feel that change exactly where it was supposed to be.

Using the monitor I found that if I jog at 165 bpm I don’t go very fast, but I can sing to myself and feel barely out of breath. I can basically run until I am bored.

But when my heart rate reaches 175 bpm I find myself shorter of breath and can either tough it out or relax back to a more manageable 165 bpm.

So the key to understanding my experience of what my body was doing was to make a measurement not of my speed of running – but of my heart rate. I guess it is equivalent to watching the rev-meter on a car instead of the speedometer.

And as result I have been able to run local roads on a course that lasts 5 kilometers: further than I have ever run before. And at the end I still have energy for a not very impressive sprint. (It’s not the time that I care about – its just I would like to give my neighbours the impression that I have been running at that speed all the time)

And the reason I am writing this is because the experience has left me feeling mildly empowered and slightly relieved. Understanding what I was experiencing and being able to relate it to what other people experienced was comforting. It’s just another example of Kelvin’s maxim that until you can ‘put a number’ to something, you do not truly understand it.

And the good news is that the device is telling me that, rather than being close to death, I instead have the heart of younger man. So as long as he doesn’t want it back any time soon, that’s great!

Who is going to die in 2048?

April 9, 2014
Age Standardised UK Mortality

Graph showing Age-Standardised UK Mortality per 100,000 of population per year. In 2010 mortality was around 1100 per 100,000, so for the UK population of 60 million we would expect around 660,000 deaths per year. However if the trend continues, no one will die in 2048!

While investigating causes of death in the United Kingdom, I came across the data above. The graph shows that the age-standardised mortality in the UK has been falling since at least 1980 – and shows no signs of stopping.

Indeed, if the trend continues, then sometime around the 14th March 2048, mortality will reach zero and no one will die in the UK!

Now of course, although this data is real and correct, the trend can’t possibly continue indefinitely. But the data is nonetheless fascinating for at least three reasons.

Firstly, in the face of seemingly endless stories telling us all how unhealthy we are – it seems that the trend to lower mortality is continuing unabated, despite the obesity ‘crisis’.

Secondly, although the linear trend in the data is striking, we have no justification for extrapolating the trend into the future. Why? Because its the future! And we don’t know what is going to happen in the future.

And finally, these numbers give us a scale for considering the relative seriousness of different causes of death: that was the reason I looked up the data in the first place.

I read that air pollution causes 30,000 deaths a year in the UK and that seemed a surprisingly large number. From the graph we can estimate that mortality in 2014 is approximately 1000 deaths per 100,000 of population per annum. So that that for the UK population of 60 million, this is about 5% of deaths – which still seems shockingly high, but is a smidgeon closer to believability.

So good news all round: especially if you, like me, are a man. The mortality of men and women is shown separately below.

If the trend continues, then after millennia of ‘excess male mortality’, the mortality of men should fall below that of women in approximately 2027 and reach zero in 2042 – before the women – who will not attain immortality until 2060!

Age Standardised UK Mortality by sex

Graph showing Age-Standardised UK Mortality per 100,000 of population per year for men and women. If trends continue, male mortality will fall below female mortality in 2027 and no men will die at all after 2042!


Dave asked: Are you sure age standardised mortality means what you think it does? Age standardised mortality might drop to zero. But that is not mortality. If the plot showed mortality that would suggest life expectancy has doubled since 1980, from 50 to nearly 100.

And I replied: The calculation is this:

  • How many people died in a particular year aged (say) 69.
  • This number is then expressed as a fraction of the actual UK population who were aged 69.
  • This is then expressed as an actual number who would have died in a ‘standard population’ called the European Standard Population.

This procedure allows the relative mortality in different countries to be compared

So, if for example, the UK has a high absolute mortality for 69 year-olds, but not many 69 year olds – then this will produce a larger number when ‘age standardised’.

I have obtained one or two sets of actual death data – but I don’t know the equivalent population to divide by to get the absolute mortality per 100,000. However this data shows a similar trend with roughly the same intercept.

What does it mean? I don’t know! I think it means that we are living longer (Is that news?). I was just struck by how straight the line was and how it begged to be extrapolated!

Float or Sink: Don’t believe everything you see on the internet

January 15, 2014
Wow! A can of Diet Coke floats and a can of regular Coca Cola sinks! Mmm. Don't believe everything you see on the internet.

Wow! A can of Diet Coke floats and a can of regular Coca Cola sinks! Really? Well No. Like so many things you see on the internet, it is not so simple.

While discussing recent news stories on the dietary ‘value’ of sugar, a colleague told me he had seen a surprising video on the internet.

It showed Steven Spengler putting cans of Coca Cola and Diet Coke into a tank and the can of Diet Coke floated while the can of Coca Cola sank. The celebrity scientist then related this to the amount of the sugar in the drink. See what you think:

I told my colleague that I didn’t believe anything I saw on the internet, but that the demo was convincing: it seemed as though nature itself was voting on the evils of sugar. A sort of ‘Witch Trial’ for harmful additives.

However being the person I am, I thought I would just check. I bought a can of each drink at lunchtime, went to the lab and did the test. This is what I saw:

This is what I saw when I put the two cans into water: they both floated.

This is what I saw when I put the two cans into water: they both floated.

Yes, that’s right: both cans floated. If you look closely you can see that the Coca Cola is lower in the water, but it did not sink. So how did I get the picture at the top of the page? Simple: I heated the water.

The density of water falls slightly with increasing temperature (See the graph at the bottom of this article) and when the water was around 36 °C at the top and 32 °C at the bottom, the Coca Cola sank. But when the water cooled to between 33 °C at the top and 30 °C at the bottom, the Coca Cola floated.

I made measurements of the mass of the cans, full and empty and found that the 330 ml of Coca Cola weighed 340.2 g while 330 ml of Diet Coke weighed 330.8 g. And hence I made an estimate for the density of the two fluids, and yes, Coca Cola appears to be 2.8% denser that Diet Coke.

But then the label tells you that 330 ml of Coca Cola contains 15.9 g of sugar, whereas Diet Coke contains Aspartame which is weight-for-weight 200 times sweeter. So there is only about  0.07 g of Aspartame in a can of Diet Coke. So this isn’t really ‘news’ of any kind.

I couldn’t understand the exact 9.4 g difference in mass because the density of the fluids is affected by all the other components ingredients which could differ between products.

However both fluids were denser than water (0.5% and 3.3.% respectively) . And whether a can floats or sinks depends only partially on the density  of the liquid in the container.

It also depends on the ‘air’ gap, and the weight of the can. So Steven Spengler’s demo just relies on a simple coincidence between the average density of the US-size cans and the density of water at about room temperature.

For larger containers, the mass of the container will make less difference and so I thought that for a 2 litre PET bottle, both fluids would probably sink. Was I right? No.

The Coca Cola weighed in at 2.129 kg and the Diet Coke was 85 g lighter at 2.044 kg. But in fact these bottles have a larger air gap and so – even when heated to 48  °C – both bottles floated.

2 litre bottles of Coca Cola and Diet Coke both float in water - even when heated to 48 Celsius.

Two 2 litre bottles of Coca Cola and Diet Coke floating in water – even when heated to 48 Celsius. The blue lines scrawled on the photograph show the water level around each bottle and the value on the thermometer.


The density of water plotted as a function of temperature and fitted with a quadratic polynomial.

The density of pure water plotted as a function of temperature and fitted with a quadratic polynomial. The entire vertical range represents just 1% change of the density.

Are you sure you want to lose weight?

January 4, 2013
The relative risk of having a Body Mass Index in different ranges. Being overweight or being obese (Class I) results in a lower risk of death than being 'normal'.

The relative risk of having a Body Mass Index in different ranges. Being overweight or being obese (Class I) results in a lower risk of death than being ‘normal’. The error bars represent a 95% confidence interval.

Welcome to 2013! It may be that you – like me – are considering losing a little weight. But if you are considering a diet or exercise regime, then you should be aware that a new study confirms that although you may feel better, you will also be increasing your risk of an early death.

I have written about this before (here), but this new study by Flegal et al (available free here) confirms this result a fortiori . The work is a ‘meta’ study covering more than 2.8 million people on several continents – mainly Europe and the US – and has recorded more than 270, 000 deaths. Thus the relative risks of being overweight can be isolated from many other causes. The results (see the Figure above) confirm that being ‘overweight’ reduces the relative risk of death (the Hazard Ratio) compared with being ‘normal’. They even confirm that being Class I obese still offers a protective effect. Higher levels of obesity increase the risk the of death significantly.

This result is important in a social context where right-wing think tanks suggest that the benefits of overweight unemployed people should be cut if they fail to exercise (Guardian, BBC). Their aim – no doubt laudable – is to reduce the burden of obesity-related disease on the National Health Service. However the data here indicate that the consequence would be to increase the risk that these people would die. This is not a morally defendable position. Indeed, this work seems to me to call into question every piece of government advice on the topic of obesity.

The Journal of the American Medical Association has an accompanying editorial discussing the work:

“The optimal BMI linked with lowest mortality in patients with chronic disease may be within the overweight and obesity range. Even in the absence of chronic disease, small excess amounts of adipose tissue may provide needed energy reserves during acute catabolic illnesses, have beneficial mechanical effects with some types of traumatic injuries, and convey other salutary effects that need to be investigated in light of the studies by Flegal et al and others.”

Not all patients classified as being overweight or having grade 1 obesity, particularly those with chronic diseases, can be assumed to require weight loss treatment. Establishing BMI is only the first step toward a more comprehensive risk evaluation.

In other words being marginally overweight is a good thing if you fall ill – because you can draw on the energy in the fat if you are not eating. However its also states that there are “other salutary effects that need to be investigated in light of the studies by Flegal et al and others.” In other words “We don’t understand this, but its real”.

My body mass index is around 27 kg/m2, close to the UK average so I am technically overweight. I also feel overweight. And I understand that this increases my risk of disease (morbidity). However, it also reduces the risk that I will die from it (mortality). Nothing in life is as simple as we would like it to be!

Happy New Year.

The incredible lightness of being wrong

July 18, 2012
BBC GLobal Fat Scale

Where I lie on the BBC Global Fat Scale. Under average for the UK – over average for the world. But the text in the ‘info’ box (shown enlarged in the figure below) has  been changed. Click for larger image.

One of the most interesting features of modern media is that if one makes a mistake, it can be corrected quickly and the original error then disappears. Anyone who accidentally witnessed the transient error is then left with a feeling of bewilderment when they try to show the mistake to their friends. They may even experience a feeling of paranoia. I take advantage of this feature regularly on this blog, and the more august BBC took advantage this week when it erred in an article on Body Mass Index (BMI).

Both myself and one of my international network of informants noticed that the ‘hover over’ information text stated that having an above ‘normal’ BMI caused increased mortality – an increased risk of death. As I mentioned in a previous article, this appears to be not true. In fact being ‘overweight’ (BMI 25 to 30) appears to reduce mortality. Even being ‘obese’ (BMI 30 to 35) appears to give a reduced risk of death. Before I could write to the BBC about this, they removed the offending text, leaving me and my informant scratching our heads.

BBC GLobal Fat Scale

Detail from the image at the head of the article.

This matters. If the BBC, or the government, or a health authority, encourages people to move from the ‘obese’ or ‘overweight’ categories to the ‘normal’ category, then they are encouraging them to die earlier. Generally we consider ‘dying younger’ to be a ‘bad thing’. But that is what the statistical evidence indicates and that presumably is why the BBC chose to remove this reference.

There are two lessons to learn from this episode.

Firstly, reputable media, such as the LA Times, make a note on a page to record the fact that it has been updated since publication. This should be routine in a public organisation such as the BBC. They should not try to re-write history.

Secondly, the article in itself is profoundly flawed. It ranks a large number of countries but fails to note that the bottom half of the list are countries in which the populace is profoundly and tragically undernourished. The mortality data indicate that the ‘normal’ BMI categories may well have been drawn up in a time when our population was malnourished. Being overweight may not the most beautiful condition, but the data tells us that in terms of living longer, its the best condition to be in. And that is something this chubby 52 old (BMI = 27) feels pretty good about.


Published first at 12:09 a.m. on the 19th July 2012


July 1, 2012
BMI Mortality

The upper curve shows the relationship between body mass index (BMI) and mortality expressed as relative risk of death compared to those with a BMI in the range 22.5 to 25. The data indicates quite robustly that being in the overweight category is ‘protective’. The data point at BMI=18 includes data for all people with lower BMI and the data point at BMI=35 includes data for all people with higher BMI. The lower curve shows the distribution of BMI in the population used for the study.

Friends, I am concerned about my weight. In particular I am concerned about whether being overweight (i.e. with a body mass index or (BMI) in the range 25 to 30) is genuinely bad for me, or whether it just makes me feel bad.

I have looked at this issue before and expressed my puzzlement at how ‘normal’ ever came to be defined as having a BMI in the range 20 to 25, when as far as I could tell, it has never coincided with the central range in the population.

My puzzlement appears to be vindicated by research which shows that the relative risk of death – mortality – is lower for people who are overweight compared with people in the ‘normal range’. These conclusions are backed up by other studies. But even so, it is important to understand how the research was done in order to appreciate what it really tells us.

The research followed 11,834 individuals in Canada from 1994/5 to 2006/7, and saw how mortality was affected by their BMI at the start of the study in 1994/95. Let me stress this. It saw how the BMI statistic in 1994/5 affected the rate at which they died in the subsequent 12 years. This large population included men and women, smokers and people who never smoked, and people in all age groups.

Could the ‘BMI effect’ have been protective in young people, but harmful in older people? This might make sense, since young people are less likely to die in any case. This might have masked the effect that I would have expected to see: that being overweight was harmful. The researchers controlled for that and looked at how the relative risk of death varied with BMI categories for various sub-populations within the group. Surprisingly – to me at least – the effect was seen in all categories.

BMI Mortality versus age

BMI Mortality versus age for different subpopulations within the study. The risk of death is relative to those with BMI in the ‘normal’ range. The lowest and highest BMI points include data for all individuals at lower or higher BMI. I have missed out the confidence indicators (error bars) because they make the graph too confusing.

Mortality and Morbidity. This report recorded the BMI of a population at a point in time, and studied correlations between the BMI and the rate at which people died in the following 12 years: This is called the risk of mortality. It did not record whether the individuals concerned became ill or unwell, and did not study how their BMI affected their chance of becoming unwell – that is called the risk of morbidity. I do not have data to hand but I would be pretty sure that being in the ‘overweight’ category (as defined by a BMI in the range 25 to 30) would be a significant risk factor for diseases such as cardio-vascular disease and type II diabetes.

This data was taken amongst a population which, for possibly the first time in human history, has enjoyed essentially unrestricted access to food for several generations. This is an astonishing cultural achievement. If the data really does fall in this way for other populations – including that in the UK – it will be very interesting to understand why that occurs. It will also be interesting to hear doctors argue that ‘chubby’ people like me should lose weight – something which will increase my risk of death!

What am “I” ?

May 9, 2012
A portrait of the artist as a young man.

A portrait of the artist as a young man. This is a photograph of the collection of atoms which went by the name Michael de Podesta circa 1980. In the same way that Manchester United still exists, but no longer has any players from the 1980’s, so Michael de Podesta ‘exists’ but retains none of these particular atoms or cells. What does that mean?

I remember being an undergraduate.  And at about the time the image above was captured, I recall imagining being very tiny – smaller than an atom – and moving through the atoms of ‘my body’. I remember asking: How would I identify the edge of ‘me’?

On this tiny scale, the nuclei of atoms would be far apart, and so I would barely be able to tell whether I was inside an atom, or in-between atoms. The edge of ‘me’ would be very hard to detect: it would correspond to nothing more than a relative decrease in the frequency with which I came across a nucleus. And also to changes in the types of nucleus I encountered – with many fewer carbon nuclei in the bits that were not ‘me’.

The upshot of my imaginings was that I decided that the concept of an ‘individual’ was one that only made sense when viewed on a large scale. On a small scale, all one would see would be essentially chaotic variations. This was a pivotal moment for me,

From this journey of my imagination I understood (pace any high energy physicists out there) that in same way that a microscopic explorer would never deduce the existence of the phenomenon I called ‘me’, so looking at the components of matter would never – even in principle – result in a ‘theory of everything‘.

I was reminded of this insight today while reading an article in the Scientific American on bacteria. First I was shocked to find that: In the human body, bacterial cells outnumber human cells 10 to one. The article then explained that many dis-eases were not caused by ‘foreign’ bacteria, as had at first been imagined. Instead many dis-eases arose from a disturbance of the ”social’ balance between the vast populations of bacteria that live on (and within) our bodies.

For example, the microbe Heliobacter Pylori is associated with ‘causing’ stomach ulcers, but in normal life Heliobacter Pylori lives happily in our stomach. And it helps us by producing the hormone ‘ghrelin‘ which produces a sensation of ‘satiation’ after eating. Similarly, the bacterium Bacteriodes fragilis stops inflammatory T cells from damaging our bodies, and ts absence could be associated with auto-immune diseases.

And so biology brought me back to same question that physics had raised 30 years ago. What do I mean when I use the word ‘I’? Physics told me that ‘I’ was only blearily separated from the atoms around ‘me’. And now biology tells me that ‘I’ am a combination of my genetically unique cells, and a vast army of bacterial ‘hangers on’ that have come to live with ‘me’ essentially by chance.

The concept that dis-eases were not caused by individual poisonous foreigners, but instead resulted from an imbalance in a diverse social network resonated strongly with me. And so I thought I would share that with you. Goodnight.

See and links within for further details.

%d bloggers like this: