What Kiwis die of – Part II: The battle of the sexes

By Siouxsie Wiles 31/01/2013

Two weeks ago, we asked Kiwis what they thought they would die of, and compared their responses to the primary causes of death recorded by the Ministry of Health for the 29,204 people who died in New Zealand in 2009. Mike Dickison produced a beautiful infographic* of the causes that were responsible for more than 1% of those deaths. Obviously this data reflects what was recorded on the death certificate which may not always be the immediate cause of death. In fact, there was some discussion afterwards about the reality of ‘old age’ being an unacceptable cause of death here!

This week we decided to take a look at differences for causes of death recorded for men and women. Mike has worked his magic again and produced the infographic below. This time the little coffins each represent 25 people.

Its striking that more men die of prostate cancer than women die of ovarian, and twice as many men than women die from cancer of the bladder and kidney. But lots more women die of cerebrovascular diseases, that is strokes and brain haemorrhages, and dementia. To see if people’s perceptions tallied up with reality, we took to social media and asked people whether they thought there were any differences between what men and women died of**. So what did the Kiwis we surveyed think? Results are summarised in the slideshare presentation below.

[slideshare id=16270672&doc=whatkiwisdieofii-gendersurvey-130131024841-phpapp02]

The first striking thing to come from our survey participants is that the majority thought that more men die than women, and that more men die of diabetes, cerebrovascular diseases, traffic accidents and suicide. But not skin cancer. In actual fact, the numbers of men and women who died in 2009 was very similar, 14,615 men and 14,589 women. As you can see from the infographic, more men died in traffic accidents and diabetes, or committed suicide. But our respondents were way off with cerebrovascular diseases. And with the fact that skin cancer killed more men than women. The other question we asked related to breast cancer deaths in men. 8.2% of our respondents thought that men couldn’t get breast cancer. In fact, in New Zealand in 2009, 1 in 100 breast cancer deaths were in men.

As several people pointed out with our first infographic, these differences between perception and reality could be dangerous if they mean resources are diverted away from where they are really needed.

*Head on over to Mike’s Pictures of Numbers blog to see how the infographic came about and for a downloadable version.

** We used Twitter, Facebook, and email to entice over 100 people to fill out our survey using the SurveyMonkey website. Like all online surveys, ours should not be considered representative of the population of New Zealand, but rather of people on social media who were happy to fill out our survey.

This post is the second in a series inspired by the Guardian newspaper’s infographic ‘What we die of’ and is a collaboration between myself, chief number cruncher Dr Paul Gardner and data visualisation extraordinaire Dr Mike Dickison. Dr Paul Gardner (@ppgardne) is a Royal Society of NZ Rutherford Discovery Fellow and Senior Lecturer in Bioinformatics at the University of Canterbury’s School of Biological Sciences. He gets very excited about RNA. Dr Mike Dickison (@adzebill) is a freelance information designer with a PhD on the evolution of giant flightless birds. He quite likes ukuleles too. Dr Siouxsie Wiles (@SiouxsieW) is a Health Research Council of NZ Hercus Fellow at the University of Auckland’s Faculty of Medical & Health Sciences. She is rather keen on nasty bacteria and anything that glows in the dark.

0 Responses to “What Kiwis die of – Part II: The battle of the sexes”

  • That data is very interesting and the presentation is excellent and illustrates very well the differences. I’d love to see it done with incidence rates as well. There’s many nuances there, such as while men complete suicide more often, women attempt more frequently (incurring increased morbidity rather than deaths, including liver and renal damage) and that sort of thing doesn’t always get focused on and so it seems to me to be part of the picture missing when focusing on only one parameter.

    One thing I’d point out is that Prostate cancer isn’t analgous to Ovarian cancer, while the Prostate is part of the reproductive system the comparasion should be with Testicular cancer. Looking at the data, it’s not clear whether it’s been left out as it’s selective data or whether it’s less than 1% of deaths but it would have been interesting to check the death rates on that considering Ovarian cancer is an ‘silent’ cancer and generally is only found when extensive so mortality rates are high – according to the figures I’ve found for cancer registrations put out by the Cancer Society in 2000 there were 301 cases of Ovarian cancer with 179 deaths, for the Testis, there were 143 cases with only 7 deaths.

    The better comparasion would have been with Breast cancer, this is the leading site for cancer registrations as Prostate is for men, both involve glandular type tissue and both genders have them (FYI the direct comparable site to the Prostate is the Paraurethral gland or Skene’s gland in women).

    With the Kidney and Bladder cancer that looked as if that occurred at about twice the rate in men, so you’d expect to see that reflected in mortality figures. The Cancer Society data I looked at pointed out that about 80% of cancers are preventable, with smoking a major one but also sun exposure. A lot of education has gone on to address these and this should have an effect on later incidence and mortality. Of late, the death stats show men catching up with women in the longevity stakes, there has been more men ceasing smoking (it used to be an unacceptable thing for women to do in general, so there will be changes as for a while, women increasingly took up the habit), improvements in care for cardiovascular disease has meant what used to be a leading killer in middle age isn’t now, and there are less men working in heavy industry. e.g. “Based on 2007-09 mortality rates, a girl born today can expect to live, on average, 82.4 years while a boy might reach 78.4 years. “Life expectancy for women is still higher than it is for men, but the gap has narrowed from more than six years in 1975–77 to four years in 2007–09,” Population Statistics manager Denise McGregor says. Men are dying younger than women but their longevity has increased more dramatically since the late 1970s. “Since 1975–77, life expectancy at birth has increased by 6.9 years for females and 9.4 years for males,” says Mrs McGregor” http://www.everybody.co.nz/page-b2ee06d4-df2e-4bbe-a83c-9bc126beb0c3.aspx

    This could mean in later years things could change dramatically if the trend lasts. One thing that is a big concern is the higher rates of death from both hypertension and cerebrovascular disease in women – you could almost pull the two together as hypertension is the most important cause of cerebrovascular disease. Given that the disparity there is stronger, much stronger than with Diabetes it would seem to be clear that a lot more attention needs to be paid to hypertension and other risk factors for cerebrovascular disease in women

  • I’d be interested in seeing the data with old age taken out of the picture. Immediate cause of death for someone over 90 is not as relevant as for someone under 70. Something is going to kill you. If you live long enough, then it is likely to be “cerebrovascular diseases”. Women on average live longer than men so this will weight that category in their favour. I’d define a cutoff for old age (eg 85 yrs, 90 yrs, 1std deviation above average), and re-check the data to see if there are age related biases there.

  • WRT your first question – I saw it as a trick qn.
    100% of men and 100% of women die.
    If the birth rate of males and females is the same then the exact same number of males and females die! (or if the same proportion reach the age at which they become classified as “man” or “woman”).

    One of the things with the MoH data is that it is dependent on what the doctor decides to put on a death certificate. Consider this: Diabetes is a strong risk factor for End Stage Renal Disease (ESRD); ESRD is a strong risk factor for cardiac arrest (not to mention they will be hypertensive as well). If a person dies of a cardiac arrest who was hypertensive, in ESRD, and with diabetes what is the cause of death?

    As one funeral director said to one vicar as they followed the coffin out of a church in response to the vicar’s question “What did she die of?” – “Her heart stopped.” (true story).

  • I don’t have incidence rates for 2009, but there were only 4 deaths from testicular cancer that year (compared with 562 from prostate cancer), so I left it off the graphic. It’s great to have informed commentary from medical folks about the complexities involved in ascribing cause of death!

  • ” Something is going to kill you. If you live long enough, then it is likely to be “cerebrovascular diseases”.

    IMHO I wouldn’t be keen on taking out older age cohorts, people are living longer (and I’d say healthier lives) in general and what they die of is still important and would have implications for such things as planning/providing health services or deciding on areas of funding for health. That group deserves just as much recognition as middle aged men and women, and in my earlier post I pointed out that men are gaining much more in longevity than women and are set to catch up in the near future. The current four year gap can’t actually explain the large difference between deaths from both Hypertension and Cerebrovascular disease between men and women because the difference in longevity is not that large (and the trend is over time time for it to lessen) nor is Cerebrovascular disease primarily/exclusively a disease of very aged people.

    Cerebrovascular disease, in particular Strokes, cause a lot of deaths in younger people (figures come from the World Stroke Day 2010 information):
    – Second leading cause of death for people above the age of 60.
    – Fifth leading cause in people aged 15 to 59.

    It’s a health issue all of it’s own, and seeing it stand out there with deaths does make me wonder if there is a problem with lack of recognition and therefore timely preventative treatment such as good management of Hypertension. Rather than discarding data, it would be better to add it in, see which age cohorts are dying of what, compare it with incidence and morbidity rates (if available) from selected disorders, whatever suits.

    As for how death is certified, read this document – http://www.health.govt.nz/publication/guide-certifying-causes-death . If death is caused from one thing, such as bacterial meningitis that may be recorded as that. If it is not, and say Type 2 diabetes complicated by ESRD the format is to put the the disease, injury or complication (such as sepsis) which directly preceded death, then if that first part was caused by any antecedent cause, this is entered, then any other antecedent cause or causes. They do take into account judgement of the most important antecedent condition that contributed to death. This means primary and secondary causes of death and the complications that directly caused death are all recorded and become part of the data. Definitely not a subjective exerciseby the doctor certifying death.