GEOG 20016 Fertility, Mortality and Social
Change Health and mortality transitions A/Prof Tim Adair The Nossal Institute for Global Health Melbourne School of Population and Global
Health Introduction Why is mortality of interest to demographers? • Measuring the level and patterns of mortality has been a
focus of demography since its inception. ➢Work of John Graunt in the 17th century in London – weekly reports called the Bills of Mortality – measured
how many people were dying and what from ➢Warned of the bubonic plague and other infectious diseases • Mortality change → more moderate impact on population
growth and age composition compared with fertility. ➢However, increasing longevity is a major reason why
population aging is a significant issue throughout much of the
world. Definitions Mortality, ie deaths, has a clear definition and occurs
at a distinct point in time. • What causes death is however more difficult to
determine. Morbidity: the measurement of disease, injury and
disability. Morbidity is defined as “any departure,
subjective or objective, from a state of physiological or
psychological well-being” (Porta 2014: 189). It is one thing to prevent people from dying, but are
they living healthy lives? Porta, M., (ed.), 2014. A Dictionary of Epidemiology, (sixth edition), Oxford
University Press, New York. Health and mortality transitions • Mortality and morbidity – Indicators – Data sources – Theory – Global patterns and trends – Future prospects Key mortality indicators Crude death rate The crude death rate is the number of deaths in a
defined period (usually a calendar year) per 1,000
people. Crude death rate = Number of deaths in calendar year Midyear population
∗
1000 Key mortality indicators Age-specific death rate The age-specific death rate is the crude death rate for a
specific age group. Age-specific rates are usually expressed per 1,000 or
100,000 population in the age group. Important to understand the likelihood of people dying
at different ages. Age − specific death rate, 25 − 29 in 2020 = Deaths aged 25 − 29 in 2020
Population aged 25 − 29 at 30 June 2020
∗
1000 Key mortality indicators Early age mortality indicators Under five mortality rate = Number of deaths of children aged less than 5 years Live births
∗
1000 Infant mortality rate = Number of deaths of infants aged less than 1 year Live births
∗
1000 Or… the probability of dying from birth until age 1 year, or age 5 years. Key mortality indicators Life expectancy The average number of years of life a population is
expected to live, based on current mortality conditions – Normally calculated as life expectancy at birth – But useful to understand life expectancy at other ages, eg
life expectancy at 65 years is important for designing
retirement income systems Key mortality indicators Years of life lost (YLLs) A measure of premature mortality. It is calculated as the
difference between age at death and the longest
possible life expectancy for a person at that age. ➢If someone dies at age 35, it contributes more to YLLs
than someone dying at age 85. Where do mortality data come from? In higher-income countries, vital registration systems record all deaths
and provide (mostly) high quality data on the cause of death. – Doctor completes a medical certificate of cause of death, from which the
underlying cause of death is assigned. – Underlying cause is most useful for health policy purposes – it is what
initiated events leading to the death But around one-third of all deaths in the world are not registered, and
less than half of all deaths have an accurate cause of death. In these countries,
surveys, censuses
and statistical models
are relied upon for
mortality data. Life expectancy has increased substantially
almost everywhere in the past 100-150 years Source: Dattani, S, Rodés-Guirao, L, Ritchie, H, Ortiz-Ospina, E & Roser, M (2023) – “Life Expectancy” Published online
at OurWorldInData.org. Retrieved from: ‘https://ourworldindata.org/life-expectancy’ [Online Resource] Global life expectancy GBD, 2018 Long-term life expectancy trends Long-term increases in life expectancy – complex
interaction of: – socio-economic factors, such as increased income,
economic development, education (especially maternal)
and urbanization – also more proximate behavioural, environmental and
medical factors such as improved nutrition and sanitation,
reduced tobacco consumption and medical advances
(prevention and treatment) (Gu et al. 2013, Oeppen and
Vaupel 2002). Rectangularisation of mortality Adapted from Wilmoth & Horiuchi (1999) FUTURE??? Epidemiological transition Describes global long-term trends in life expectancy and
causes of death (McKeown 2009, Omran 2005), similar
to demographic transition. Age of pestilence and famine: – Mortality is high (especially in childhood) and life
expectancy fluctuates between 20 and 40 years due to
droughts and famine. – Most of human history has been spent in this stage. – Deaths were primarily caused by infectious disease such
as water- and food-borne diseases, the plague and
tuberculosis. Epidemiological transition Age of receding pandemics: – Mortality began to decline, especially in children, and life
expectancy increased to around 50 years. – Food supply increased due to agricultural productivity and
nutrition improved, while sanitation improved due to better
water supply, sewage, and food handling. – Child mortality reductions were caused by declines in
infectious diseases, female literacy and public health
education programs (the latter from the late 19th and early
20th century). Epidemiological transition Age of degenerative and man-made diseases: – Mortality continued to decline, including at older ages, and life
expectancy rose to 70 years. – Non-communicable diseases such as heart disease and cancers
caused an increasing proportion of deaths. – This phase occurred in the mid-20th century in Western countries
and is present in many present-day developing countries. Age of delayed degenerative diseases (fourth stage?) – further decline in death rates and ages at death become even
older – rectangularisation of mortality – increased prevalence of degenerative diseases such as
Alzheimer’s and Parkinson’s disease
– found in many Western countries today What about COVID-19 and other (re)-emerging
infectious diseases? Shift in causes of death, 1990-2019, worldwide 1990 2019 GBD, 2020 Ratio Group 2 : Group 1 = 4.13 Life expectancy = 73.5 Ratio Group 2 (non-comm.) :
Group 1 (comm.) = 1.76 Life expectancy =
65.1 33% 58% 9% Communicable Non-communicable Injuries 18% 74% 8% Shift in causes of death, 1990-2019, worldwide GBD, 2020 Stage of epidemiological transition in various
regions of the world 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% %
die th
from ro m
each
cause
grow up Region,
Ratio of Non-communicable : Communicable deaths Injuries Non-communicable Communicable GBD, 2017 What are the risk factors for mortality? Risk factor: “attribute, characteristic or exposure of an individual
that increases the likelihood of developing a disease or injury”
(WHO 2016) – these inform policy interventions. Leading risk factors for mortality globally are:
1. High blood pressure 2. Tobacco smoking 3. Poor diet 4. Air pollution 5. High blood sugar levels 6. Obesity/overweight 7. High cholesterol 8. Kidney dysfunction 9. Child and maternal malnutrition 10. Alcohol use GBD, 2020 Smoking and lung cancer, Australia Trends in tobacco consumption and respiratory or lung cancer
(age-standardised death rate), males, Australia, 1887-2006 Adair et al., 2011, Reconstruction of long-term tobacco consumption trends
in Australia and their relationship to lung cancer mortality Future prospects for longevity Much debate about how quickly life expectancy is likely to rise in coming
decades Useful to examine historic trends in the record level of life expectancy Since the mid-19th century the global record level of female life expectancy has
increased at a remarkably steady rate of ~3 months per year Roser, M. (2020) – “The rise of
maximum life expectancy”
Published online at
OurWorldInData.org. Retrieved
from:
‘https://ourworldindata.org/the- rise-of-maximum-life-expectancy’
[Online Resource] Future prospects for longevity Oeppen and Vaupel (2002): record level of life
expectancy will continue to rise in the future at the same
pace as in the past Life expectancy not approaching a ceiling because no
slowdown in the lowest-mortality countries People reaching older age with lower prevalence of
disability, while there will be future developments in
technology and biomedicine. Past pandemics such as HIV/AIDS are past their
mortality peak. Historic under-estimation of future life expectancy
increases. Future prospects for longevity However: • Other researchers are more pessimistic • Increasing prevalence of obesity (especially in the
USA) and its demonstrated impact on increasing
death rates • Effect of high smoking levels in large populations
such as China and Indonesia. • Challenges in continuing the declines in infectious
diseases • Emergence of respiratory infections such as COVID- 19 • Re-emergence of diseases that had been eradicated
(cholera), anti-microbial resistance USA – falling life expectancy falling in some
counties from 1980-2014 Dwyer-Lindgren et al. (2017), Inequalities in Life Expectancy Among US Counties, 1980 to 2014: Temporal Trends and Key Drivers,
JAMA Intern Med What about COVID-19? The COVID-19 pandemic has
resulted in a major mortality
“shock” in many countries where
low and continually falling
mortality rates have been
norm for decades. Scholey, J, et al, 2022, Life expectancy changes
since COVID-19, Nature Human Behavior Change in life expectancy in 2020-21 Future prospects for longevity A more measured view is that historic declines in
mortality among children and young adults cannot
continue, so life expectancy is unlikely to continue to
rise at 3 months per annum (Bongaarts 2006). ➢Potential for further pandemics ➢Ever-present risk of mortality from natural disasters and
conflict ➢Impact of climate change.
But improvements in biotechnology, preventive and
curative medicine and drug treatment should lead to
reduction in mortality at older ages and support future
increases in life expectancy (Bongaarts 2006, 2014). Are people living healthier lives? Although (almost all) populations are living longer, are
they living healthier lives? As people increasingly live to older ages, they are likely
to be at risk of longer periods of morbidity. Increasing proportion of deaths from chronic diseases,
which typically affect a person for significant periods of
time before they die. Consequences for individuals, their families and care- givers, governments and health care providers. Morbidity Traditionally, demographers have typically studied
mortality rather than morbidity ➢Mortality is a component of population change and
therefore is a key determinant of the size, growth and
structure of a population. Easier to measure mortality – clearer definition, more
data Morbidity data do exist but are commonly from
disparate sources which are not necessarily complete
nor use consistent definitions.
But morbidity is a key measure of population-level
health
Key morbidity indicators Incidence: “the number of instances of illness
commencing, or of persons falling ill, during a given
period in a specified population” (Porta 2014: 144). Prevalence: “the total number of individuals who have
the condition (eg, disease, exposure, attribute) at a
particular time (or during a particular period) divided by
the population at risk of having the condition at that time
or midway through the period” (Porta 2014: 223). Porta, M., (ed.), 2014. A Dictionary of Epidemiology, (sixth edition), Oxford
University Press, New York. Key morbidity indicators Years lived with a disability (YLDs): provides a better understanding of the severity of the morbid
condition.
prevalence of the condition (ie illness, injury or disability)
multiplied by a disability weight that measures the
severity of that condition (Vos et al 2012). Disability weights: the extent of “health loss” associated with
a certain condition (Salomon et al 2015).
Scale from 0: state of full health to 1: death
For example, stroke with severe long-term consequences is
rated as 0.552 and mild hearing loss is rated as 0.010
(Salomon et al 2015).
Indicators that combine morbidity and
mortality Disability-adjusted life years (DALYs): – Measure of total disease burden in a population – The sum of: • non-fatal burden, as measured by YLDs, and • fatal burden, as measured by years of life lost (YLLs) – a measure
of premature mortality. At each age, they are calculated as the
difference between age at death and the standard life expectancy
at that age. – The number of DALYs is the number of years lost in the
population due to living with a disability and early
death. Indicators that combine morbidity and
mortality Health-adjusted life expectancy (HALE). – Measures the average length of time a person can expect
to life without illness, injury or disability. – Similar to the conventional life expectancy measure, but
accounts for morbidity (GBD 2016 DALYs and HALE
Collaborators 2017).
– The difference between life expectancy and HALE is
termed non-healthy life expectancy. Morbidity data sources Wide range of potential data sources to use to measure
morbidity. Challenges are that much morbidity is
“hidden”, the consistency of definitions of different
conditions, and how to measure the severity of different
conditions: – Hospital data – Disease registries – Health surveys – Epidemiological studies – Administrative data Morbidity rates have been declining GBD Collaborators (2020) Age-standardized YLD rate (per 100,000), by sex, global,
1990 and 2019 10,141 9,822 12,070 11,719 11,105 10,773 0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 1990 2019 1990 2019 1990 2019 Male Female Both YLD
ra te Leading causes of morbidity differ from those
for mortality GBD Collaborators (201) Leading causes of YLDs (age-std. rate per 100,000, GBD cause level 3), by sex, global, 2019 Males Females Cause YLDs Cause YLDs Low back pain 671 Low back pain 884 Age-related and other hearing loss 524 Gynecological diseases 731 Diabetes mellitus 475 Headache disorders 725 Depressive disorders 452 Depressive disorders 702 Headache disorders 440 Other musculoskeletal disorders 566 Other musculoskeletal disorders 362 Age-related and other hearing loss 476 Dietary iron deficiency 299 Dietary iron deficiency 469 Falls 275 Anxiety disorders 445 Anxiety disorders 275 Diabetes mellitus 413 Road injuries 266 Oral disorders 302 Oral disorders 265 Neck pain 299 Blindness and vision loss 264 Endocrine, metabolic, blood, and
immune disorders 294 Neonatal disorders 248 Blindness and vision loss 291 Chronic obstructive pulmonary disease 247 Osteoarthritis 265 Neck pain 235 Falls 252 Compression of morbidity A reduction of the average proportion of a person’s life
that is spent in a state of morbidity (Porta 2014: 53). That is, if people live longer, they are also spending a
smaller % of their lives in ill health/disability. Evidence generally suggests that compression of
morbidity is not happening. Morbidity comes from a wide range of conditions that
have not benefitted from large-scale preventative
measures (Salomon et al 2012).
Non-healthy life expectancy has been increasing
with longer life expectancy – shows that
compression of morbidity has not been happening GBD HALE Collaborators (2020) HALE, non-healthy life expectancy, life expectancy and % of life expectancy that is non-healthy, by sex, global, 1990 and 2019 55.8 62.6 58.2 64.9 7.2 8.4 9.8 11.2 63.0, 13.0% 71.0, 13.4% 68.0, 16.8% 76.1, 17.3% 0 10 20 30 40 50 60 70 80 90 1990 2019 1990 2019 Male Female Y ea rs HALE Non-healthy LE Non-healthy life expectancy is longer in
countries with longer life expectancy HALE, non-healthy life expectancy and life expectancy, by sex,
global, 2019 SDI: Socio-Demographic Index.
GBD Collaborators (2020) 56.2 59.7 64.0 66.1 68.2 57.4 61.0 66.4 69.3 70.2 7.7 8.1 8.1 8.4 10.4 9.9 10.7 11.0 11.3 13.5 63.9 67.8 72.1 74.4 78.6 67.3 71.7 77.4 80.6 83.8 0 10 20 30 40 50 60 70 80 90 Low SDI Low-middle SDI Middle SDI High-middle SDI High SDI Low SDI Low-middle SDI Middle SDI High-middle SDI High SDI Male Female Y ea rs HALE Non-healthy LE Within countries, morbidity is common
lower in higher socio-economic groups YLD age-standardized rate by socioeconomic group (per
1,000), Australia, 2018 AIHW (2021) 111 106 98 95 82 98 0 20 40 60 80 100 120 Q1 (lowest) Q2 Q3 Q4 Q5 (highest) Australia YLD
ra te Socio-economic group Indigenous Australians have far higher
morbidity rates than non-Indigenous Cause Indigenous
YLD rate Non-Indigenous YLD
rate Rate
difference % of YLD
health gap Cardiovascular 9.1 5.3 3.8 4.2 Mental/substance
abuse 54.7 22.9 31.8 35.4 Injuries 13.0 3.5 9.4 10.5 Respiratory 23.2 11.3 11.9 13.3 Cancer 1.7 2.1 –0.4 –0.5 Endocrine 4.8 2.0 2.9 3.2 Kidney/urinary 4.9 0.7 4.2 4.7 Neurological 14.4 6.7 7.7 8.6 Gastrointestinal 2.4 2.4 0.0 –0.1 Musculoskeletal 29.0 21.3 7.7 8.6 Infectious diseases 4.3 1.1 3.2 3.6 Infant/congenital 1.6 0.8 0.7 0.8 Blood/metabolic 2.9 0.8 2.1 2.3 Hearing/vision 6.4 3.9 2.4 2.7 All other diseases
groups 11.3 9.0 2.3 2.6 All diseases 183.6 94.0 89.6 100.0 Age-standardized YLDs (per 1,000) for leading causes, and contribution to
health gap between Australian Indigenous v non-Indigenous, Australia 2011 AIHW (2017), Table S8.4 Policy implications While populations are living longer lives, they are also
spending more years in a state of morbidity ➢rise in non-communicable diseases. Requires often expensive funding of treatment and
care.
Creates a significant burden on families to provide
unfunded care for their loved ones. There has been relatively low funding of non- communicable diseases, which are overwhelmingly the
main cause of mortality and morbidity. Policy implications Most United Nations Sustainable Development Goals
related to health focus on childhood and infectious diseases
diseases – But only a few for non-communicable diseases – Even less for morbidity. Policy interventions to reduce smoking in high-income countries (eg higher taxes, advertising
restrictions, anti-tobacco advertising, plain packaging)
are a successful example of how policy can reduce
non-communicable diseases (eg cardiovascular
diseases, respiratory diseases, lung cancer) Needs to be replicated in countries with high smoking
prevalence. Policy implications Leading risk factors of mortality and morbidity including
high body-mass index, high blood sugar, blood pressure
and cholesterol, and dietary risks. Unlike smoking, there are fewer prominent examples of
policies to improve diets and exercise. Many national and subnational governments have taxed
sugar content in drinks (eg Mexico) and banned trans
fats in restaurants (eg New York City). These risk factors pose a challenge to further increases
in life expectancy in all countries. Improving the evidence base for health
decision-making Need evidence to understand the levels, trends and
differentials in health status of populations, and for
policy makers to develop appropriate interventions to
reduce mortality and improve health. Many of the charts and tables I have shown you are
estimates based on statistical models, because
many countries have data of insufficient quality. Around one-third of all deaths in the world are not
registered, and even less have an accurate cause of
death. Death registration systems – current global
status Adair, T. et al., Bulletin of the World Health Organization, (2023) Death notification • There is commonly a lack of reliable means for the initial
notification of the occurrence of the community (ie non- facility) deaths. ➢ In contrast to deaths that occur in facilities, which generally have a
formal system to notify the death and medically certify its cause. • Initial notification of the occurrence of a death is a crucial first
step in the process towards eventual registration of the death ➢
and to provide stronger evidence on community deaths. Ward recorders are used in
Papua New Guinea to notify
deaths that occur in remote
villages Improving evidence about who dies of what Automated verbal autopsy Problem of obtaining cause of deaths that occur outside of facilities
– most death in low- and middle-income countries. Lack of doctors
in many areas. Verbal autopsy is an interview of the deceased’s family using a
standard, brief questionnaire to gather information on signs and
symptoms experienced before death Diagnosis of cause of death based on this information using an
algorithm. Provides evidence about what
people are dying from. Verbal autopsy results Leading causes of community deaths in Myanmar,
2018/19, derived using verbal autopsy. Bo KS, Firth SM, Phyo TPP, Mar NN, Zaw KK, Kapaw NH, et al. (2023) Estimating causes of community death of adults in Myanmar
from a nationwide population sample: Application of verbal autopsy. PLOS Glob Public Health 3(11): e0002426.
https://doi.org/10.1371/journal.pgph.0002426 • Global Burden of Disease data and visualisations: http://ghdx.healthdata.org/gbd-results-tool https://vizhub.healthdata.org/gbd-compare/