rss
Postgrad Med J 88:679-683 doi:10.1136/postgradmedj-2012-130877
  • Original article

Falling mortality rates in Type 2 diabetes mellitus in the Wirral Peninsula: a longitudinal and retrospective cohort population-based study

  1. Banjo Adekunle Afolabi3
  1. 1Department of Community and Child Health, University of Chester, Chester, Cheshire, UK
  2. 2Department of Medicine, Wirral University Teaching Hospital Foundation NHS Trust, Upton, Merseyside, UK
  3. 3Department of Statistics, Aintree University Hospitals NHS Foundation Trust, Liverpool, Liverpool, UK
  1. Correspondence to Dr Chukwuemeka Nwaneri, Department of Community and Child Health, University of Chester, Chester, CRV 253, Riverside Campus, Cheshire CH11SL, UK; c.nwaneri{at}chester.ac.uk
  1. Contributors HC, DBJ, and CN designed and planned the study. KC had done preliminary evaluations on the study. All authors developed the protocol. CN and BA did the statistical analyses and alongside other authors interpreted the data and prepared the manuscript. All authors contributed to the review and writing of the paper.

  • Received 21 February 2012
  • Accepted 23 July 2012
  • Published Online First 17 August 2012

Abstract

Objectives To determine the life expectancy and mortality rates in patients with type 2 diabetes mellitus when compared with the UK general population; to measure the years of life lost.

Design Longitudinal and retrospective cohort study.

Setting The Wirral Peninsula in the northwest of England.

Participants Total of 13 620 patients with type 2 diabetes mellitus on the Wirral Diabetes Register.

Main Outcome measure All-cause mortality, from 1 January 2000 to 31 December 2007.

Results Over the 8-year period of the study, there were a total of 16 692.5 person-years lived and 3888 deaths; 2041 (52.5%) males and 1847 (47.5%) females with corresponding mean ages at death of 75.6±10.3 years and 80.2±10.2 years, respectively. Although prevalence rates increased linearly (from 1.06% in 2000 to 4.39% in 2007) a decrease in mortality rates (from 117 to 46 per 1000 population) in both sexes was observed. This coincided with a progressive fall in cardiovascular risk factors in this population. A survival time curve of life lived until death showed that males had 8.0 years reduction in life span and females' life span was reduced by 9.6 years when compared with UK general population. In both sexes, life expectancy was reduced by between 2 and 11 years dependent on the age of diagnosis, with males showing a greater degree of reduction.

Conclusion Type 2 diabetes mellitus is associated with a significant reduction in life expectancy, more markedly in men, and in those diagnosed before age 70 years. However, annual mortality rates have fallen progressively in our population and may contribute to longer survival and life expectancy in future years.

Introduction

It is well established that Type 2 diabetes mellitus (T2DM) is associated with a reduced life span.1 ,2 This is particularly marked in men compared with women.3 The cause of death in the majority of patients with T2DM is cardiovascular disease (CVD)4–7 associated with an increased incidence of CVD attributable to diabetes mellitus.8 In 2007–2008, the diabetes Quality Outcome Framework (QOF) prevalence for Wirral and England was 4.3% for males and 3.9% for females, whereas in 2008–2009, the diabetes QOF prevalence was 4.7% and 4.3%, respectively.9 Most published recommendations from National Committees, including the National Institute of Health and Clinical Excellence (NICE), and the British Hypertension Society Guidelines, indicate an increasing recognition of the importance of cardiovascular risk factors in determining outcome in patients with T2DM. Previous large-scale studies, including the United Kingdom Prospective Diabetes study (UKPDS) have indicated that blood pressure and glycaemic control are independent risk factors in macro- and microvascular complications of patients with T2DM.10 Targets for glycaemia, following the findings of the UKPDS studies, suggest maintenance of HbA1c at less than 7%. Identified targets of total cholesterol (<4 mmol/l) and low density lipoprotein cholesterol (<2 mmol/l) are recommended.

A recently published study in The Netherlands (ZODIAC-10) showed a comparable median life expectancy for T2DM with the general population, but a history of raised cardiovascular risk factors and albuminuria reduced life span.11 Consistent with recommendations of national bodies, increasingly tight control of lipids and blood pressure is encouraged. Studies in Cardiff, UK, and in the USA, have demonstrated a reduced life expectancy in patients with T2DM.12 ,13 Previous studies published in the 1990s indicate that average loss of life for a person living with T2DM is 7.5 years.14 However, this is age related, and individuals who develop diabetes at younger ages are known to have even more years of loss of life.15–17

The objective of this study was to determine life expectancy and mortality rates in patients with T2DM in the Wirral Peninsula compared with the general population of Wirral and England. We aimed to measure years of life lost and the median survival times in these patients.

Methods

Study design

A longitudinal and retrospective cohort study was undertaken of the Wirral Diabetes Register of patients with T2DM between 1 January 2000 and 31 December 2007 inclusive.

Setting

The Wirral University Teaching Hospital is a Foundation Trust that provides services to a population of 308 500 (48% males and 52% females) in Wirral, and is located in the northwest of England, UK. It comprises a quarter of the total Merseyside population. It has wide demographic variations consisting of 98% white population compared with a national UK average of 91%, and has a relatively increasing ageing population, and a relatively low population of younger people compared with England overall.18 ,19 Box 1 describes the demographics of the Wirral population.

Box 1

Demographics of the Wirral population

  • 19% of Wirral's population is aged >65 years and 63.5% is aged between 15 and 64 years; compared with a distribution across England of 15.9% and 66.3%, respectively.18

  • It is estimated that by 2026, the population in Wirral aged >65 will increase by 35%, while the overall population of Wirral will increase by just 3%;

  • The Wirral region is the 60th most socially deprived area of the 326 English districts, and is in the bottom 20%;

  • Wirral has extreme polarity of income distribution highlighting wide socioeconomic differentiation. Life expectancy in Wirral's most deprived areas is 4.5 times less than the Wirral average, and the quality-adjusted life expectancy is 5.3 times lower.20

Participants/data sources

The diabetes register established in 1997, provides a wealth of extensive and high-quality data on all patients diagnosed with diabetes in this area providing a large historical cohort of people with diabetes. The register contains both demographic and clinical information on all patients with diabetes in a data system commonly used within the National Health Service (NHS). Data is stored for a period of 8 years following the death of a patient. A sample of data on the register is audited annually to ensure continuous quality control, and all laboratory data are downloaded directly from the central hospital laboratory computer. Mortality statistics for England (and UK) are based on the information retrieved when death is registered. The completeness of the death register is determined by direct review of General Practitioner's practice records by the Diabetes Register Manager and triangulated with the proportion of deaths recorded by the Office for National Statistics.

For the register, a diagnosis of T2DM is defined according to clinical criteria set by the WHO, and includes age of onset greater than 40 years at diagnosis and not requiring treatment with insulin.

Quantitative variables

We obtained death registration data from the Wirral University Teaching Hospital and the Office for National Statistics, United Kingdom. Death certificates showed date of death, cause of death and underlying cause of death. The cause of death was coded to the International Classification of Diseases, 10th revision (ICD-10 code E10-E14).

Statistical methods

Using the statistical package SPSS (V.18.0) we analysed the survival time function of number of years lived from diagnosis to the time of death, and age at death for men and women at respective years using life table methodology (table 1). The life table death rate is therefore the age-adjusted death rate resulting from the weighting of age-specific death rate by the life table stationary population. Life expectancy, defined as the arithmetic mean of the years lived by a group of individuals from their starting age until all have died, is usually derived from life tables.

Table 1

Median age of death and age at death, and duration of diabetes for men and women in years between 2000 and 2007 in Wirral

Future life expectancy at five different ages was chosen within the limits of the data source; at 40, 50, 60, 70 and 80 years of age and above. In calculating life expectancy, we used the Reed-Merrell methodology21; here, the endpoint is mortality from all causes and the number of ‘years alive’ at each individual's age, where age intervals ‘x’ equal a 1-year interval, ‘nx’ is the number of individuals with T2DM who survive at the beginning of each age interval. The mortality rates, ‘dx’ for different ages and sexes were calculated as a measure of the number of people with T2DM who died within each of the age intervals. The proportion of people with T2DM who survive to the beginning of ‘x’ age intervals, ‘lx’ equals proportional death risks, ‘qx’, the midpoint survivorship, ‘Lx’, the total years left to be lived by T2DM survivors to age, ‘Tx’, and the life expectancy, ‘ex’, were calculated using the Reed-Merrell approach: where ex=Tx/lx.

We constructed table 1 based on annual medians of age at death, age at diagnosis and duration of diabetes. Table 2 shows comparative life expectancies of individuals of the general population in England and the Wirral T2DM population of the same age as at the age at diagnosis. Data identified in table 2 was produced by the Office of National Statistics in the United Kingdom.22 ,23

Table 2

Comparative life expectancies between the UK general population (UK ONS, 2007) and the Wirral T2DM population between periods 2000 and 2007

The reduction in life span is calculated using the following formula: Constructed Life expectancy (national death rates) minus Constructed Life expectancy (death rates for people with T2DM).

We also computed the annual prevalence and mortality rates in both sexes (table 3).

Table 3

Annual prevalence, all-cause mortality and sex-specific mortality rates

Results

A total of 13 620 T2DM patients were recorded on the register within the period of study, with 3888 deaths; 2041 (52.5%) males and 1847 (47.5%) females. The baseline characteristics of the participants were differentiated according to gender between the periods 2000 and 2007, as shown in table 1.

Average duration of diabetes prior to death was not statistically different between the sexes, (Wilcoxon p>0.05), whereas, differences in ages at death, and ages at diagnoses in both males and females were statistically significant (Wilcoxon p<0.05). The median age of death for men and women in years between 2000 and 2007 (table 1) showed higher median age in females than in males by approximately 3–6 years. Overall, age at diagnosis, duration of diabetes and median age at death for males and females were 77 and 82 years, 68 and 74 years, and 6 and 6 years, respectively. At age 40 years, life expectancy is 30.8 years (life span of 70.8 years), 9 years less than the corresponding age in the general population of England; at age 70 years, the life expectancy is 12.6 years (life span of 82.6 years) in males. Males diagnosed at 50, 60, 70 and 80 years have the corresponding ages at death of 73.1, 76.9, 82.6 and 85.9 years, respectively; whereas in females, ages at death were 74.7, 77.7, 82.0 and 87.6 years, respectively. Consequently, the increasing age at death of older T2DM males and females is distributed according to age of diagnosis.

The survival time curve shows an improvement when compared with previous studies,24 although still below the English general population (see table 2). Figure 1 illustrates the survival function curves of cumulative survival against age at death, duration of diabetes and age at diagnosis. There is no statistically significant change in the duration of diabetes between males and females, although there were statistical significances between age at diagnosis and age at death during these periods. Furthermore, the result shows that regardless of increasing trends in prevalence rates over the 8-year period, there is progressive decrease in all-cause mortality and sex-specific mortality rates (table 3). During this period, there was a progressive decline in several cardiovascular risk factors in the total Wirral T2DM population. There was a progressive decline in mean total cholesterol from 5.1 mmol/l to 4.3 mmol/l; p value <0.001, and mean low density lipoprotein cholesterol from 3.6 mmol/l to 2.2 mmol/l; p value <0.001. In the year 2000, all-cause mortality rate was 177 per 1000 population, whereas in the year 2007, the mortality rate was 46 per 1000 population. Females show a more dramatic reduction in mortality rates than males.

Figure 1

Survival curves of males and females among T2DM in Wirral in relation to age at death, duration of diabetes and age at diagnosis, between periods 2000 and 2007.

Discussion

Evidence shows that in the Wirral general population, males and females have life expectancies at birth of 75.7 and 80.9 years, respectively; at 65 years of age, life expectancies for males and females in Wirral were 16.6 and 19.3 years, respectively.25 In line with previous studies, our analyses suggest that there is a reduced life expectancy of patients with T2DM in the Wirral Peninsula during the study period between January 2000 and December 2007. The survival curves indicate a mortality rating of approximately 230%, that is, 130% over baseline. Illustrating the extent and degree of reduction in life expectancies between the T2DM in Wirral and the general population of Wirral, data from the ONS shows that at birth in males there is approximately a 6-year reduction in life expectancy.

When compared with the general population of England, there is a further 2-year reduction, making a total of 8 years reduction in life expectancy. At age 65 years, the life expectancies for males (and females) in England, Wirral general population, and T2DM in Wirral were 17.5 years (and 20.2 years), 16.6 years (and 19.3 years), and 15.1 years (and 15.9 years), respectively. The mean life span for T2DM males and females differ from the English general population by 8 and 10 years, respectively, comparable with data released by Diabetes UK in 2010.26 When compared with the outcomes of other researchers in the UK, the mean age at death of 77.0 years (75.6 years for males and 80.2 years for females) appears better24 but below the 81.4 years (77.3 years for males and 81.5 years for females) of the UK Government Actuary's data for the general population. This reduction in life expectancy is more marked in males and in patients diagnosed before the age of 70 years.

Males diagnosed at the age of 50 years have an approximate 6–7 years reduction in life expectancy; those diagnosed at the age of 70 years have a 1-year reduction in life expectancy. For females there is approximately an 8-year reduction in life expectancy for those diagnosed at the age of 50 years, and a 4-year reduction in life expectancy if diagnosed at the age of 70 years. The difference in females is partly due to an increased life expectancy in females in the non-diabetic population. At all ages, females have a greater life expectancy following the diagnosis of T2DM than males of between 2 and 3 years. The relative slight upward trend in the median age at death for both males and females is statistically significant with evidence of a rise in life expectancy in T2DM patients, although not when compared with the general non-diabetic population.

In T2DM patients, the mortality rates in both males and females decreases at a time when a rise in prevalence rates is observed. It is likely that the increase in prevalence rate reflects greater detection of patients in primary care as a result of improved screening. It is therefore conceivable that earlier detection of T2DM may result in an overall fall in mortality rates due to increased numbers of short-duration patients within this cohort.

However, at the same time, we observed a progressive fall in cardiovascular risk factors with lowered mean blood pressure and lipid profiles but no change in the mean population HbA1c level within the total Wirral T2DM population.

In 2005, the Wirral model of diabetes care was reconfigured with greater emphasis on a protocol-driven service based in primary care. Improvement in mortality rates appears to have begun prior to this date. It should also be recalled that vascular disease, which is the commonest cause of death in these patients, is considered to be a long-term progressive condition which evolves over several years. It is therefore uncertain whether the fall in mortality rates reflects improving cardiovascular risk factors or is due to other factors, such as effectiveness of patient management or to a combination of these.

Strengths and limitations

The strength of our study lies in the utilisation of an unbiased selection of patients with T2DM using the quality-controlled Wirral Diabetes Register, including those treated by primary and secondary care, and those unable to visit clinics. Evidence exists to show that many patients develop diabetes several years before eventual diagnosis. However, this cannot be factored into these figures and is a limitation. However, given the large sample size, the longitudinal (8 years) period of the study, and the congruent data from the Office of National Statistics, we are confident that the findings are accurate.

Interpretation of the results

It is evident from the analyses that T2DM is associated with a significant reduction in life expectancy. This reduction is particularly marked in men and those diagnosed before the age of 70 years.

Death certification is accessible in electronic format making such large-scale cohort studies practicable and feasible, although it is known that death certification in general is notoriously unreliable,27 ,28 making the register an ideal tool for exploring life expectancy analysis.

The individuals represented by these statistics were treated over the 8 years in a manner that reflects national policies and recommendations. Within this population, it is known that there has been a statistically significant downward trend in lipid values and also blood pressure readings during this period, but there has been no parallel improvement in glycaemic control.29 A follow-on study will therefore explore survival in relation to the HbA1c levels and the relationship between social deprivation, mortality and life expectancy.

More recently, there has been a trend to lowering treatable targets for blood pressure and lipids in patients with T2DM, but whether this will be associated with a more significant increase in life expectancy remains to be determined.

As is evident from this study, there is an improved life expectancy and survival time among the Wirral T2DM patient population when compared with other research findings, but there is room for improvement making the need for further detailed analyses of the register data an important priority.

Generalisability and comparison with other studies

We believe that the results are in agreement with the observations of a reduction in life expectancy by other authors.12 ,13 ,30 ,31 However, our results contrast with the outcome of a recently published study on life expectancy in a large cohort of T2DM (ZODIAC-10) in The Netherlands between 2001 and 2007. This study showed a normal life expectancy among T2DM patients when compared with the general population.11 However, this was a selective study excluding certain groups of patients (eg, patients unable to visit their GPs, very ill, with terminal disease and cognitively impaired) and known predictors of cardiovascular mortality, and had a small sample size (n=973) with a median follow-up of 5.4 years.

We believe that the findings of our study indicate that T2DM remains a highly significant adverse influence on life expectancy in the UK, supporting the need for the continual delivery of quality-based interdisciplinary healthcare.

Main messages

  • In the Wirral region, there is a fall in mortality rates in T2DM population.

  • There is reduced but improving life expectancy between T2DM population and the Wirral general population.

  • Rising incidence of T2DM in Wirral; mirrors the national trend.

Current research questions

  • Factors which determine the fall in mortality parameters in Wirral.

  • The role of socioeconomic factors in the reduction of life expectancy in T2DM in Wirral.

Key references

▶ Scarborough P, Bhatnagar P, Wickramasinghe K, et al. Coronary heart disease statistics 2010 edition. London: British Heart Foundation. 2010.

▶ Office for National Statistics. Population Estimates for UK, England and Wales, Scotland and Northern Ireland, mid 2006. http://www.ons.gov.uk/ons/publications/re-reference-tables.html?edition=tcm%3A77–213674 (accessed 13 Aug 2012).

▶ JSNA Consultation Document. Wirral Population. Chapter 2. 2011:1–39. http://info.wirral.nhs.uk/ourjsna/wirral2009–10/wirralpopulation/

Acknowledgments

We are indebted to assistance from Peter Marsh and Gordon Langshaw of the Informatics department at the Arrowe Park Hospital, Wirral University Teaching Hospital NHS Foundation Trust, Upton, in data retrieval and collection.

Footnotes

  • Funding Dr Nwaneri is a Gladstone Doctoral Fellow at the University of Chester.

  • Competing interests None.

  • Ethics approval The study protocol was approved by the local ethics committee of the University of Chester and Integrated Research Application System (IRAS) exemption.

  • Provenance and peer review Not commissioned; externally peer reviewed.

References


Free sample
This recent issue is free to all users to allow everyone the opportunity to see the full scope and typical content of PMJ.
View free sample issue >>

Don't forget to sign up for content alerts so you keep up to date with all the articles as they are published.