Advertisement

Regional Trends

, Volume 42, Issue 1, pp 60–79 | Cite as

Regional health inequalities in England

  • Amy Ellis
  • Robert Fry,
Article

Abstract

Abstract

This article aims to provide an up-to-date picture of regional health inequalities in England. Health inequalities can start early in life and persist not only into old age, but subsequent generations. To address these health inequalities, a better understanding is needed of how health compares across the country.

This article brings together a range of health indicators for each of the nine government office regions, making comparisons across regions and against England as a whole. Indicators include life expectancy, alcohol consumption, smoking, drug usage, child obesity and mortality rates by cause. Bringing these indicators together in this manner provides a fuller picture of health differences between regions, instead of looking at each indicator in isolation.

The article confirms other studies showing that the north-south divide between regions persists, as the overall picture of health is better in the south than it is in the north. However, the article also reveals exceptions where some health indicators do not fit in with this trend.

Introduction

Which English region has the highest level of alcohol consumption? In which region are cancer incidence rates high? How does life expectancy at birth vary between the regions?

Over the last 50 years, there have been impressive social, economic and health improvements in this country. People from every class and region are healthier and live longer than ever before. Unfortunately, not everyone shares the benefits of these improvements.

Health inequalities can start early in life and persist not only into old age but subsequent generations. To reduce these health inequalities, the health gap needs to be narrowed between disadvantaged groups or communities and the rest of the country. This is an aim of Public Service Agreement (PSA) 18, to promote better health and wellbeing for all.

This article looks at inequalities among 18 health indicators by region. The aim is to provide an up-to-date picture of regional health inequalities and determine how the indicators compare against England as a whole. Bringing these indicators together in this manner provides a fuller picture of health differences between regions, instead of looking at each indicator in isolation.

The indicators analysed in this article are:
  • male and female life expectancy at birth in 2006–08

  • age-standardised mortality rates in 2008 – all causes of death, cancer, diseases of the respiratory and circulatory systems

  • age-standardised cancer incidence rates in 2005–07 – breast cancer (females) and lung cancer (males and females)

  • infant mortality rates in 2007

  • drug use among 16 to 24-year-olds – British Crime Survey 2008/09

  • average weekly alcohol consumption (males and females) – General Lifestyle Survey (GLF) 2008

  • current smokers (males and females) – GLF 2008

  • self-reported limiting longstanding illness – GLF 2008

  • childhood obesity in 2008/09, (reception and Year 6) – National Child Measurement Programme (NCMP)

Further information on the sources and context for these indicators is available in the Annex: Data sources.

The values for each indicator have been standardised (Box 1) and compared against the national figure for England, creating the spine charts shown in this article. The patterns emerging from the spine charts are discussed in detail in the section Regional Comparisons.1
Table 1

An example – Prevalence of childhood obesity in reception years

The robustness of the data (the width of the confidence intervals – Box 2) can affect whether significant differences are identifiable (Box 3). Caution needs to be exercised when analysing regional health data, largely because a lot of the indicators are based on sample surveys with small regional samples.

Box 1: Spine charts range standardisation methodology

The method used to create the spine charts, using the regional values for each indicator, is explained here.

The indicators featured are all different, some being proportions, some rates and some years. To ensure each region’s set of health indicators are comparable on one chart, the data are standardised. The specific technique, known as range standardisation, compares each value to the minimum value for that indicator if a high value is ‘good’ (such as life expectancy) or the maximum value if a high value is ‘bad’ (such as alcohol consumption). This is then standardised in relation to the England value. These figures are then altered so that England falls on zero, and any values below zero are worse, and any values above are better. These standardised figures are then plotted to create the spine charts.

An example of this calculation is shown in Table 4.1 for childhood obesity in reception year in the North East.

The proportion of children in reception year who were obese was 10.2 per cent in the North East in 2008/09. As higher proportions of obesity are bad, this figure must be standardised using the maximum value of the full set of regional figures (11.2 per cent in London), using the following calculation:

The value -0.1845 has been plotted on the spine chart for the North East (Figure 4.5), revealing that it had higher reception obesity levels than England.

This calculation is repeated for each indicator in turn.

Box 2: Confidence intervals

The GLF and NCMP provide a selection of the indicators analysed in this article. These are sample surveys which interview a sample of the population of interest. The estimated rates or proportions resulting from these surveys will rarely be identical to the true population value, and an indication of the accuracy of these estimates is provided using confidence intervals.

In addition, rates or figures based on a small number of events (such as deaths) or within a small population can also be subject to variability. Examples in this article include the age-standardised mortality rates, infant mortality rates, and cancer incidence rates. The estimates of these can be subject to fluctuation due to chance alone. For this reason, confidence intervals are also calculated around these figures.

Confidence intervals give a range in which the true value for the population is likely to fall. Upper and lower 95 per cent confidence intervals mean there is a 95 per cent chance that the range contains the ‘true’ rate.

For further information about calculating confidence intervals, please see the General Lifestyle Survey (GLF) 2008 Appendix C: Sampling Errors.

Box 3: Statistically significant differences

This article focuses solely on regional differences which are statistically significant, as opposed to differences that could have occurred by chance. The differences are assumed to be significant if the confidence intervals of the two comparator figures do not overlap. If there is an overlap, the values could fall within the same range, and no significant difference can be inferred. For example, the 95 per cent confidence intervals for the prevalence of childhood obesity in reception class children for the North East and London are 9.8–10.6 and 11.0–11.5 respectively. Given that these intervals do not overlap, there is a statistical significant difference between the two regions.

In this article, the same method (overlapping confidence intervals) has been used to compare regional values with England to see if significant differences exist, even though it is acknowledged that a value for a region is not independent from the England value that includes data from the region.

Regional comparisons

Figure 4.2 provides a set of spine charts showing the chosen health indicators for each region, whereas Figure 4.3 compares the regions for each indicator.
Figure 4.2

Health indicator spine charts by region – standardised values, 2006 to 2008

Figure 4.3

Health indicator spine charts by indicator – standardised values, 2006 to 2008

The England values for each indicator are represented by the centre line at zero, the actual non-standardised values of each indicator are shown in Table 4.4. If the indicator's bar is to the right-hand side of this, it suggests that region or indicator performed ‘well’ in comparison to England, whereas if it is on the left-hand side, it appears the region or indicator performed ‘badly’. A regional example using London shows that the levels of childhood obesity are ‘worse’ than England, and the drug use prevalence is ‘better’. Where no bar is shown against a region, this represents a value equal to that of England.
Table 4

England indicators

These spine charts show the indicators’ standardised values for each region (Box 1) compared against England, but each of these comparisons may not actually be a statistically significant difference (Box 3). Please refer to the section for each region to see whether the differences are significant.

The following section compares how the selected 18 health indicators performed in each region, and compared with England as a whole, referring to statistically significant differences only (Box 3). Background information for each indicator can be found in the Data Annex.

In summary, this article shows that the northern regions generally do less well than the midlands and London, and the best performing regions are the East of England and southern regions. This regional pattern corroborates findings from previous health inequality publications such as the Association of Public Health Observatories (APHO) Health Profiles 2009.

North East

The spine chart for the North East (Figure 4.5) shows a picture of poor health when compared with England. The North East, along with the North West and Yorkshire and The Humber, all show a similar number of indicators that were worse than England. The life expectancy for both males and females was lower than the England average and significantly lower than all other regions except for the North West.
Figure 4.5

Spine chart for North East

Poor life expectancy is linked to the other indicators that were lower than the England figure. Age-standardised mortality rates from all cancer, respiratory, circulatory and ‘all causes’ were worse than England and the majority of regions.

Incidence rates of lung cancer for both men and women (82 and 53 newly diagnosed cases in 2005–7 per 100,000 population respectively – Table 4.6) were greater than the rates for England (59 and 36 cases respectively) and all other regions.
Table 6

North East indicators

However, some indicators such as incidence of breast cancer, childhood obesity in reception years, alcohol consumption and drug use were not significantly worse than the England average.

Surprisingly, the proportion of males smoking was much lower than the England average. Because of a large confidence interval around this figure, it was only significantly different from the region with the highest proportion of male smokers, the North West (Figure 4.7).
Figure 4.7

Spine chart for North West

North West

In common with the North East, the spine chart for the North West shows a picture of poor health when compared with England. Along with the North East and Yorkshire and the Humber, the region shows a similar number of indicators that are worse than those of England. The life expectancy for both males and females was lower than the England average and lower than all other regions except for the North East.

Poor life expectancy is linked to other indicators that were also different to England. The age-standardised mortality rates from all cancer, respiratory, circulatory and ‘all causes’ were significantly worse than England and the majority of regions.

The proportions of male smokers and drug use amongst 16 to 24-year-olds (25 and 27 per cent respectively) appear much higher than the England averages (21 and 23 per cent) and are the highest across all regions. However, these differences were not significantly different compared with England as the confidence intervals for these statistics are wide.

Lung cancer rates for both males and females (70 and 46 people per 100,000 respectively) were higher than the England averages and only lower than one region, the North East.

In contrast, the levels of some indicators were more moderate and close to the England average. For example, breast cancer, childhood obesity (reception), long-term limiting illness and infant mortality indicators were not significantly different compared with England.4
Table 8

North West indicators

Yorkshire and The Humber

The Yorkshire and The Humber spine chart also illustrates poor health generally, although more moderate for some indicators when compared with the North East and the North West. For example, life expectancy for males was higher than the North East and North West, although it was lower than England.
Figure 4.9

Spine chart for Yorkshire and The Humber

The region shows the highest level of alcohol consumption (29 and 21 per cent respectively), measured as the percentage of males/females drinking more than 8/6 units of alcohol at least once in the survey week. The level was higher than the England average and the majority of regions.5
Table 10

Yorkshire and The Humber indicators

East Midlands

The spine chart for the East Midlands shows that the health in this area was similar to the England average for many of the indicators. It clearly shows that the health in this area was better than that demonstrated in North East, North West and Yorkshire and The Humber regions.
Figure 4.11

Spine chart for East Midlands

Life expectancy in this area for both males and females was close to the England average. It is the only region where the life expectancy for males was not significantly different to the England average.

There are some indicators that appear to be better than the England average – such as alcohol consumption and smoking – but these were not significantly different to the England average. The level of child obesity was low in this region and this was better than the England average.

To add to the mixed picture of health in this region, there are some indicators that were worse than the England average – such as breast cancer, long-term limiting illness, drug use and infant mortality. However, only one indicator was significantly worse compared with England – deaths from all causes.

The infant mortality rate in 2007 appeared to be higher than the England average according to the spine chart, but was only significantly higher than the South East. Comparisons with England and the other regions were not significantly different.6
Table 12

East Midlands indicators

West Midlands

The West Midlands’ spine chart shows a mixed picture of health across the range of indicators. Life expectancy for males and females was lower than the East Midlands and the England average.
Figure 4.13

Spine chart for West Midlands

The low life expectancy relates to several indicators that performed worse than the England average. Childhood obesity (reception and Year 6), deaths by all causes, deaths by respiratory causes, and infant mortality were all higher than the figures for England.

Most of these indicators had moderate values compared with the northern regions but the rate of infant mortality was the highest across all regions and was significantly higher than four of the regions.

By contrast, some indicators were better than the England average. The level of alcohol consumption by females (measured as the percentage of females drinking more then 6 units of alcohol at least once in the survey week) was lower than the England average (11 per cent versus 15 per cent). The incidence of female lung cancer was also lower than the England average.7
Table 14

West Midlands indicators

East of England

The spine chart for the East of England shows one of the best depictions of health across all regions, with almost every indicator better than the England average.
Figure 4.15

Spine chart for East of England

Of the 18 indicators, 11 performed significantly better than the England average. This included child obesity (reception and Year 6), deaths from all causes and deaths from cancer, respiratory and circulatory conditions. It was the region with the lowest percentage of females who consumed more than 6 units of alcohol on one or more occasions in a survey week. This was lower than the England figure and the North West and Yorkshire and The Humber.

The indicators which were better than average related to the high life expectancy in this region. Only the South East and the South West had a significantly higher life expectancy.8
Table 16

East of England indicators

London

The London spine chart shows that health in this region was broadly better than, or closer to the England average.
Figure 4.17

Spine chart for London

Infant mortality, long term limiting illness, alcohol consumption, smoking percentages, respiratory deaths, circulatory deaths and incidence of lung cancer were all indicators that were not significantly different to the England average.

There are several indicators that stand out as performing much better than the England average. The incidence of breast cancer was lower than England and all the regions. The percentage of 16 to 24-year-olds using drugs was lower than the North East, North West, East Midlands, East of England and the South East.

In contrast, the level of child obesity (11 per cent for reception years and 21 per cent for Year 6) in London was significantly higher than the England average (10 and 18 per cent respectively) and every other region.

Life expectancy seems to be in line with the picture of good health in London. Only the South East, South West and the East of England had significantly higher life expectancy.9
Table 18

London indicators

South East

The South East spine chart depicts a very positive picture of health, with almost every indicator better than the England average. The region had similar rates and percentages to the East of England and South West regions. This is reflected in the high life expectancy, which was higher than the England average and all other regions.
Figure 4.19

Spine chart for South East

This high level of life expectancy is linked to the better than average values for the majority of the indicators. In particular, the mortality rates for cancer, respiratory, circulatory and all causes were lower than the England average and most of the regions.

Other indicators also performed strongly with 12 out of the 18 indicators significantly better than the England average. Levels of childhood obesity and infant mortality were both lower than the England average and many of the regions.

The only indicator that dramatically differs from the rest was the percentage of drug use amongst 16 to 24-year-olds (26 per cent). It was higher than the other southern regions; London (17 per cent), and the South West (19 per cent).10
Table 20

South East indicators

South West

The South West spine chart shows an indication of relatively good health, similar to that seen in the South East and the East of England. Life expectancy was very similar to the South East, East of England and London and was better than the England average.
Figure 4.21

Spine chart for South West

Like the South East, the high life expectancy appears to be linked to the low mortality rates for cancer, respiratory and circulatory causes, all of which were lower than the England average.

There were also low levels of lung cancer and child obesity, both lower than the England average and most of the northern and midland regions. The region also had one of the lowest levels of drug use amongst 16 to 24-year-olds, lower than the North West and the South East, but not compared with the other regions due to the relatively wide confidence intervals.

There were three indicators that performed worse than the England average, but only one of these indicators was significantly lower than the England average. The South West had the highest level of breast cancer across all regions (135 cases per 100,000 population) and was higher than the England average (123 per 100,000 population) and all other regions.11
Table 22

South West indicators

Conclusions

This article has shown that there are many regional variations, leading to inevitable health inequalities. Overall, there was a north–south divide in health experiences. The northern regions (North East, North West and Yorkshire and The Humber) were at one end of the scale where health experiences were generally poorer than average, but in the south (South East, South West, East of England and London) the health experiences were largely better than average. The East Midlands and the West Midlands appeared to be around the England average.

This regional trend is consistent with findings from previous health inequality publications such as the Association of Public Health Observatories (APHO) Health Profiles 2009. See the References section for more north–south divide publications.

This article has highlighted that some indicators do not follow this north–south divide for health experiences. Some indicators which oppose the general trend are:
  • High levels of childhood obesity in London – this contrasts the majority of indicators doing better than England in this region. This finding is backed up by the APHO Profiles

  • High proportion of drug use in the South East – goes against the trend of good health in this region

  • High levels of breast cancer incidence in the South West – opposing the picture of good health in this region

Notes

Supplementary material

References

  1. Regional Trends online data tables, March 2010, Office for National Statistics, available at: www.statistics.gov.uk/statbase/Product.asp?vlnk=14161
  2. Health inequalities, Department of Health, available at: www.dh.gov.uk/en/Publichealth/Healthinequalities/index.htm
  3. PSA Delivery Agreement 18: Promote better health and wellbeing for all, June 2009, HM Government, available at: www.hm-treasury.gov.uk/d/pbr_csr07_psa18.pdf
  4. Methods for Area classification for statistical wards, Neighbourhood Statistics, available at: www.statistics.gov.uk/about/methodology_by_theme/area_classification/wards/methodology.asp
  5. Association of Public Health Observatories (APHO) Health Profiles 2009, available at: www.apho.org.uk/default.aspx?QN=HP_REGIONS_2009
  6. People and places: A 2001 Census atlas of the UK, University of Sheffield, 2004, available at: www.sasi.group.shef.ac.uk/publications/peopleandplaces.htm and North–south split ’getting wider’, BBC News, 30 June 2004, available at: http://news.bbc.co.uk/1/hi/uk/3852227.stm
  7. Health Profile of England 2008, Department of Health, January 2009, available at: www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsStatistics/DH_093465 and
  8. North–south divide widens in health map of England, The Guardian, 23 October 2007, available at: www.guardian.co.uk/society/2007/oct/23/health.northsouthdivide
  9. Life expectancy at birth and at age 65 by local areas in the United Kingdom, 2006–08, Online data tables, Office for National Statistics, available at: www.statistics.gov.uk/StatBase/Product.asp?vlnk=8841
  10. Life expectancy at birth and at age 65 by local areas in the United Kingdom, 2006–08, Statistical Bulletin, October 2009, Office for National Statistics, available at: www.statistics.gov.uk/pdfdir/liex1009.pdf
  11. Social Trends 39, Office for National Statistics, available at: www.statistics.gov.uk/downloads/theme_social/Social_Trends39/Social_Trends_39.pdf
  12. Life expectancy at birth 2006–08, General Details, Neighbourhood Statistics, available at: http://neighbourhood.statistics.gov.uk/dissemination/MetadataDownloadPDF.do?downloadId=25602
  13. Regional Trends online data tables, March 2010, Office for National Statistics, available at: www.statistics.gov.uk/statbase/Product.asp?vlnk=14161
  14. Regional Trends online data tables, March 2010, Office for National Statistics, available at: www.statistics.gov.uk/statbase/Product.asp?vlnk=14161
  15. Health Statistics Quarterly 44, Winter 2009, Office for National Statistics, available at: www.statistics.gov.uk/downloads/theme_health/HSQ44.pdf
  16. Infant mortality rate 2003–05, General Details, Neighbourhood Statistics, available at: http://neighbourhood.statistics.gov.uk/dissemination/MetadataDownloadPDF.do?downloadId=20396
  17. Regional Trends online data tables, March 2010, Office for National Statistics, available at: www.statistics.gov.uk/statbase/Product.asp?vlnk=14161
  18. Mortality statistics: Deaths registered in England and Wales by area of usual residence, Online 2007 data tables, Office for National Statistics, available at: www.statistics.gov.uk/statbase/product.asp?vlnk=15229
  19. Social Trends 39, Office for National Statistics, available at: www.statistics.gov.uk/downloads/theme_social/Social_Trends39/Social_Trends_39.pdf
  20. Health Statistics – Technical Assistance, Tools of the Trade, Confidence Intervals for a Crude Rate, Pennsylvania Department of Health, available at: www.health.state.pa.us/hpa/stats/techassist/cicruderate.htm
  21. International Classification of Diseases (ICD), World Health Organisation, 2010, available at: www.who.int/classifications/icd/en/
  22. Regional Trends online data tables, March 2010, Office for National Statistics, available at: www.statistics.gov.uk/statbase/Product.asp?vlnk=14161
  23. British Crime Survey and other surveys, Home Office, available at: www.homeoffice.gov.uk/rds/bcs1.html
  24. Table 4.4 in the Drug Misuse Declared 2008/09 publication provides data on all drug use estimates, available at: www.homeoffice.gov.uk/rds/pdfs09/hosb1209.pdf
  25. Results from the General Lifestyle Survey (GLF) 2008, Office for National Statistics www.statistics.gov.uk/downloads/theme_compendia/GLF08/GeneraLifestyleSurvey2008.pdf
  26. Smoking and drinking among adults 2008, General Lifestyle Survey, Office for National Statistics, available at: www.statistics.gov.uk/downloads/theme_compendia/GLF08/GLFSmoking&DrinkingAmongAdults2008.pdf
  27. Appendix B: Sample Design and Response, General Lifestyle Survey, Office for National Statistics, available at: www.statistics.gov.uk/StatBase/Product.asp?vlnk=5756&Pos=2&ColRank=1&Rank=272
  28. Appendix C: Sampling error, General Lifestyle Survey, Office for National Statistics, available at: www.statistics.gov.uk/StatBase/Product.asp?vlnk=5756&Pos=2&ColRank=1&Rank=272
  29. Overview Report, General Lifestyle Survey 2007, Office for National Statistics, available at: www.statistics.gov.uk/downloads/theme_compendia/GLF08/GLFoverview2008.pdf
  30. Regional Trends online data tables, March 2010, Office for National Statistics, available at: www.statistics.gov.uk/statbase/Product.asp?vlnk=14161
  31. National Child Measurement Programme: England, 2008/09 school year, Tables, December 2009, NHS The Information Centre for Health and Social Care, available at: www.ic.nhs.uk/statistics-and-data-collections/health-and-lifestyles/obesity/national-child-measurement-programme-england-2008-09-school-year
  32. National Child Measurement Programme: England, 2008/09 school year, December 2009, NHS The Information Centre for Health and Social Care, available at: www.ic.nhs.uk/webfiles/publications/ncmp/ncmp0809/NCMP_England_2008_09_school_year_report_2.pdf
  33. Regional Trends online data tables, March 2010, Office for National Statistics, available at: www.statistics.gov.uk/statbase/Product.asp?vlnk=14161
  34. International Classification of Diseases (ICD), World Health Organisation, 2010, available at: www.who.int/classifications/icd/en/

Copyright information

© Crown copyright 2010

Authors and Affiliations

  • Amy Ellis
    • 1
  • Robert Fry,
    • 1
  1. 1.Office for National Statistics

Personalised recommendations