Inequalities in health perception

Background

Healthy life expectancy (reported in the Demographics Dashboard) is a summary measure of how long and well people live, which provides important context for population health strategies. It is calculated through a combination of mortality rates and the proportion of people self-reporting good health taken from the Annual Population Survey. Respondents to the survey are asked to rate their health from very good to very bad. The same question was included in the 2021 Census and therefore responses to this provide a once-in-a decade-opportunity to understand disparities in how people perceive their health between groups with different characteristics. Analysis of the data also provides insight into the reasons behind healthy life expectancy inequalities.

Age matters most

29,250 people from Central Bedfordshire completed the General Health question in the 2021 Census and there were around 44,560 people (15.1%) who characterised themselves as not being in good health. As expected, age was the most significant factor affecting people’s perception of health, with the proportion of people in Central Bedfordshire reporting good or very good health reducing from 97.7% for the youngest group (0 to 15 years) to 62.8% for the oldest (65 years and over).

It is forecast that the older age groups will grow fastest and therefore it can be expected that there will be an increasing number of people not in good or very good health over time.

Figure 1: Population pyramid for Central Bedfordshire 2021,2025 and 2035

LOOKING FORWARD
Based on the local population forecasts, there are expected to be 64,100 people aged 65 to 84 and 11,400 people aged 85 and over by 2033, rising to 74,700 and 14,900 by 2043. If the proportions not in good health in 2021, 28,100 in 2033 and 33,300 in 2043 could be expected not to be in good health.

However, it is also important to understand the significance of differences in health perception between groups of people with different characteristics, irrespective of age.

General health perception in Central Bedfordshire compared

Age standardisation is a statistical method to do identify what is important beyond age. Using direct age standardisation 84.3% of people reported good or very good health in Central Bedfordshire, 11.7% fair health and 4.0% bad or very bad health, compared to 81.7% good or very good, 13.0% fair and 5.3% bad or very bad for England overall. This means a significantly lower proportion of people in Central Bedfordshire reported bad or very bad health than in England. However, significantly more females reported not being in good health in Central Bedfordshire than males.

Table 1: Directly age standardised general health Central Bedfordshire

Table source: Analysis of Census 2021 data downloaded from: https://www.ons.gov.uk/datasets/create/filter-outputs/d3bb0376-bf4a-4ea2-80c8-01e07948b8cd#get-data

Beyond age and sex, significant inequalities were apparent based on deprivation, ethnicity, living situation, economic and employment status, and disability and caring responsibilities. To quantify these inequalities, indirect age standardisations were calculated. These provide ratios for each different characteristic compared to the population of Central Bedfordshire overall, which is scored at 100. A ratio higher than 100 indicates that the risk of having bad or very bad health in the population with a specific characteristic is higher than would be expected in the standard population of Central Bedfordshire and a ratio lower than 100 means that the risk is lower.

For example, males who lived in households with four dimensions of deprivation had a ratio of 455, meaning they were four and a half times more likely to have bad or very bad health than the overall score for males in households in Central Bedfordshire.

The outcomes of this data analysis are presented in the Dashboard below. This is followed by a breakdown of what it means for the people of Central Bedfordshire now and projections of what could happen in the future based on the changes in population predicted in the Local population forecasts.

General health perception Dashbaord

Deprivation

Nationally, deprivation, as measured by the Index of Multiple Deprivation (IMD), is strongly correlated with healthy life expectancy.^[1] This is reflected locally, where, over the three-year period from 2018 to 2020, there was an 8.4-year gap for males and an 8.9-year gap for females in healthy life expectancy between the most and least deprived wards in Central Bedfordshire.^[2] (See the Demographics dashboard for more).

In the 2021 Census, deprivation was not based on the IMD but was calculated instead from other measures within the Census. For example, employment deprivation was identified if any member of a household who was not a full-time student, was either unemployed, or economically inactive due to long-term sickness or disability. As the health dimension was calculated from the general health question as well as those who identified as disabled, that dimension is not included in this analysis.

In the 2021 Census data, the relationship between deprivation and perceived health remained strong. The more dimensions of deprivation that were apparent in a household, the greater the poor health ratio, with those deprived in all dimensions approximately 4.5 times more likely to have bad or very bad health, compared to the Central Bedfordshire baseline. Of the individual dimensions, employment deprivation was most associated with not being in good health, with a ratio of 247 for females and 236 for males.

Figure 2: Females in deprived households with significantly higher proportions of people reporting bad or very bad health

Figure 3: Males in deprived households with significantly higher proportions of people reporting bad or very bad health

Ethnicity, primary language and length of UK residence

Central Bedfordshire does not have the same ethnic diversity as its near neighbours Luton, Bedford Borough and Milton Keynes City, however, trend data for the last twenty years show the population has become more ethnically diverse over time. Since 2001, the percentage of residents from ethnic minority groups has increased, while the percentage of White British residents has decreased in line with similar trends across the East of England and nationally.

Figure 4: Population of Central Bedfordshire by ethnicity 2001-2021

The percentage of Central Bedfordshire residents who were White British reduced from 94% to 88.7% between 2001-2011, and then down to 83.5% in 2021 although the number of local residents that were White British increased by nearly 75,000 from 2001 to 2021. In addition, the ethnic diversity within the white group increased with the percentage of white people from places other than the UK increasing from 3.3% in 2001 to 4.2% in 2011, reaching 6.7% by 2021.

Across all ethnic minority groups other than White British (English, Northern Irish, Scottish, Welsh and British) 45% were born in the UK in 2021, compared to 42% in 2011.

The ethnic mix also changed between 2011 and 2021, with the Other White group, which includes people from Eastern Europe, increasing 120% and adding nearly 8,500 people, the largest net population change of any ethnic group apart from White British.

The next largest increase in numbers was among the Indian population (2,150, 85%) and the Pakistani and Bangladeshi were among the fastest growing groups. They grew 197% and 167% respectively, although this equated to around 1,200 residents combined.

Black, Black British, Black Welsh, Caribbean or African population increased overall, driven by growth in the Black African population (+110%) and those with Caribbean heritage (+75%).

The ‘Any other ethnic group’ population saw the fastest growth, trebling in size (+300%) and it also saw the sharpest rise across the East of England and nationally. This is not clearly defined but does include Afghans.

These trends are expected to continue over the next two decades based on national migration.

Figure 5: Forecast of ethnicity population change 2023 to 2043

Being from an ethnic group other than White British does not correlate with lower healthy life expectancy. However, there are some specific ethnic groups that report higher proportions of people not in good health than others.

In Central Bedfordshire, people from the Gypsy and Irish Traveller communities had the highest poor health ratio among ethnic groups and therefore the poorest perception of their health when compared to Central Bedfordshire overall. This is the case for both females and males with ratios approximately 3.5 times worse than the Central Bedfordshire baseline.

Figure 6: Ethnic minority groups with significantly higher proportions of people not in good health by sex

The number affected is small (under 150 were not in good health in 2021), however, it is very likely that the Census significantly underestimates the size of Gypsy and Traveller communities as many members do not complete the Census. It is also important to understand that these communities are distinct from the Roma community that was counted in the Census for the first time in 2021. By contrast, the proportions of both males and females from the Roma community reporting bad or very bad health were similar to the Central Bedfordshire average.

Males and females from the Bangladeshi community also reported significantly higher proportions of bad or very bad health compared to Central Bedfordshire overall. This was reinforced by analysis of main languages spoken. Bengali with Sylheti and Chatgaya are the major languages of Bangladesh and males who use these languages as their main language are twice as likely and females over two and a half times more likely to consider their health bad or very bad health compared to the overall population of Central Bedfordshire.

The Bangladeshi population grew 167% between the 2011 and 2021 Censuses and even assuming no additional increase in the proportion of Bangladeshi people, there will be a considerable increase in the number of people from Bangladeshi and therefore increases in the number of people in bad health could be expected.

LOOKING FORWARD
Based on the local forecasts of growth for ethnic groups in Central Bedfordshire, there are expected to be around 2,100 Bangladeshi people experiencing bad or very bad health in 2033, rising to approximately 2,400 in 2043.

Among females, there are also significantly higher proportions of people reporting bad or very bad health with mixed White and Black Caribbean ethnicities, and this group accounted for more females not in good health than the Gypsy and Traveller and Bangladeshi groups combined.

It is also noteworthy that only people born in the UK had a poor health ratio that was significantly higher than the average for Central Bedfordshire. This suggests that people coming to the UK are generally in good health.

Figure 7: Poor health ratio by length of residence in the UK and sex

Living situation

The way people live – the accommodation they live in, who they live with and their relationships – can all impact their health and wellbeing.

Housing characteristics

In Central Bedfordshire all dwelling types other than detached, including flats and terraced houses and and semi-detached houses or bungalows, people who rent, and those with limited or no central heating, all reported higher proportions of people with bad and very bad health. Males who live in commercial buildings were also more likely not to be in good health.

Around two thirds of Central Bedfordshire’s population (approaching 200,000 people) live in dwelling types that have significantly higher proportions of people who view themselves as not being in not good health.  Approximately three quarters of all people who identify as not being in good health live in these types of dwellings.

The proportion of people who reported not being in good health also varied based on the way they pay for their homes. People who rent are more likely to report poor health than those who own their own home and there were 77,300 people in Central Bedfordshire who lived in rented accommodation in 2021. Of these, 50% rented from the local authority or another form of social renter. In total 15,200 people living in rented accommodation were not in good health, which accounted for 36% of males and 33% of females not in good health.

20,000 people lived in buildings with heating systems associated with higher proportions of people not in good health. Of these, 3,900 reported not being in good health, which meant that 9.0% of males and 8.8% of females not in good health had heating systems that were associated with poorer health.

Projections aren’t possible as these numbers cannot be added together as many will live in accommodation with two or more of these features. For example, a person could live in a purpose-built block of flats, rented from the local authority which had electricity as its only form of heating. However, it is still useful to see the housing characteristics associated with higher proportions of people not in good health as the numbers are considerable.

Figure 8: Characteristics of housing for which significantly more females report not good health

Figure 9: Characteristics of housing for which significantly more males report not good health

Overcrowded homes

Overcrowding is associated with poor health and lower educational attainment^[1] and, in Central Bedfordshire, those who live in accommodation with more people than bedrooms reported significantly higher proportions of people not in good health, with ratios of 154 for females and 136 for males. Furthermore, males who live in accommodation that meets the formal definition for the correct number of bedrooms for the number of people^[2] also had a significantly higher proportion reporting not being in good health, with an even higher ratio of 144, while males who lived in properties with more bedrooms than necessary also had a slight but significantly higher proportion of people not in good health.

In 2021, there were 95,500 living in accommodation with occupancy ratings with higher proportions of people not in good health and 57% of these were in homes with a rating suggesting they had fewer bedrooms than necessary. Furthermore, only females in households with more bedrooms than required and both sexes in households with considerably more bedrooms than required had a lower proportion of people reporting they were not in good health than the Central Bedfordshire baseline. This suggests that the method of calculating the number of bedrooms required may not be adequate for health improvement purposes.

Social isolation

People’s close social relationships can impact on their physical and mental health. Loneliness is particularly associated with increasing mortality and a higher risk of some cardiovascular, metabolic and neurological disorders as well as poor mental health.^[3] A striking feature brought out by the analysis of the General Health question in the 2021 Census was the relationship between higher proportions of people reporting they were not in good health and those living in single adult households.

For both males and females, a significantly higher proportion of people not living in a couple (with the exception of males married or in a registered civil partnership but not with tier partner and widowed females) reported not being in good health. The highest ratio for females was for those who were separated (155) and for males was for those who had never been married or in a civil partnership (144). The lowest ratio among people living alone was for those widowed (110 for males). Female widows had a ratio that was similar to the Central Bedfordshire baseline (103).

A total of 82,600 adults (aged 16 years and over) were not living in a couple in Central Bedfordshire in 2021, of whom 61% were single and had never been Married or in a civil partnership and 18,350 reported not being in good health. Adults not living in a couple accounted for 56% of females and 32% males not in good health.

There is a similar pattern for single person and single parent households across both sexes, for those aged 65 and under. However, for older people, men living alone have a higher ratio, while it is women living in a couple in which one partner is under and the other older than 65 years, that have a higher ratio, which may suggest a carer relationship exists.

LOOKING FORWARD
If the 2021 proportions were applied to the forecast population, there would be 91,000 adults not living in a couple in 2033, rising to 103,000 in 2043. Of these 20,200 in 2033 and 23,000 in 2043 could be expected not to be in good health.

Figure 10: Characteristics of living conditions for which significantly more females report not being in good health

Figure 11: Characteristics of living conditions for which significantly more males report not being in good health

Disability and unpaid caring

The percentage of people providing unpaid care decreased in Central Bedfordshire between 2011 and 2021, however, this may reflect more an increase in the number of people who provide care but do not see themselves as carers than a true decline in the number of carers. Indeed, the number of people living longer with long term conditions is increasing and therefore so is the demand for care. Unpaid carers – family, friends and volunteers – make a considerable contribution to meeting this need. However, it can also impact on the health of the carers.

For both males and females who provided 20 hours or more of unpaid care, the poor health ratio was significantly above the baseline for Central Bedfordshire and broadly, the more hours of care provided, the higher the proportion of people not in good health. Females who provided 20 to 49 hours or care had a ratio of 151, which increased to 154 for those providing 50 hours or more of unpaid care, while males had the same ratio of 146 for both categories.

Such small differences despite more hours might reflect additional emotional, physical and financial hardships faced by people providing 20 to 49 hours care. Many also have other responsibilities, such as employment and childcare as well as providing significant levels of care. Males providing up to 19 hours of care also had a significantly ratio, although at 108 much lower than their counterpart providing more hours of care. However, the picture was more positive for females providing up to 19 hours of care, who had a similar ratio to the baseline (97).
In 2021, there were a total of 23,550 people who identified themselves as providing unpaid care. This compares to a little under 3,000^[1] who were receiving Carers Allowance in the quarter ending August 2021 (the period when the Census took place). Of the people identified as unpaid carers in the Census 5,726 reported they were not in good health, again suggesting a gap between need and the support being received.

LOOKING FORWARD
If the 2021 proportions were applied to the forecast population, there would be 28,600 people providing unpaid care in 2033, rising to 32,600 in 2043. Of these 6,950 in 2033 and 7,900 in 2043 could be expected not to be in good health.

Figure 12: Characteristics of disability and unpaid caring and poor health ratios for females

Figure 13: Characteristics of disability and unpaid caring and poor health ratios for males

Females who identified themselves as being disabled in the Census were approximately 2.7 times more likely to report not being in good health than the Central Bedfordshire baseline, while males were 2.9 times more likely. Approaching two thirds of all the people who reported that they were not in good health in 2021 were also disabled. In 2021, there were 28,400 people considered disabled under the Disability Act. This number has been rising and is expected to continue doing so.

LOOKING FORWARD
If the 2021 proportions were applied to the forecast population, there would be 34,150 disabled people in 2033,
rising to 38,600 in 2043. Of these 22,000 in 2033 and 25,000 in 2043 could be expected not to be in good health.

Where people with poorer perceived health live

Just as there are differences in how people with different characteristics perceive their health, so are there differences based on where they live.

Generally, areas with higher population density have higher proportions of people who perceive themselves not to be in good health, although the least dense areas are not necessarily those with the lowest proportions. Furthermore, wards with similar proportions of people not in good health accounted for approaching half of all males and a little over a third of all females not in good health.

Figure 14: Poor health ratio for females by ward

Figure 15: Poor health ratio for males by ward

Economic activity, employment and education

Across sexes, there is a noticeable relationship between general health and economic activity, and the type of work undertaken is particularly important. Males not in employment and who have never worked were 1.6 times more likely not to be in good health in 2021 than the Central Bedfordshire baseline. For females, the ratio was slightly less: they were 1.5 times more likely not to be in good health. In 2021, there were 17,000 people in Central Bedfordshire who had never worked of whom 5,000 reported they were not in good health. Even more concerning were 57,800 people who hadn’t worked for 12 months or more, of whom 21,800 considered themselves not in good health.

Significantly more people who were economically inactive reported bad or very bad health. For those who in work, there is a positive relationship between managerial, administrative and professional occupations and good health, while there is a negative relationship for others. Both males and females in elementary occupations, sales and customer service occupations, and caring, leisure and other service occupations are all more likely to not be in good health. Males who are process, plant or machine operatives, or in skilled trades, also have a higher poor health ratio compared to the Central Bedfordshire baseline. These jobs often require fewer and lower qualifications and the relationship between qualifications and not being in good health follows a similar pattern to occupation. The group with no qualifications has a significantly higher proportion of people not in good health compared with the Central Bedfordshire baseline, while qualifications of 2 or more A levels or equivalent and higher, saw significantly lower proportions of people not in good health.

Finally, not having a permanent place of work is also associated with a higher proportion of people not in good health. Approximately 23,000 working people had no fixed location for work in 2021, 81% of whom were male. The ratio was closer to the Central Bedfordshire baseline for men (109), compared to 123 for women, but both were significantly higher.

Figure 16: Characteristics of employment, economic activity and education for which significantly more females report not good health.

Figure 17: Characteristics of employment, economic activity and education for which significantly more males report not good health.

LOOKING FORWARD
If the 2021 proportions were applied to the forecast population, there would be 20,000 people who had never worked in 2033, rising to 22,700 in 2043. Of these 6,100 in 2033 and 6,900 in 2043 could be expected not to be in good health.

References

^[1] Male R² = 0.68; Female R² = 0.65. https://fingertips.phe.org.uk/indicator-list/view/612WCOn3ZU#page/10/gid/1938133452/pat/15/par/E92000001/ati/502/are/E09000002/iid/94240/age/1/sex/4/cat/-1/ctp/-1/yrr/1/iid2/90362/age2/1/sex2/2/cat2/-1/ctp2/-1/yrr2/3/cid/4/tbm/1

^[2] https://centralbedfordshire.jsna.uk/jsna/population-place/demographic-dashboard/

^[3] Goux DM, Maurin E. The effect of overcrowded housing on children’s performance at school. Journal of Public Economics 2005;89(5):797–819.

^[4] The number of bedrooms the household requires is calculated according to the Bedroom Standard, where the following should have their own bedroom: 1. adult couple 2. any remaining adult (aged 21 years or over) 3. two males (aged 10 to 20 years) 4. one male (aged 10 to 20 years) and one male (aged 9 years or under), if there are an odd number of males aged 10-20 5. one male aged 10-20 if there are no males aged 0-9 to pair with him. 6. repeat steps 3-5 for females 7. two children (aged 9 years or under) regardless of sex 8. any remaining child (aged 9 years or under) An occupancy rating of: * -1 or less implies that a household’s accommodation has fewer bedrooms than required (overcrowded) * +1 or more implies that a household’s accommodation has more bedrooms than required (under-occupied) * 0 suggests that a household’s accommodation has an ideal number of bedrooms.

^[5] Hawkley, L.C. Loneliness and health. Nat Rev Dis Primers 8, 22 (2022). https://doi.org/10.1038/s41572-022-00355-9

^[6] Local Authority StatXplore CA in Payment