Survey response: Comparison of key estimates

Unless stated otherwise, the Health Survey for England Feasibility Study (HSE FS) content and wording of questions are based on the HSE 2019 and HSE 2019 face-to-face surveys. Any substantive differences in content or wording used in the HSE FS are described in the relevant section below. Where appropriate, key estimates from the HSE FS are compared with HSE 2019, the latest year for which published results are available.     

A comparison of key survey estimates between HSE FS and HSE 2019 is presented to aid understanding as to whether a change in data collection methods has impacted survey estimates, the survey overall and its individual elements. Comparisons should be treated with caution due to the differences in sample and survey design.  

This section includes, for adults aged 16 and over, a comparison of key estimates on self-assessed general health, longstanding illness, and acute sickness as well as doctor-diagnosed diabetes.

In the HSE face-to-face survey interviewers ask about any longstanding health conditions [21] with instructions to probe for more conditions, recording up to six.  

The proportions of adults reporting ‘very good’ or ‘good’ health were similar for the HSE FS and HSE 2019 (76% and 75%, respectively) (years: p=0.162). This pattern was similar across men and women (77% in the HSE FS and 75% in HSE 2019 for men; 75% and 74% for women respectively) (years and sex: p=0.528).

 A lower proportion of adults reported ‘bad’ or ‘very bad’ general health in the HSE FS compared with HSE 2019 (5% and 7%, respectively) (years: p<0.001). This pattern was evident for men and women: 4% and 7% among men and 6% and 8% among women (years and sex: p<0.001).     

The proportion of adults reporting a longstanding illness was lower in the HSE FS than in HSE 2019 (37% and 43% respectively) (years: p<0.001). Among men, the difference was larger: 33% in the HSE FS  and 40% in HSE 2019. This difference was less marked for women (41% and 45%, respectively) (years and sex: p<0.024).

The prevalence of having at least one longstanding illness was higher among women (41%) than men (33%) in the HSE FS (sex: p<0.001).

The proportion of adults reporting acute sickness [22] was similar in the two surveys (16% in both surveys) (years: p=0.709). In the HSE FS, 14% of men and 18% of women reported acute sickness compared with 13% of men and 19% of women in HSE 2019. (years and sex: p=0.435).

Further information is available from Table 14. 

In the HSE FS, the proportion of adults reporting doctor-diagnosed diabetes was 8% in men and 5% in women (sex: p=0.053). The prevalence increased with age, from 2% of adults aged 16 to 24 to 15% of adults aged 75 and over (age: p<0.001). There were no differences between men and women nor across age groups compared with HSE 2019 (years and sex: p=0.809) (years and age: p=0.815). 

Further information is available from Table 15. 

This section includes a comparison of key estimates on adults’ health- related behaviours: cigarette smoking, e-cigarette use, alcohol consumption, physical activity, and fruit and vegetable consumption.  

Participants were asked about: smoking cigarettes; places where people smoke; giving up smoking; and use of e-cigarettes; nicotine replacement products, and cigars and pipes.  

The HSE questions required minimal change for the HSE FS online questionnaire. The showcard used on the face-to-face survey to illustrate examples of vaping devices was included as an embedded picture in the online survey. The same content was used for the paper questionnaire, but the HSE 2018 paper self-completion was also used to aid the design, particularly with regards to the complex routing and participant instructions, which required clear sign posting throughout.

In the HSE FS, 10% of adults were current cigarette smokers, compared with 16% in HSE 2019 (years: p<0.001). Conversely, a higher proportion in the HSE FS than in HSE 2019 had never smoked cigarettes regularly (69% and 60% respectively). (years: p<0.001)

In the HSE FS, similar proportions of men and women were current cigarette smokers [23] (10% and 9% respectively). In HSE 2019, the proportions currently smoking were higher, (years and sex: p=0.203) 18% of men and 15% of women were current smokers. Similar proportions in the two surveys were former smokers, with the difference accounted for by people who had never smoked regularly [24] 67% of men and 71% of women in the HSE FS and 55% of men and 65% of women in HSE 2019.

Further information is available from Table 16.

A higher proportion of younger adults than older adults were current cigarette smokers: 13% of those aged 16 to 24 compared with 4% of those aged 75 and over (age: p<0.001). A similar pattern was evident in HSE 2019 (21% and 6%, respectively) (years and age: p=0.793).

Further information is available from Table 17.

Among all adults, there was no difference in the median number of cigarettes smoked (10 cigarettes smoked per current smoker) in the HSE FS and HSE 2019 (years: p=0.455).

In the HSE FS, 5% of all adults were identified as current e-cigarette users. This proportion was 6% in HSE 2019 (years: p=0.027). The proportions of adults who had tried using e-cigarettes but were not current users was 13% in the HSE FS and 15% in HSE 2019 (years: p<0.001).    

Further information is available from tables 18.

Participants were asked about the frequency of drinking alcohol in the last 12 months and the number of drinking days in the last week. Further questions were asked to provide the average weekly consumption [25] and the maximum amount drunk on a single day in the last week. [26]

It was not possible to use the HSE questions directly for the paper mode due to the complexity of the routing. Hence, the drinking section from the HSE 2018 young person’s self-completion booklet was used to inform the design of the paper questionnaire. Questions were designed in a grid format for the type and amount of alcohol drunk on the heaviest drinking day. The paper questionnaire also included the images of different wine glass sizes to aid participant recall; this show card was also incorporated into the online questionnaire. 

Alcohol consumption is reported in terms of units of alcohol; one unit of alcohol is 10ml by volume of pure alcohol. Figure 6 shows the conversion factors used. Those who drank bottled or canned beer, lager, stout or cider were asked in detail about what they drank (e.g. normal strength or strong beer), and this information was used to estimate the amount in pints.

A programming error within the web questionnaire was identified in the data cleaning stage whereby participants who reported drinking wine in glasses on any one day during the last 12 months were not asked about the size of the glass. This resulted in the loss of some data for 2,311 participants who completed the online questionnaire. As a result, the usual approach to deriving units of wine drunk for these participants was not possible as the glass size was unknown.

An alternative approach for calculating units of wine drunk was applied, a method used on HSE until 2007. Where glass size was unknown the derivation of units consumed was based on an average glass size (175ml) and a conversion factor of 2.0 units was applied (see Figure 6 for conversion factors). So, the assumption was made to count all glasses of wine as 2.0 units of alcohol. Using an average size glass would systematically give underestimates of consumption for some participants, and overestimates for others. This differs to the current approach used on HSE, and therefore precludes any comparisons to HSE 2019 estimates (and those from previous years) and should be reviewed with caution. Table 22 presents estimates for unadjusted HSE 2019 data compared with adjusted HSE FS estimates. A discussion comparing the adjusted and non-adjusted estimates in HSE 2007 can be found in the HSE 2007 report. [27] That comparison found little difference in adjusted and non-adjusted estimates.         

Figure 6: Conversion factors for estimating alcohol content of drinks  

The proportions of adults who had not drunk alcohol in the last year were similar in the HSE FS and HSE 2019 (17% and 20%). (years p=0.051) Similar patterns were evident for those who drank alcohol at least once a week (49% in the HSE FS and 48% in HSE 2019). (years p=0.517).    

In the HSE FS, 83% of adults reported drinking alcohol in the past year, with similar proportions reporting this in HSE 2019 (80%) (years: p=0.51). The proportion of adults drinking alcohol in the past year was highest among adults aged 55 to 64 and 65 to 74 (88% for both age groups).

The proportion of adults drinking alcohol in the past year increased with age, from 72% of those aged 16 to 24 to 88% of those aged 55 to 74 before decreasing for those aged 75 and over (75%) (age: p<0.001).

Similar associations with age were seen in HSE 2019: 74% of those aged 16 to 24 and 85% of those aged 55 to 74 drank alcohol in the past year (years and age: p=0.591).  

A higher proportion of men than women in both surveys drank alcohol at least once a week (55% and 43% respectively in the HSE FS (sex: p<0.001); and 55% and 41% respectively in HSE 2019) (years and sex: p=0.386).

Adults aged 55 to 74 were more likely to drink alcohol at least once a week (59% and 60%, respectively) than adults aged 16 to 24 (25%) (age: p<0.001). Again, these proportions were similar in HSE 2019: 58% for those aged 55 to 74 and 30% for those aged 16 to 24.    

Further information is available from Tables 19, 20.

There was little variation between the HSE FS and HSE 2019 in the proportions of adults who had drunk alcohol at levels putting them at increased risk (19% in both surveys) (years: p=0.762). This pattern was similar for men and women (23% in the HSE FS and 25% in HSE 2019 for men; and 14% in the HSE FS and 12% in HSE 2019 for women (years and sex: p=0.004).

The proportion of adults who had drunk at higher levels than this (defined as more than 35 units per week for women and 50 units per week for men) was 6% in the HSE FS and 4% in HSE 2019 (years: p=<0.000).

Further information is available from Table 22.

The proportion of men and women usually drinking over 14 units in a week varied across age groups, increasing with age up to the age of 45to 54 among women (20%) and for those aged 65 to 74 among men (30%) (sex and age: p<0.001).

Further information is available from Table 21.

The physical activity module uses the short-form international physical activity questionnaire (IPAQ), last included in the HSE in 2018. These questions ask about levels of moderate to vigorous physical activity (MVPA) in the last seven days.

Within this report adults reporting less than 30 minutes of moderate to physical activity (MVPA) per week are defined as inactive’ and those reporting 30 minutes or more of MVPA per week are defined as ‘active’.  

In the HSE FS 68% of adults (67% of men, 69% of women (sex: p=0.116) were defined as active compared with 73% in HSE 2018 (years: p<0.001).

Patterns in the proportion of men and women reported to be active differed slightly by age in the HSE FS (age: p<0.001). Among men, the proportion declined by age, from 73% of those aged 16 to 24 to 54% of those aged 75 and over.

Among women the highest proportion reporting to be active were those aged 35 to 54 (75%) with the lowest proportion in those aged 75 and over (46%). In HSE 2018, the highest proportion reporting to be active were those aged 25 to 44 (78% and 82% respectively), with the lowest proportion in those aged 75 years and over (41%).  (years and age: p=0.006).

In all age groups, except the oldest, the proportions of those classified as active were lower in the HSE FS compared with HSE 2018. The largest differences were among those aged 16 to 24 (68% and 77% respectively) and those aged 35 to 44 (71% and 80% respectively).

In the HSE face-to-face survey, fruit and vegetable consumption is assessed in terms of portions consumed per day (roughly 80g per portion). Questions ask about the type and quantity of fruit and vegetables consumed.

The questions and their format in the online questionnaire were similar to those used in the HSE 2018 face-to-face survey. [28] The HSE 2018 showcards were embedded within the HSE FS online questions to appear on the screen where relevant. In the face-to-face survey, interviewers code the size of the fruit mentioned by the participant using a comprehensive fresh fruit code frame (e.g. an apple would be coded as a medium fruit). It was not possible to include this level of detail in the HSE FS; instead, a few examples for each size of fruit were provided. 

The questions for the paper questionnaire were taken from the WHS 2015 paper questionnaire. Though the content was similar to HSE 2018, the

In the HSE face-to-face survey, fruit and vegetable consumption is assessed in terms of portions consumed per day (roughly 80g per portion). Questions ask about the type and quantity of fruit and vegetables consumed.

The questions and their format in the online questionnaire were similar to those used in the HSE 2018 face-to-face survey. [28]  The HSE 2018 show cards were embedded within the HSE FS online questions to appear on the screen where relevant. In the face-to-face survey, interviewers code the size of the fruit mentioned by the participant using a comprehensive fresh fruit code frame (e.g. an apple would be coded as a medium fruit). It was not possible to include this level of detail in the HSE FS; instead, a few examples for each size of fruit were provided. 

The questions for the paper questionnaire were taken from the WHS 2015 paper questionnaire. Though the content was similar to HSE 2018, the different formatting of the WHS questions meant less routing was needed. Participants were only asked to record how much of each fruit or vegetable they had consumed. A description of what was shown on the show cards was included within the questions in the HSE FS paper questionnaire, but a visual picture was not. For the fresh fruit questions, participants were asked to record overall how much of each size fruit they had consumed instead of recording up to 15 separate fruits and the quantity of these separate fruits they had consumed.

38% of all HSE FS adults reported consuming five or more portions of fruit and vegetables. In HSE 2018 this proportion was 28%, 10 percentage points less than in the HSE FS (years: p<0.001).

Similar proportions were found among men and women (37% and 38% respectively consuming five or more portions) (sex: p=0.901). Proportions were lower in HSE 2018 among both men and women compared with the HSE FS, with a slightly larger difference in percentage points for men (12 percentage points) than women (8 percentage points) consuming five or more portions of fruit and vegetables (years and sex: p=0.003).

Proportions of adults reporting consuming no fruit and vegetables were also lower in HSE 2018 compared with the HSE FS. In HSE 2018 the proportion of adults who reported consuming no fruit and vegetables was 8% compared with 10% in the HSE FS (years: p=0.002). The proportion of adults consuming at least one but less than five portions of fruit and vegetables was higher in HSE 2018 compared with the HSE FS (years p<0.001).

In the HSE FS, the mean number of portions of fruit and vegetables consumed was 4.6. Similar results were reported for men (4.6) and women (4.5) (sex: p=0.238). In HSE 2018 a lower mean number of portions of fruit and vegetable was reported for all adults (3.7) (years: p<0.001). The mean portion of fruit and vegetables was lower among men (3.5) and women (3.8) in HSE 2018 compared with the HSE FS (years and sex: p=0.005).