Health Survey for England, 2021 part 1

For a full description of the derivation of the prediction equations, see Health Survey for England predicting height, weight and body mass index from self-reported data.

Prediction equations were based on linear regression modelling with the measured values of height and weight as the outcome variable and the self-reported values of height and weight plus age (in single-year bands but trimmed at 90 years) as independent variables. Linear, quadratic and cubic terms were entered for self-reported height and weight, and linear and quadratic terms were entered for age.

Adjusted height for men: =   292.2272 + (-3.3763324 × height) + (0.0217649 × height²) + (-0.0000364 × height³) + (0.0788123 × age) + (-0.0012560 × age²)

Adjusted height for women:  = 160.3468 + (-1.7360242 × height) + (0.0159004 × height²) + (-0.0000321 × height³) + (0.1119681 × age) + (-0.0016046 × age²)

Adjusted weight for men: = 9.9126 + (0.6653765 × weight) + (0.0036857 × weight²) + (-0.0000125 × weight³) + (0.0319996 × age) + (-0.0001959 × age²)

Adjusted weight for women: = 1.2812 + (0.9487828 × weight) + (0.001376 × weight²) + (-0.00000755 × weight³) + (0.0083899 × age) + (-0.0000739 × age²)

Worked example

For women, the adjusted height is calculated from self-reported height in cm as follows:

adjusted height = 160.3468 + (-1.7360242 × height) + (0.0159004 × height²) + (-0.0000321 × height³) + (0.1119681 × age) + (-0.0016046 × age²)

Therefore, for a female aged 50 with a self-reported height of 163cm, the adjusted height is calculated as follows:

adjusted height = 160.3468 + (-1.7360242 × 163) + (0.0159004 × 163²) + (-0.0000321 × 163³) + (0.1119681 × 50) + (-0.0016046 × 50²) ≈ 162.4cm

For more information: Tables A1, A2.

This section of the report presents findings on the differences between BMI, overweight and obesity from the self-reported and adjusted values of height and weight.

Mean values of self-reported and adjusted height and weight (and BMI derived from these) were calculated. To compare self-reported and adjusted values, the differences (self-reported minus adjusted) were calculated between means (BMI) and between prevalence levels of overweight (including obesity) and obesity. Negative values for the difference indicate lower mean BMI for self-reported versus corrected data.

On average, mean BMI derived from self-reported height and weight was lower than mean BMI derived from adjusted height and weight (1.0kg/m² men, 1.1kg/m² women). The gap between self-reported and corrected mean BMI increased with age. Among those aged 16 to 24, mean BMI from self-reported height and weight was 0.7kg/m² and 0.8kg/m² lower than mean BMI from adjusted height and weight for men and women, respectively. The equivalent figures among those aged 75 and over were 1.7kg/m² and 1.8kg/m² respectively.

Using BMI from self-reported height and weight for classification, 59% of men and 49% of women were classified as either overweight or obese; the equivalent estimates based on BMI from adjusted height and weight were 69% and 59%, respectively. Thus, for both sexes, the prevalence of overweight including obesity was 10 percentage points lower for self-reported versus corrected data.

The gap between the estimates of the proportion of adults classified as either overweight or obese increased with age, from 5 percentage points lower for self-reported versus corrected data among those aged 16 to 24 to 17 percentage points lower in those aged 75 and over. This pattern was similar for both sexes.

19% of men and 21% of women were classified as obese based on BMI from self-reported height and weight; 25% of men and 26% of women were classified as obese based on BMI from adjusted height and weight. Thus, obesity was 7 and 6 percentage points lower for self-reported versus corrected data for men and women respectively.

The gap between the estimates of the proportion of adults classified as obese increased with age, from 3 percentage points lower among those aged 16 to 24 to 10 percentage points lower in those aged 75 and over. This pattern was similar for both sexes.

The magnitude of the differences between the self-reported and adjusted estimates of overweight and obesity in HSE 2021 were in line with the differences between the self-reported and interviewer-measured data presented in the methodology study based on HSE 2011-2016 data.

In the HSE 2011-2016 data, the proportion of adults classified as either overweight or obese was 8 and 9 percentage points lower for self-reported versus interviewer-measured data for men and women, respectively. The proportion of adults classified as obese was 7 (men) and 6 (women) percentage points lower for self-reported versus interviewer-measured data.

For more information: Table A4.