Participants and data collection

The HSE is a continuous, annual, cross-sectional health examination survey of people living in private households in England. Details of the survey methods and protocol have been published elsewhere (Mindell et al. 2012). In brief, HSE used a multistage stratified probability sampling technique. At each selected household, HSE collected data in two home visits. At the first (interviewer) visit, each participant was interviewed and completed a structured questionnaire regarding personal demographic, social and health information. At the second (nurse) visit, measurements of blood pressure (BP), waist circumference (WC), and hip circumference, and non-fasting blood samples were obtained. Laboratory parameters for this study were glycated haemoglobin A1c, total cholesterol, and high-density lipoprotein (HDL)- cholesterol.

Inclusion and exclusion criteria

All individuals aged 18 years and above who had a nurse visit in the CVD survey years 2003, 2006, 2011, or 2017 were included. In 2003, 2006, 2011, and 2017, a total of 14,836, 14,142, 8610, and 7997 adults were interviewed (new sample was interviewed each year), respectively. Household response rates achieved were 73%, 68%, 59%, and 55%, respectively. Non pregnant women and participants with complete information of CVD risk factors were included, providing a final sample of 23,501 participants.

Groups of study participants

The analytical sample was categorized into three age groups as young adults (18-54 years), middle-aged adults (55-74 years), and older adults (≥ 75 years) (Pearson-Stuttard et al. 2021). To compare the CVD risk factors in the population by diabetes status, participants were also sub-categorized into those with versus those without diabetes. Participants having diabetes diagnosed by their physician and/or having a survey-measured HbA1c ≥6.5% or 48mmol/mol were classified as having diabetes. Those not reporting doctor-diagnosed diabetes and with HbA1c <6.5% or <48mmol/mol were classified as without diabetes (Health Survey for England. 2017).

Demographic and anthropometric measurements

Ethnicity was categorized as White, Black, Asians, and others; locality type into urban and rural; and qualification levels into no qualification, below degree qualification and degree/National Vocational Qualification 4 (NVQ4) and above. Height was measured by asking participants to stand in an erect position with the head in the Frankfort plane using a portable stadiometer with a measuring scale in centimeters (cm) then converted into meters (m). Weight was measured with a digital weighing machine in kilograms (kg) on a flat surface. Using non-stretchable tape, waist circumference (WC) was measured midway between the centre points of the lower margin of the ribs and the top of the pelvis. BP was measured by the nurse with the participant in a relaxed sitting position using an appropriately sized cuff on the right arm, following a standardized protocol using an Omron digital monitor (Omron HEM-907, Omron Healthcare Co Ltd, Kyoto, Japan). This procedure was repeated three times at one-minute intervals after a five-minute rest. The mean of the second and third values was used for systolic (SBP) and diastolic pressure (DBP).

Definitions of CVD risk factors

History of cardiovascular disease was defined as reporting a history of IHD, angina, heart attack, stroke, abnormal heart rhythm, heart murmur, or other heart condition. Obesity was classified as generalized obesity and central obesity. The body mass index (BMI, or Quetelet Index) was calculated as weight in kilograms divided by height in metres squared (kg/m2). Participants were classified into four mutually exclusive categories: underweight (<18.5kg/m2); normal weight (18.5-24.9kg/m2); overweight (25.0-29.9kg/m2); and obese (≥30.0kg/m2) (Scheelbeek et al. 2019). Generalized obesity was defined as BMI ≥30 kg/m2. Abdominal or central obesity was considered as WC >102 cm in men and >88 cm in women in accordance with the report of the National Cholesterol Education Program Adult Treatment Panel III guideline (Hirani et al. 2008). Combined obesity was defined as participants having generalized obesity or central obesity.

Hypertension was defined as SBP/DBP ≥140mmHg/≥90mmHg or reporting taking medication prescribed for high BP. Untreated hypertension refers to those cases who had survey-detected raised blood pressure but were not taking medication for hypertension (Health Survey for England 2003). Smoking status was classified into current smoker, ex-smoker (used to smoke cigarettes regularly or had smoked at least 100 cigarettes in the past but who had quit smoking at the time of interview), or never smoked (Tompkins et al. 2021). Drinking frequency was categorized into three categories: those consuming alcohol at least three days a week as frequent drinker; at least once a week or month as an occasional drinker; and those who were non-drinkers or had consumed alcohol only once or twice in the past 12 months as rare/non-drinker. Excessive drinkers were considered as consuming more than 14 units in a week (Homlmes et al. 2020).

HSE used the National Institute of Health and Clinical Excellence (NICE) guideline to define raised total cholesterol as ≥5 mmol/L; HDL-cholesterol was defined as low at a level of less than 1.0 mmol/L (Tompkins et al. 2021). Total cholesterol (TC) to HDL ratio was calculated by dividing TC by HDL-cholesterol. Dyslipidaemia was defined as having TC to HDL (TC/HDL) ratio ≥6 or using lipid lowering medicines. Untreated dyslipidaemia refers to those cases who had a history of abnormal lipid levels or had TC/HDL ≥6 and were not taking any lipid-lowering medications (Christianson et al. 2006).

Ethical approval and informed consent

Ethical approval was granted prior to data collection by a relevant NHS Research Ethics Committee at the time of the survey. Verbal consent prior to enrolment in the study and written consent prior to taking biological samples were obtained in each survey from each study participant. Further ethical approval was not needed for this secondary analysis.

Statistical analysis

The survey design variables were used to take account of the complex survey design of the Health Survey for England (HSE). Non-response weights for the interview, nurse visit, and blood sampling, as appropriate, were used to obtain better nationally-representative estimates (Health Survey for England 2003). Age standardization was carried out using the direct standardization method. Statistical Package for Social Sciences (SPSS) version 20 was used for analysis.

Independent sample t-tests were used for continuous variables while chi-square tests were used for categorical variables for comparing results between participants with and without diabetes. Univariate and multivariate logistic regression were used to assess the association of risk factors among diabetic and non-diabetic adults in England. Within each age group, we computed adjusted odds ratios (AORs) for each risk factor to compare those with diabetes and with those without diabetes. The confounding risk factors were adjusted for age, sex, locality, ethnicity, qualification, survey year, cardiovascular disease, raised BP, dyslipidaemia, combined obesity, current smoker, and excessive drinker. Interaction effects between each variable and survey year was carried out individually in separate adjusted models to examine whether the effect of a variable on the risk of diabetes was changing over time, only significant interaction effects were included in the final model. P-value <0.05 was considered statistically significant in all tests used in this study.