Background Prescribing is the most common clinical intervention in the NHS, with annual costs exceeding £9 billion. Understanding differences in patterns of prescribed medicine use over time informs clinical practice and epidemiological research. We compare trends in prescribed medicine use by body mass index (BMI) and age to better understand the factors associated with increased prescribing.

Methods Repeated cross-sectional analysis of nationally-representative Health Surveys for England 1994–2015 (n=42 216 participants aged 20+ years with measured BMI and medicine use data). Sex-specific logistic regression models with main effects and interactions between BMI (reference: normal-weight; obese: BMI ≥30 kg/m²), age, and survey year on taking any prescribed medicine in the last week (excluding smoking cessation products and contraception) were assessed adjusting for smoking and education. Analyses were repeated for polypharmacy (3+ medicines), and for cardiovascular and non-cardiovascular medicines. Results are presented as fully-adjusted Odds Ratios (OR) with 95% Confidence Intervals (95% CIs).

Results Overall, the age-standardised prevalence of prescribed medicine use between 1994 and 2015 increased from 37.8% (95% CI 36.7% to 39.0%) to 46.7% (45.2%–48.2%) in men and from 45.7% (44.6%–46.8%) to 53.2% (51.8%–54.6%) in women. By 2015, use of 3+ medicines had doubled to 24.6% (23.4%–25.8%) in men and to 27.2% (26.1%–28.3%) in women. Among those taking any medicine, polypharmacy rose by 1.7 times to 42.8% (39.9%–45.6%) and 45.1% (42.9%–47.2%) respectively.

Prescribed medicine use increased over time more sharply with age. However, after age-adjustment, the increase in prevalence over time was greatest in obese women (BMI-by-year interaction: p=0.003). The odds of obese women taking any prescribed medicine in the last week were 1.5 times higher than those for normal-weight women in 1994 (OR: 1.49; 95% CI 1.28 to 1.73), but had increased to 2.1 in 2015 (2.14; 1.82–2.53). Increased medicine use over time was greatest in obese men for cardiovascular medicines (BMI-by-year interaction: p=0.036). The odds of obese men aged 50–59 years taking any prescribed cardiovascular medicine in the last week were 2.1 times higher than those for normal-weight men of the same age in 1994 (2.08;1.46–2.95). The equivalent odds had increased to 3.0 in 2015 (2.98; 2.10–4.21).

Conclusion Higher BMI is associated with increased prescribing over the last 20 years regardless of age, reflecting secular rises in levels of awareness and of treatment of obesity and other co-morbid conditions, such as hypertension and diabetes, increased availability of effective secondary prevention medicines (e.g. statins, ACE inhibitors), lower thresholds for their use, and greater adherence to guidelines for their prescription.