Vitamin D prescribing in children in UK primary care practices: a population-based cohort study

Abstract

Objective: To examine temporal changes in the incidence and patterns of vitamin D supplementation prescribing by general practitioners (GPs) between 2008 and 2016.

Design: Population-based cohort study.

Setting: UK general practice health records from The Health Improvement Network.

Participants: Children aged 0 to 17 years who were registered with their general practices for at least 3 months.

Outcome measures: Annual incidence rates of vitamin D prescriptions were calculated, and rate ratios were estimated using multivariable Poisson regression to explore differences by sociodemographic factors. Data on the type of supplementation, dose, dosing schedule, linked 25-hydroxyvitamin D (25(OH)D) laboratory test results and clinical symptoms suggestive of vitamin D deficiency were analysed.

Results: Among 2 million children, the crude annual incidence of vitamin D prescribing increased by 26-fold between 2008 and 2016 rising from 10.8 (95% CI: 8.9 to 13.1) to 276.8 (95% CI: 264.3 to 289.9) per 100 000 person-years. Older children, non-white ethnicity and general practices in England (compared with Wales/Scotland/Northern Ireland) were independently associated with higher rates of prescribing. Analyses of incident prescriptions showed inconsistent supplementation regimens with an absence of pre-supplementation 25(OH)D concentrations in 28.7% to 56.4% of prescriptions annually. There was an increasing trend in prescribing at pharmacological doses irrespective of 25(OH)D concentrations, deviating in part from UK recommendations. Prescribing at pharmacological doses for children with deficient status increased from 3.8% to 79.4%, but the rise was also observed in children for whom guidelines recommended prevention doses (0% to 53%). Vitamin D supplementation at pharmacological doses was also prescribed in at least 40% of children with no pre-supplementation 25(OH)D concentrations annually.

Conclusions: There has been a marked and sustained increase in vitamin D supplementation prescribing in children in UK primary care. Our data suggests that national guidelines on vitamin D supplementation for children are not consistently followed by GPs.

Keywords: epidemiology; paediatric endocrinology; paediatrics.

Figure 1

(A)Time trends in vitamin D supplementation prescribing in children using all components of the case definition together and each source of case identification independently; (B) Time trends in vitamin D supplementation prescribing in children by different age groups. crude incidence rates with 95% confidence intervals represented by vertical bars. 2007: SACN: update on vitamin D 2008: NICE: improving the nutrition of pregnant and breastfeeding mothers and children in low-income households 2011: the National Diet and Nutrition Survey (NDNS) rolling programme: data from years 1 & 2 2012: Department of Health: Vitamin D advice on supplements for at risk groups 2012: British Paediatric and Adolescent Bone Group’s position statement on vitamin D deficiency 2012: RCPCH: vitamin D: position statement 2013: RCPCH: guide for vitamin D in childhood 2014: NICE: vitamin D: increasing supplement use in at-risk groups 2015: NOS: vitamin D and bone health 2016: SACN: vitamin D and health report 2016: NICE clinical knowledge summaries: vitamin D deficiency in children SACN, Scientific Advisory Committee on Nutrition; NICE,The National Institute for Health and Care Eexcellence; RCPCH, Royal College of Paediatrics and Child Health; NOS, National Osteoporosis Society

Figure 2

(A) 25-hydroxyvitamin D concentrations recorded in the 90 days prior to or on the index date (n=12 277); (B) 25-hydroxyvitamon D concentrations recorded between 91–180 days after index date (only children with known levels, n=1035). Records with ambiguous unit of measurements were categorised as 'undetermined'.

Figure 3

Dosages of incident vitamin D prescriptions presented in equivalent daily dose among children with: (A) no pre-supplementation 25(OH)D concentrations; (B) 25(OH)D concentrations less than 25 nmol/L; and (C) 25(OH)D concentrations between 25–50 nmol/L in the 90 days prior to their incident prescriptions.