In a recent study published in the journal Nature, researchers examined the original framework of the National Health Service (NHS) Diabetes Prevention Programme (DPP) implemented in the United Kingdom (UK).
This program is considered the most extensive behavior change program for prediabetes globally.
They examined the program in order to determine whether it was effective in improving the health of people with prediabetes, especially their glycated hemoglobin (HbA1c), excess body weight, and serum lipid levels, as well as risk factors related to cardiovascular health.
The NHS DPP targeted people with non-diabetic hyperglycemia (or prediabetes) and delivered face-to-face quality-assured intensive lifestyle and behavior change support to prevent or delay the onset of type 2 diabetes (T2DM).
Specifically, it offered weight loss, diet, and physical activity goals to its target population over 13 group sessions implemented in nine months. It operated at a large scale, offering 100,000 referrals in 2021.
T2D and diabetes-related deaths continue to surge worldwide, necessitating the implementation of population-level measures to prevent or delay T2D onset, enhance its diagnosis, and address its cardiovascular risk factors.
Clinical trials, such as the United States (US) Diabetes Prevention study, provided proof-of-principle that changes in lifestyle and behavior were effective when delivered during one-to-one sessions with incentives. A recent meta-analysis also showed that lifestyle modifications could reverse prediabetes in adults.
However, it remains unclear whether behavior change programs work in real-world settings. The most common reason for this is skepticism among clinicians that lifestyle counseling does not work for most people representing the general population as they have low health literacy and no willingness to engage.
Accordingly, at present, a substantial proportion of adults in England who have prediabetes do not take part in intensive lifestyle counseling because of system-level (unavailability of NHS DPP) or physician- and patient-level reasons.
About the study
In the present study, researchers used de-identified electronic health records (EHRs) of over two million patients from across one-fifth of general practices in the UK.
These records were obtained from the Clinical Practice Research Datalink (CPRD) Aurum and the NHS England Hospital Episode Statistics Admitted Patient Care (HES APC) databases. This data broadly represented the national population concerning geographical coverage, socioeconomic deprivation, gender, and age.
These individuals aged 18 to 80 underwent HbA1c evaluation between 1 January 2017 and 31 December 2018, which indicated their HbA1c levels ranged between 42–47 mmol mol−1 in the past year.
The researchers applied the regression discontinuity approach, a credible quasi-experimental strategy, for causal effect estimates of referral to intensive lifestyle counseling on all study outcomes.
They used a fuzzy regression discontinuity (FRD) design that assigned treatment probabilistically to estimate the effect of the patient presenting just above the eligibility threshold, resulting in the complier average causal effect (CACE). CACE is analogous to the intention-to-treat (ITT) effect in a target trial.
The primary outcome was the change in HbA1c levels from baseline to the final follow-up. Then, there were secondary outcomes, such as changes in body weight, body mass index (BMI), systolic and diastolic blood pressure (SBP & DBP), serum cholesterol, and triglyceride levels.
The exploratory analyses investigated the effect of program referral on the probability of newly prescribed diabetes medications, blood pressure- or lipid-lowering medications, any T2D-related complication (e.g., ophthalmic, neurological, renal), mortality, and hospitalization for a major adverse cardiovascular event (MACE).
The study follow-up began six months after the baseline HbA1c evaluation and ended with an outcome or censoring, for instance, due to death.
Note that the NHS DPP began phased roll-out in 2016, with waves 1, 2, and 3 starting from 1 June 2016, 1 April 2017, and 1 April 2018. Since the timing of each NHS DPP wave varied, the researchers presented a difference-in-differences analysis to compare patients from GPs of waves 1 and 2 with patients from practices of wave 3 (control).
Finally, the researchers presented an analysis using the regional variation in NHS DPP coverage as an instrumental variable (IV) for the actual receipt of program referrals.
The researchers noted that 26,970 patients were referred to a behavior change program or intensive lifestyle counseling anytime during the 12 months after the baseline HbA1c evaluation, of which 77.7% received the NHS DPP referral.
In robustness analysis, where they restricted the referral window to three months after the baseline HbA1c evaluation and only considered the NHS DPP referrals, merely 620 patients received referrals. Referral effects were similar, albeit with a slightly larger reduction in HbA1c than in the primary analysis.
There is mixed evidence of improvements in glycaemic control among people with prediabetes from controlled trials; on the contrary, this study substantiates indications from prior correlational studies suggesting beneficial effects of NHS DPP participation on HbA1c levels and weight control.
They found that the beneficial effect of referral to intensive lifestyle counseling on HbA1c at follow-up was significant (−0.10 mmol mol−1 to −0.85 mmol mol−1). Although the clinical significance of a 0.85 mmol mol−1 reduction in HbA1c is difficult to quantify at an individual level, it is meaningful at the population level.
Among patients eligible for the NHS DPP in the primary cohort, 28.1% of patients started the intervention. Scaling up the effect of program referral in the selected bandwidth reduced patients’ HbA1c concentration by ~3 mmol mol−1, given they strictly adhered to referral interventions.
Having an HbA1c concentration above the eligibility threshold for the NHS DPP was associated with a small increase in the probability of being prescribed diabetes medication shortly after treatment assignment, which increased at follow-up. There was no discontinuity in newly prescribed lipid- and blood pressure-lowering medications.
In secondary analyses, referral to intensive lifestyle counseling markedly reduced BMI by −1.35 kg m−2 and body weight by −2.99 kg. However, referral to intensive lifestyle counseling did not markedly reduce diabetes complications, emergency hospitalization for MACEs, and mortality in exploratory analyses.
Moreover, while both men and women significantly improved their BMI, effect estimates suggested more improvements in men than women.
In the difference-in-differences analysis, the group–time average treatment effect estimates favored implementing the NHS DPP to improve glycemic control.
The analysis using the regional variation in NHS DPP coverage also showed a significant beneficial effect of program referral on follow-up HbA1c evaluation.
This study supports further investment in structured, population-level behavioral interventions and targeted prevention strategies for individuals at risk for T2D, especially those not in the purview of conventional care pathways.
Importantly, these programs might also extend benefits to other non-communicable diseases such as cancer or infectious diseases (e.g., influenza, coronavirus disease 2019 [COVID-19]), which can have more serious effects in those with diabetes.
Overall, the study highlights a promising route to improve population-wide health more broadly.