Lifestyle Interventions and HbA1c in Prediabetic Patients
By Ellen Feldman, MD
Altru Health System, Grand Forks, ND
- The National Health Services Diabetes Prevention Programme, a community-based program in the United Kingdom, offers intensive lifestyle counselling through a minimum of 13 group sessions across nine months targeting patients with elevated HbA1c or fasting glucose levels who are not diagnosed with diabetes.
- The analysis examined specific methods to determine the causal relationships between program referral and improvement in key cardiovascular risk factors.
- Program referral led to a modest yet clinically significant average reduction of 0.85 mmol/mol in HbA1c, notable reductions in weight and body mass index, no substantial change in blood pressure or total cholesterol levels, and a significant increase in diagnosis of type 2 diabetes.
SYNOPSIS: A comprehensive analysis of medical records shows that referral of patients with prediabetic risk factors to an intensive lifestyle change program results in significant improvements in critical cardiovascular risk factors, including hemoglobin A1c levels and obesity; these changes are more pronounced in men than in women.
SOURCE: Lemp JM, Bommer C, Xie M, et al. Quasi-experimental evaluation of a nationwide diabetes prevention programme. Nature 2023;624:138-144.
Type 2 diabetes mellitus is a largely preventable, potentially devastating, and unfortunately increasingly common metabolic disease.1 Research consistently shows that controlled lifestyle modifications, such as improving diet quality and increasing physical activity, are effective in preventing this disease. However, the real-world effects of lifestyle interventions remain less explored.1,2
To address this gap, Lemp et al analyzed large swaths of data from electronic medical records culled from 20% of general practices across England covering from early 2017 to mid-2020. This period marked the launch of the National Health Service Diabetic Prevention Programme (NHS DPP), designed to prevent type 2 diabetes through a community-based approach.3 The program offers a minimum of 13 group counseling sessions over nine months, focusing on topics such as nutrition and physical activity. Along with other eligibility criteria, the program targets patients with glycated hemoglobin A1c (HbA1c) levels above 6%.
Given that there was no defined control group built into NHS DPP, the researchers adopted a quasi-experimental design. This method, common in health and behavioral sciences, makes use of existing conditions when creating a control group would be impractical or unethical.4
In this case, Lemp et al employed (among other statistical maneuvers) a regression discontinuation design. This approach leverages the “quasi-random” division for program eligibility based on HbA1c levels, comparing the health outcomes of participants near the threshold — patients with HbA1c values just above the threshold value (and, thus, who were eligible for the program) vs. those who had HbA1c values just below the threshold value.
This innovative approach attempted to reveal the program’s effectiveness in a naturalistic or “real-world” setting, offering insights into the potential of intensive lifestyle interventions in preventing type 2 diabetes among at-risk populations.
To confirm the legitimacy of this approach, extensive analysis was devoted to comparing characteristics of the two groups (above and below the threshold), comparing other cohorts, and determining that the active treatment group (above -threshold group) did not qualify for any other specialized programs or services, such as cancer screening or more frequent medical visits. More than 2 million records were analyzed during the study period.
Results
The effect of being eligible for intensive lifestyle counseling was measured at a 0.10 mmol/mol reduction in mean HbA1c after 12 months compared to those who were not eligible. Meanwhile, the effect of being referred for intensive lifestyle counseling was measured at a 0.85 mmol/mol reduction in mean HbA1c after 12 months compared to those who were not eligible.
Only 17% of eligible patients were referred to NHS DPP or similar intensive counseling programs. Of this group, only 28.1% reported attending at least one session (per electronic medical records).
Estimates by extrapolation revealed that, if all eligible patients completed a course of NHS DPP, mean Hb1Ac would have declined by 3.00 mmol/mol. Secondary outcomes included changes in weight, body mass index (BMI), blood pressure, cholesterol, and diagnosis of type 2 diabetes. The results are summarized in Table 1.
Table 1. Outcomes for Patients Eligible for and Referred to the NHS DPP Study |
||||
| Measurement | Eligible for NHS DPP (Baseline HbA1c Above Threshold) |
P Value | Referred to NHS DPP | P Value |
HbA1c (mmol/mol) (n = 298,822) |
-0.10 (95% CI, -0.16 to -0.03) |
0.006* |
-0.85 mmol/mol 95% CI, -1.46 to -0.24) |
0.006* |
Weight (kg) (n = 208,111) |
-0.33 (95% CI, -0.48 to -0.18) |
< 0.001* |
-2.99 (95% CI, -4.38 to 1.61) |
< 0.001* |
BMI (kg/m2) (n = 184,087) |
-0.15 (95% CI, -0.21 to 0.09) |
< 0.001* |
-1.55 (95% CI, -1.88 to 0.83) |
< 0.001* |
Systolic blood pressure (mmHg) (n = 243,292) |
-0.24 (95% CI, -0.6 to 0.11) |
0.175 |
-2.03 (95% CI, -4.96 to 0.91) |
0.176 |
Cholesterol to HDL ratio (n = 178,852) |
0.00 (95% CI, -0.02 to 0.02) |
0.741 |
-0.03 (95% CI, -0.12 to 0.14) |
0.741 |
Diagnosis of type 2 diabetes mellitus (risk difference) (n = 327,756) |
0.39 (95% CI, 0.21 to 0.57) |
< 0.001* |
3.66 (95% CI, 2.00 to 5.33) |
< 0.001* |
*Statistically significant result NHS DPP: National Health Service Diabetic Prevention Programme; HbA1c: hemoglobin A1c; CI: confidence interval; BMI: body mass index; HDL: high-density lipoprotein |
||||
Gender Comparison
When results were stratified by gender, there were significant improvements in men (but not in women) who were referred to NHS DPP in multiple areas, including reduction in HbA1c, weight, and BMI. While women still showed a reduction in HbA1c, the decrease was more substantial for men. Table 2 exhibits these results.
Table 2. NHS DPP Program Results Stratified by Gender |
||||
| Measurement | Men Eligible for NHS DPP | Men Referred to NHS DPP | Women Eligible for NHS DPP | Women Referred to NHS DPP |
Change in HbA1c |
-0.15 (95% CI, -0.24 to -0.05) P = 0.002* |
-1.34 (95% CI, -2.2 to -0.38) P = 0.002* |
-0.05 (95% CI, -0.13 to 0.04) P = 0.264 |
-0.42 (95% CI, -1.16 to 0.32) P = 0.265 |
Change in BMI |
-0.19 (95% CI, -0.15 to -.12) P < 0.001* |
-1.73 (95% CI, -2.36 to -1.09) P < 0.001* |
-0.11 (95% CI, -0.18 to -0.03) P = 0.006 |
-0.92 (95% CI, -1.57 to -0.26) P = 0.006 |
Change in weight |
-0.56 (95% CI, -0.78 to -0.35) P < 0.001* |
-5.28 (95% CI, -7.33 to -3.24) P < 0.001* |
-0.14 (95% CI, -0.32 to 0.05) P = 0.140 |
-1.21 (95% CI, -2.82 to 0.4) P = 0.140 |
*Statistically significant result NHS DPP: National Health Service Diabetic Prevention Programme; HbA1c: hemoglobin A1c; CI: confidence interval; BMI: body mass index |
||||
Commentary
Lemp et al tackled a complex issue often encountered in behavioral health research: How can we assess the effectiveness of a behavioral intervention in real-world settings, beyond the confines of controlled experiments? This question is particularly relevant in addressing prediabetes, where researchers and clinicians have questioned if cultural, financial, and other barriers would make participation in an intensive lifestyle change program impractical or unattractive to many individuals.1-3
Using a regression discontinuation design, the researchers cleverly leveraged the eligibility threshold to form a control group, offering causal insights rather than mere associations. Overall, the findings suggest that referral to a lifestyle intervention program is effective in lowering HbA1c and reducing weight — although the effect is more pronounced in men.
Notably, the study revealed a small yet statistically significant reduction in HbA1c levels among those merely eligible for the program, although this change lacked clinical importance. However, for the 17% of participants referred to NHS DPP, the lowering of 0.85 mmol/mol represents a move toward both statistical and clinical significance. Moreover, the estimate of a 3 mmol/mol decrease in HbA1c levels among the 28% of those who reported participation in the program underscores the potential benefits of the program. This suggests that the observed improvement in the HbA1c levels among the referred group largely was influenced by those who actively engaged with the program.
One of the primary challenges of this research involves the incomplete data regarding patient referrals to and engagement with the NHS DPP sessions. The researchers were restricted to information contained in the EMR, but patient attendance at NHS DPP was not necessarily meticulously documented in the medical record. Thus, Lemp et al inferred participation from any mention of attendance in the records. This method introduces uncertainty regarding tracking consistent participation and fails to shed light on the underlying causes of the low referral rate or identify traits of patients more likely to participate in the program. A deeper analysis of these and other related factors could substantially improve our grasp of diabetes prevention in general.
Additionally, the almost paradoxical finding that the diagnosis of diabetes type 2 increased in the group referred to NHS DPP must be considered. Lemp et al postulated that this reflected the difficulty of diagnosing metabolic disorders in participants who were not followed closely (non-participants). There may be other explanations, including that the referred group were more at risk for this disorder at baseline. Follow-up studies may want to clarify and investigate further.
In summary, this study illuminates the tangible benefits of lifestyle intervention programs for diabetes prevention in real-world settings. Highlighted are the effectiveness of such interventions in lowering HbA1c and weight, particularly among men.
The study’s limitations and unexpected findings open avenues for future research to explore participant characteristics, engagement, and the reasons behind low referral rates, potentially deepening our understanding of how to effectively implement these interventions. Healthcare providers can readily draw on the current results to enhance and refine clinical practice by recognizing the importance of behavioral interventions in managing and preventing diabetes. Although this study may present more questions than answers, it serves as a significant stride forward toward understanding and enhancing diabetes prevention strategies.
References
- Lin X, Xu Y, Pan X, et al. Global, regional, and national burden and trend of diabetes in 195 countries and territories: An analysis from 1990 to 2025. Sci Rep 2020;10:14790.
- Galaviz KI, Weber MB, Kara Suvada BS, et al. Interventions for reversing prediabetes: A systematic review and meta-analysis. Am J Prev Med 2022;62:614-625.
- National Health Service. NHS Diabetes Prevention Programme. https://www.england.nhs.uk/diabetes/diabetes-prevention/
- Harris AD, McGregor JC, Perencevich EN, et al. The use and interpretation of quasi-experimental studies in medical informatics. J Am Med Inform Assoc 2006;13:16-23.
A comprehensive analysis of medical records shows that referral of patients with prediabetic risk factors to an intensive lifestyle change program results in significant improvements in critical cardiovascular risk factors, including hemoglobin A1c levels and obesity; these changes are more pronounced in men than in women.
Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.