Insulin regimen simplification is achieved in this algorithm by changing to or adding glargine as basal insulin and adding noninsulin agents to replace meal-time insulins. Long-acting insulins include insulin detemir (Levemir; Novo Nordisk) and Neutral Protamine Hagedorn insulin. Meal-time insulins include insulin lispro (Humalog; Lilly), insulin aspart (NovoLog; Novo Nordisk), and insulin glulisine (Apidra; Sanofi-Aventis) All patients were 65 years or older with type 2 diabetes, took 2 or more insulin injections/d, and had at least 1 episode of hypoglycemia (glucose level <70 mg/dL) during a 5-day period of continuous glucose monitoring.
aManagement of hyperglycemia in type 2 diabetes, 2015: a patient-centered approach: update to a position statement of the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD) (http://www.ncbi.nlm.nih.gov/pubmed/25538310). Further simplification agents were chosen based on risk of hypoglycemia, cost, adverse-effect profile, effect on weight, and effectiveness, as recommended.
bIf fasting glucose levels are higher than the goal but prelunch or dinner glucose levels reach the goal, change the glargine dose to bedtime at the same dose.
Munshi MN, Slyne C, Segal AR, Saul N, Lyons C, Weinger K. Simplification of Insulin Regimen in Older Adults and Risk of Hypoglycemia. JAMA Intern Med. 2016;176(7):1023-1025. doi:10.1001/jamainternmed.2016.2288
Hypoglycemia is a serious adverse event, especially in older patients with diabetes, and is associated with poor outcomes.1 Intensive insulin regimens add a large burden of self-care to older patients and increase the risk of hypoglycemia.2,3 Treatment guidelines and expert consensus recommend deintensification of treatment regimens in older adults.4 However, the effect of insulin regimen simplification on the risk of hypoglycemia or uncontrolled hyperglycemia is not known. In addition, the lack of an algorithm to guide insulin regimen simplification may account for missed opportunities to deintensify insulin regimens in older patients with diabetes.5
In a single-arm, intervention study, we recruited 65 adults 65 years or older, with type 2 diabetes (positive stimulated serum C-peptide levels) on 2 or more insulin injections/d with 1 or more episodes of hypoglycemia (≤70 mg/dL) over a 5-day period of continuous glucose monitoring (CGM). Simplification was achieved by switching multiple-dose insulin regimens to once-a-day glargine with or without noninsulin agents over a 5-month period, followed by a 3-month no-contact (no interaction with study staff) period to assess sustainability of the simplified regimen. The algorithm used to simplify the regimen included adding and/or changing long-acting insulin to basal insulin glargine, decreasing and discontinuing meal-time insulin, and simultaneously adding noninsulin agents (Figure).
We simplified the insulin regimen in 65 patients (mean [SD] age, 76  years; diabetes duration, 23  years; insulin injections/d, 3.7 [1.3]; hemoglobin A1c [HbA1c] level, 7.7% [1.2%]). Overall, 15 patients (23%) were living alone, 16 patients (25%) had cognitive dysfunction, 26 patients (40%) had depression, and 16 patients (26%) had a history of falls within the past 6 months. Hypoglycemia duration decreased at 5 and 8-month (P < .001) without any change in HbA1c levels (Table) after simplification. There was an improvement in HbA1c levels in patients whose baseline HbA1c levels were between 8% and 9% (mean [SD] improvement, −0.52% [0.5%]; P < .001) or above 9% (mean [SD] improvement, −1.7% [2%]; P = .03). There was a small worsening in those with baseline HbA1c levels below 7% (mean [SD] worsening, 0.37% [0.7%]; P = .03), while no change was noted in those with HbA1c levels between 7% and 8% (mean [SD] change, −0.06% [0.6]; P = .80). Diabetes-related distress score, measured by Problem Areas In Diabetes (PAID),6 improved at 5 months and remained low at 8 months (P < .001). Interestingly, the duration of hypoglycemia was not different when subjects were stratified by HbA1c levels of less than 7.0%, 7.1% to 8.0%, 8.1% to 9.0%, and more than 9.0% at baseline (mean duration, 278 min/5-day CGM), 5-month (mean duration, 124 min/5-day CGM), or 8-month (mean duration, 88 min/5-day CGM).
Our results show that insulin regimen simplification can reduce the risk of hypoglycemia without compromising glycemic control. After simplification, 36 (60%) of the study patients only needed 1 additional agent with basal insulin, 23 (38%) needed 2 agents, and 1 (2%) needed 3 agents. In addition, the total dose of the insulin was significantly reduced after simplification. An important benefit of the simplified regimen was that many patients were able to adjust the dose of glargine when glucose levels changed with other health issues during the no-contact period.
Another important finding of the study is that the level of HbA1c did not correlate with duration of hypoglycemia before or after simplification of the regimen, suggesting that HbA1c level is a poor predictor of the risk of hypoglycemia. Our previous study has shown that the HbA1c levels may not correlate with estimated average glucose in older populations.7 The combination of these findings suggests that liberating the HbA1c goals in frail older patients is not adequate to protect against the risk of hypoglycemia. The type of glucose-lowering agents and the strategy for their use are important to lower the risk of hypoglycemia.
Our study shows that (1) simplification of insulin regimens in older adults can decrease hypoglycemia risk and disease-related distress without compromising glycemic control; and (2) HbA1c levels may not predict the risk of hypoglycemia in the older population and should not be used as the sole parameter for goal setting.
Corresponding Author: Medha N. Munshi, MD, 110, Francis St, LMOB 1B, Boston, MA 02215 (firstname.lastname@example.org).
Published Online: June 6, 2016. doi:10.1001/jamainternmed.2016.2288.
Author Contributions: Dr Munshi had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Munshi, Segal, Weinger.
Acquisition, analysis, or interpretation of data: Munshi, Slyne, Segal, Saul, Lyons, Weinger.
Drafting of the manuscript: Munshi, Slyne, Saul.
Critical revision of the manuscript for important intellectual content: Munshi, Slyne, Segal, Lyons, Weinger.
Statistical analysis: Munshi, Slyne.
Obtained funding: Munshi.
Administrative, technical, or material support: Munshi, Slyne, Segal, Saul, Lyons, Weinger.
Study supervision: Munshi, Weinger.
Conflict of Interest Disclosures: Dr Munshi received an investigator-initiated grant from Sanofi and works as a consultant for Sanofi and NovoNordisk. Ms Segal is an advisory board member for Lilly USA.
Funding/Support: This study was supported by an investigator-initiated grant from Sanofi (grant No. Lantu_L_05685).
Role of the Funder/Sponsor: Sanofi had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.