Abstract
Objective To measure the effect of buprenorphine-naloxone as opioid substitution therapy on glycemic control in patients with type 2 diabetes mellitus and opioid use disorder.
Design Retrospective cohort study and secondary data analysis.
Setting Northwestern Ontario.
Participants Patients with diabetes receiving opioid substitution therapy, as well as patients with diabetes only, who live in 6 remote First Nations communities.
Main outcome measures Glycated hemoglobin A1c values during a 2-year time period in the 2 groups.
Results Over a 2-year period, there was an absolute decrease of 1.20% in mean glycated hemoglobin A1c values in patients with diabetes who also received opioid substitution therapy, compared with patients with diabetes who were not being treated for opioid dependence, whose values rose by 0.02%.
Conclusion Patients with diabetes who also suffer from opioid use disorder achieve significant (P = .011) improvement in glycemic control when treated with buprenorphine-naloxone substitution therapy compared with other patients with diabetes. Treating opioid use disorder with buprenorphine-naloxone substitution therapy has an unintended positive effect on diabetes management.
Opioid use disorders and type 2 diabetes mellitus (T2DM) might coexist in many patients. This overlap occurs in many First Nations communities in northern Ontario, where both diseases have high prevalence.1
Diabetes rates in Indigenous Canadian populations have ranged from 2.7% to 19%, with some estimates of age-standardized prevalence as high as 30%.1 First Nations persons living on-reserve are at the highest risk of diabetes, with a prevalence of 15.3% for people aged 18 years and older, compared with 6.0% for non-Aboriginal populations.
The 2008 to 2010 First Nations Regional Health Survey reported that 6.8% of Ontario on-reserve respondents used opioids without a prescription.2 Prescription opioid abuse prevalence has been estimated to be between 35% and 50% in several Nishnawbe Aski Nation communities.3 The number of Indigenous people seeking treatment for prescription opioid use disorders in Ontario tripled between 2009 and 2014.4 In response to this public health and social crisis, a number of communities in northwestern Ontario have initiated opioid use disorder treatment programs combining psychosocial interventions with buprenorphine-naloxone substitution.5 There is a paucity of data available on whether participating in the buprenorphine-naloxone programs has an effect on glycemic control in patients with diabetes.
Both opioid use disorders and T2DM are chronic medical illnesses in which noncompliance with treatment is a common problem influenced by psychosocial factors.6 Opioid exposure is consistently associated with poorer glycemic control, as indicated by significantly elevated glycated hemoglobin A1c (HbA1c) levels relative to control groups (P < .05).7 Duration of addiction is a mediating factor, with substantially higher HbA1c levels among patients who had been using opioids for 2 or 5 years versus those who had been addicted for 5 months or 1 year.7
Different treatment methods for opioid use disorders have varying effects on diabetes control. Methadone maintenance therapy (MMT) in patients without T2DM is associated with increased sugar intake, elevated body mass index, and changes in glucose metabolism akin to those observed in T2DM patients.8 Conversely, acute administration of buprenorphine to laboratory animals is associated with reduced sugar consumption; this effect is reduced with chronic buprenorphine administration, with test animals consuming less sugar but a normal overall amount of calories.8 This effect is also seen with opioid antagonists such as naltrexone.8 A retrospective observational study comparing MMT and buprenorphine maintenance therapy with regard to incidence of T2DM diagnosis found that patients receiving MMT were significantly more likely to be diagnosed with T2DM, even after controlling for confounding variables in regression analysis (P = .0458). However, among patients who were diagnosed with T2DM, HbA1c results were not significantly different between the MMT and buprenorphine maintenance therapy groups, nor were they considered to be elevated.9
METHODS
One of the authors (D.T.) noticed a trend of improved HbA1c values in patients with diabetes treated with buprenorphine-naloxone in her First Nations community practice. This prompted a literature search and subsequent retrospective study.
The literature search was carried out using MEDLINE and EMBASE from 1982 to 2015 for the following terms and combinations: buprenorphine and/or naloxone, naloxone, narcan, opioid antagonist, and naltrexone; and diabetes mellitus, type 2, diabetes control, t2dm control, diabetes mellitus control, hyperglycemia, non-insulin dependent diabetes, and glucose metabolism. Naloxone was marginally associated with an increased risk of hyperglycemia. No articles focused on the effect of buprenorphine on glycemic control or treatment of patients with addiction.
Written permission was obtained from 6 First Nations communities in northern Ontario to participate in the study. The 6 communities being studied had buprenorphine-naloxone substitution programs ranging in size from 33 to 160 patients. The total population of the 6 communities was 4388 and included 526 patients receiving buprenorphine-naloxone and 573 patients with diabetes. A total of 62 patients had both opioid substitution therapy and diabetes, and these patients were the study group. The remaining 511 patients with diabetes functioned as the control group.
Using anonymized data from electronic medical records, we examined the HbA1c levels in patients with diabetes along a 2-year continuum in patients participating in a buprenorphine-naloxone substitution program compared with those not participating in such a program. The initiation of the buprenorphine-naloxone programs roughly coincided with the adoption of electronic medical records in the communities. The 62 study patients who had diabetes and were participating in buprenorphine-naloxone substitution therapy were identified. A total of 511 control participants consisting of patients with diabetes who were not prescribed buprenorphine-naloxone were also identified. Hemoglobin A1c levels from the beginning of the study period (July to December 2013), which were the earliest data in the electronic medical record and roughly correlated with the start of most of the community programs, were compared with HbA1c levels from January to July 2015. Independent 2-sample t tests were performed for equal variances. The study was approved by the Sioux Lookout Meno Ya Win Health Centre Research Review and Ethics Committee.
RESULTS
The average change in HbA1c level of the study group was an absolute decrease of 1.20%. It varied by community from an increase of 0.34% to a decrease of 2.30%, with 5 of 6 communities showing an improvement in this measure of glycemic control (Table 1).
In comparison, the control group of patients with diabetes and without buprenorphine-naloxone treatment experienced an average absolute rise in their HbA1c results of 0.02%. This average control group change in HbA1c level varied by community from an increase of 0.52% to a decrease of 0.50%, with half of the communities having a small increase and the other half a small decrease.
The absolute difference between the HbA1c levels in the 2 groups was 1.22% (P = .011), which is both clinically and statistically significant.
DISCUSSION
In this study, participation in a buprenorphine-naloxone program was associated with a decrease in HbA1c level compared with not participating in such a program. The decrease of 1.20% is clinically significant and favourably compares to the decrease associated with oral diabetes medications such as α-glucosidase inhibitors (1%), biguanides such as metformin (1%), dipeptidyl peptidase 4 inhibitors (0.75%), sulfonylureas (1.25%), and thiazolidinediones (1.25%).10 It is interesting that the study group had a higher baseline HbA1c level than the control group did (9.76% vs 8.90%) despite having a younger average age. This appears to reflect the burden of opioid use disorder on diabetes control shown by previous research.7 By the end of the study this relationship was reversed and the study group’s HbA1c level was lower than that of the control group (8.57% vs 8.91%). Perhaps participation in buprenorphine programs will reduce disease burden and diabetes complications in the long term. This might have an even bigger effect considering the relatively young age of the patients with both comorbidities.
While it is possible that some of the effect might be medication-related, such a large effect is likely owing to improved self-care and improved adherence to treatment of all health issues, as previous research on the pharmacobiology of buprenorphine administration did not find such a large effect. It is conceivable that patients participating in these programs also have more contact with health care professionals in regard to substance use disorders and might therefore receive improved follow-up of other health issues such as diabetes.
Limitation
One limitation of this study was the difference in age between the study group and the control group. Those in the diabetes-only control group were older than those in the buprenorphine-naloxone–treated group by an average of 13 years. However, as all eligible study participants with diabetes were included, this age difference likely reflects the difference in prevalence of opioid use disorder among different age strata in the communities, with people aged 20 to 50 most predominantly affected. Our study also had a greater proportion of female participants than male, which reflects the different sex distribution of diabetes diagnoses in these communities.
Our control group was made up of patients with diabetes who were not prescribed buprenorphine-naloxone. This group would include participants without opioid use disorder but might also include participants with untreated or undiagnosed opioid use disorder. However, this inclusion does not change the positive effect that the treatment of opioid use disorder had on diabetes control, as all patients had the percentage of change in their HbA1c level compared against their own baseline HbA1c level. As this was a retrospective observational study, we did not control for the initiation of various diabetes medications. As the control group was substantially larger than the study group, we believe such an analysis would not change the results. Furthermore, we postulate that the change in glycemic control is related to positive lifestyle changes, including improved adherence to diabetes medications. Therefore, changes in diabetes medications prescribed would not obfuscate the results.
All of the patients in this study were First Nations Canadians living in remote communities, who are known to experience high rates of T2DM.1 It is unknown if the results would be similar for other Indigenous populations, other ethnic groups, or urban populations.
Conclusion
This study demonstrates that patients with diabetes and opioid use disorder achieve improved glycemic control when enrolled in a community-based opioid substitution program. Both diseases have multiple long-term sequelae. Early investment in such community-based opioid use treatment programs might have many subsequent health and cost benefits.
Notes
EDITOR’S KEY POINTS
The glycated hemoglobin A1c levels in patients with diabetes were examined along a 2-year continuum. Patients with diabetes participating in a buprenorphine-naloxone substitution program were compared with those not participating in such a program.
Participation in a buprenorphine-naloxone program was associated with a decrease in glycated hemoglobin A1c level compared with the control group. The mean absolute decrease of 1.20% is clinically significant and statistically different compared with the control group (P = .011).
While it is possible that some of the effect might be related to medications, such a large effect is likely owing to improved self-care and improved adherence to treatment of all health issues.
POINTS DE REPÈRE DU RÉDACTEUR
On a mesuré les niveaux d’hémoglobine glycosylée A1c chez des diabétiques durant une période ininterrompue de 2 ans. On a comparé les diabétiques qui suivaient un programme de remplacement des opiacés par la buprénorphine-naloxone avec ceux qui ne participaient pas à un programme de ce type.
Chez les participants au programme de buprénorphine-naloxone, on a observé des niveaux d’hémoglobine A1c inférieurs à ceux des patients du groupe témoin. La diminution moyenne de 1,20% est importante sur le plan clinique et elle est statistiquement significative par rapport au groupe témoin (P = .011).
Même s’il est possible qu’une partie de cet effet soit due à des médicaments, il est probable qu’un effet de cet ordre soit plutôt le résultat d’une amélioration de l’hygiène personnelle et d’une meilleure fidélité aux traitements de tous les problèmes de santé.
Footnotes
This article has been peer reviewed.
Cet article a fait l’objet d’une révision par des pairs.
Contributors
All authors contributed to the concept and design of the study; data gathering, analysis, and interpretation; and preparing the manuscript for submission.
Competing interests
None declared
- Copyright© the College of Family Physicians of Canada