DIABETICMedicine DOI: 10.1111/dme.12810

Research: Complications Association between hypoglycaemia and impaired hypoglycaemia awareness and mortality in people with Type 1 diabetes mellitus A.-S. Sejling1,2, B. Schouwenberg3, L. H. Færch1, B. Thorsteinsson1,4, B. E. de Galan3 and U. Pedersen-Bjergaard1,4 1 Department of Cardiology, Nephrology and Endocrinology, Nordsjællands University Hospital, Hillerød, 2Faculty of Health Sciences, University of Southern Denmark, Odense, Denmark, 3Department of Internal Medicine, Radboud University Medical Centre, Nijmegen, The Netherlands and 4Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark

Accepted 15 May 2015

Abstract Aims To examine whether severe hypoglycaemia and impaired hypoglycaemic awareness, a principal predictor of severe hypoglycaemia, are associated with all-cause mortality or cardiovascular mortality in Type 1 diabetes mellitus. Methods Mortality was recorded in two cohorts, one in Denmark (n = 269, follow-up 12 years) and one in the Netherlands (n = 482, follow-up 6.5 years). In both cohorts, awareness class was characterized and numbers of episodes of severe hypoglycaemia either during lifetime (Danish cohort) or during the preceding year (Dutch cohort) were recorded. In addition, episodes of severe hypoglycaemia were prospectively recorded every month for 1 year in the Danish cohort. Follow-up data regarding mortality were obtained through medical reports and registries (Danish cohort). Results All-cause mortality was 14% (n = 39) in the Danish and 4% (n = 20) in the Dutch cohort. In either cohort, neither presence of episodes with severe hypoglycaemia nor impaired hypoglycaemia awareness were associated with increased mortality in age-truncated Cox proportional hazard regression models. Variables associated with increased risk of all-cause mortality in both cohorts were evidence of macrovascular disease and reduced kidney function. Conclusions Severe hypoglycaemia and hypoglycaemia unawareness are not associated with increased risk of all-cause or cardiovascular mortality in people with Type 1 diabetes mellitus.

Diabet. Med. 33, 77–83 (2016)

Introduction Hypoglycaemia is the most frequent acute complication of insulin treatment. It has a negative impact on perceived quality of life and constitutes the principal limiting factor for achieving and maintaining optimum glycaemic control in people with Type 1 diabetes mellitus [1]. Patients and healthcare professionals are particularly fearful of severe hypoglycaemia, which requires help from another person to restore blood glucose levels, thus creating a risk of physical damage (e.g. in traffic) [2]. It is estimated that people with Type 1 diabetes experience, on average, at least one severe hypoglycaemic episode every year [3], but the distribution is highly skewed [4]. People with Type 1 diabetes who have impaired hypoglycaemic awareness or hypoglycaemia Correspondence to: Anne-Sophie Sejling. E-mail: [email protected]

ª 2015 The Authors. Diabetic Medicine ª 2015 Diabetes UK

unawareness, which affects up to one third, are at highest risk of such events [2,4–6]. The potential contribution of severe hypoglycaemia to mortality in people with diabetes mellitus has received much attention since excess mortality during intensive glycaemic treatment of Type 2 diabetes mellitus was reported in the ACCORD study [7], particularly in those who have a history of cardiovascular disease. In that study, as well as in the ADVANCE study, both examining the effect of strict glycaemic control in Type 2 diabetes mellitus, occurrence of severe hypoglycaemia was associated with excess mortality from both cardiovascular and non-cardiovascular causes [8,9]. It is currently under discussion whether severe hypoglycaemia is a marker for an increased risk of adverse clinical outcomes associated with comorbidities rather than a direct cause [9,10]. People with Type 1 diabetes have a much higher rate of severe hypoglycaemia, albeit less cardiovascular comorbidity

77

DIABETICMedicine

What’s new? • The study is the first to show that neither measures of exposure to severe hypoglycaemia nor impaired hypoglycaemia awareness are associated with an increased risk of mortality in two European cohorts followed for 6.5–12 years. • This is an important finding, given the general anxiety about severe hypoglycaemia in people with Type 1 diabetes mellitus and its role in limiting the achievement of optimum glucose control. compared with people with Type 2 diabetes mellitus [12,13]; thus, studying the potential association between severe hypoglycaemia and mortality is particularly relevant in the population with Type 1 diabetes as it might yield clearer results regarding a relationship between the two. No studies of an association between occurrence of severe hypoglycaemia or impaired hypoglycaemic awareness and mortality have been published for Type 1 diabetes. We therefore studied the relationship between severe hypoglycaemia and long-term all-cause mortality and cardiovascular mortality in two cohorts of people with Type 1 diabetes. As impaired hypoglycaemic awareness is the major known risk marker for chronic recurrent severe hypoglycaemia [4], an analysis of the association between hypoglycaemia awareness status and mortality is also included.

Hypoglycaemia and mortality in Type 1 diabetes  A.-S. Sejling et al.

following data were collected: number of episodes of severe hypoglycaemia during lifetime and number of episodes of severe hypoglycaemia during the last 2 years before entering the study. Severe hypoglycaemia was defined as an episode for which third party assistance was needed in order to restore blood glucose levels. Awareness of hypoglycaemia was assessed by questionnaire by using the Pedersen– Bjergaard method, which characterizes participants as either aware, impaired aware or unaware of hypoglycaemia [5,13]. The cohort was subsequently followed for 1 year. During this period any episode of severe hypoglycaemia was recorded and validated by a telephone interview within 24 h [14]. Furthermore, data on severe hypoglycaemia requiring medical assistance at a hospital during the entire follow-up period was retrieved from the National Patient Register. The 12-year follow-up data on mortality and morbidity were assessed through searches of participants’ medical records as well as Danish registries including the Central Office of Civil registration, the Danish Cause of Death Register and the National Patient Register. Four participants lost to follow-up were included in the survival analyses until their last contact but they were not included in the registration of the number of episodes with severe hypoglycaemia requiring hospitalization within the last 11 years of the study and therefore not included in this part of the analyses.

Dutch cohort

Participants and methods Data were derived from prospective observational studies in two independent outpatient cohorts from Denmark and the Netherlands. The studies were approved by local ethics committees in Denmark and the Netherlands. All participants signed a written informed consent form before entering the studies.

Danish cohort

The Danish cohort included 269 participants with Type 1 diabetes mellitus recruited from the diabetes outpatient clinic at Nordsjællands Hospital Hillerød in 1999–2001 (Table 1) and followed for an average of 12 years until 2012. Inclusion criteria were age ≥ 18 years and diabetes > 2 years. Candidates who were pregnant, were on haemodialysis, or had severe concomitant disease were excluded from participation. Type 1 diabetes was defined as need for insulin therapy from the time of diagnosis and unstimulated C-peptide concentrations < 300 pmol/l or stimulated (venous blood glucose concentration >12 mmol/l) C-peptide concentrations < 600 pmol/l. At baseline the participants filled in a questionnaire including questions about hypoglycaemic exposure. The

78

The Dutch cohort included 486 participants with Type 1 diabetes recruited from the outpatient diabetes clinic of the Radboud University Medical Centre in 2006–2008, excluding four participants without any contact after the baseline visit. All participants were classified as having Type 1 diabetes according to the need for immediate insulin therapy and related clinical information. At baseline, clinical data were collected from the participants’ medical records. Hypoglycaemia awareness status was classified according to a Dutch modification of the Clarke questionnaire, consisting of five questions, where participants with a score 130/80, % any macrovascular complication, % diabetic nephropathy, % microalbuminuria, % neuropathy, % Awareness status, % Normal hypoglycaemia awareness Impaired hypoglycaemia awareness

Patients deceased

60 45  13 20  12

57 42  11 19  10

74 62  11 29  15

70  13 8.6  1.2 4.9

70  13 8.6  1.2 4.4

73  12 8.8  1.1 7

18 8 10 17 34

13 4 7 14 26

51 26 29 32 80

44 44 12

44.5 44.5 11

41 41 18

46 46  14 26  13

47 45  13 25  12

45 59  16 39  18

63  13 (7.9  1.2) 93

63  13 (7.9  1.2) 93

65  15 (8.1  1.4) 80

62 14 10 11 37

61 12 16 11 36

80 65 29 20 78

68 32

69 31

55 45

Statistics

Survival analyses were performed based on age-truncated univariate and multivariate Cox proportional hazards regression models. The model takes into account age at entry and at follow-up and therefore adjusts for the variability in age in the cohort. This is preferable because of the strong association between age and mortality. With the use of an agetruncated model it is also possible to analyse the effect of diabetes duration after adjusting for age which would otherwise have been difficult because of the strong association between age and diabetes duration. The proportional hazard assumption was assessed by Schoenfeld residuals and goodness-of-fit testing. In the multivariate models six to eight events per degree of freedom were allowed. Analyses were carried out for each of the hypoglycaemic variables. Data were processed using the SPSS software package (Version 20.0, IBM Corporation, Armonk, NY, USA) and the SAS software package (version 9.3, SAS Institute Inc, Cary, NC, USA). The level of statistical significance was chosen as 5 episodes 31 0.83 (0.24–2.8) Severe hypoglycaemia (2 years before baseline) n = 268 0 episodes (ref) 55 1 1 episode 14 0.68 (0.15–3.0) 2–5 episodes 20 0.90 (0.32–2.5) >5 episodes 11 0.92 (0.37–2.3) Severe hypoglycaemia (1 year prospective registration) n = 227 0 episodes (ref) 64 1 1 episode 17 0.78 (0.32–1.9) >1 episodes 19 0.63 (0.23–1.7) Severe hypoglycaemia (requiring medical assistance during follow-up) n = 265 No (ref) 69 1 Yes 15 0.94 (0.48–1.9) Awareness status n = 267 Normal awareness (ref) 44 1 Impaired awareness 44 1.0 (0.49–2.09) Unaware 12 0.72 (0.29–1.81) Dutch cohort Severe hypoglycaemia (1 year before baseline): n = 482 No (ref) 79 1 Yes 21 1.7 (0.64–4.4) Awareness status: n = 482 Normal awareness (ref) 68 1 Impaired awareness 32 1.2 (0.48–2.9)

Cardiovascular mortality

P

0.8 0.5 0.8

1 0.35 (0.04–2.9) 0.84 (0.21–3.4) 0.46 (0.11–1.9)

0.3 0.8 0.3

0.9 0.9 0.7

1 0* 0.42 (0.05–3.4) 1.17 (0.30–4.5)

1 0.4 0.8

0.6 0.4

1 0.60 (0.12–3.0) 1.1 (0.27–4.3)

0.5 0.9

0.9

1 1.1 (0.35–3.8)

0.9

1 0.5

1 1.4 (0.43–4.6) 0.67 (0.13–3.5)

0.6 0.6

0.3

1 3.3 (0.88–12.2)

0.08

0.7

1 3.6 (0.90–14.5)

0.07

P

*No cardiovascular disease death in this group.

80

ª 2015 The Authors. Diabetic Medicine ª 2015 Diabetes UK

Research article

DIABETICMedicine

Table 4 Baseline variables associated with increased risk of mortality in the two cohorts Danish cohort Baseline variables associated with increased mortality Male gender Duration of diabetes (per year) Presence of any macrovascular complication Presence of diabetic nephropathy (Danish cohort) or MDRD*

Association between hypoglycaemia and impaired hypoglycaemia awareness and mortality in people with Type 1 diabetes mellitus.

To examine whether severe hypoglycaemia and impaired hypoglycaemic awareness, a principal predictor of severe hypoglycaemia, are associated with all-c...
75KB Sizes 0 Downloads 10 Views