Skip to main content
Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Prevalence and Incidence of Hypoglycaemia in 532,542 People with Type 2 Diabetes on Oral Therapies and Insulin: A Systematic Review and Meta-Analysis of Population Based Studies

  • Chloe L. Edridge ,

    Affiliation University of Leicester, Department of Health Sciences, Leicester Diabetes Centre, Leicester General Hospital, Leicester, United Kingdom

  • Alison J. Dunkley,

    Affiliation University of Leicester, Diabetes Research Centre, Leicester Diabetes Centre, Leicester General Hospital, Leicester, United Kingdom

  • Danielle H. Bodicoat,

    Affiliation University of Leicester, Diabetes Research Centre, Leicester Diabetes Centre, Leicester General Hospital, Leicester, United Kingdom

  • Tanith C. Rose,

    Affiliation University of Leicester, Diabetes Research Centre, Leicester Diabetes Centre, Leicester General Hospital, Leicester, United Kingdom

  • Laura J. Gray,

    Affiliation University of Leicester, Department of Health Sciences, University Road, Leicester, United Kingdom

  • Melanie J. Davies,

    Affiliation University of Leicester, Diabetes Research Centre, Leicester Diabetes Centre, Leicester General Hospital, Leicester, United Kingdom

  • Kamlesh Khunti

    Affiliation University of Leicester, Diabetes Research Centre, Leicester Diabetes Centre, Leicester General Hospital, Leicester, United Kingdom



To collate and evaluate the current literature reporting the prevalence and incidence of hypoglycaemia in population based studies of type 2 diabetes.

Research Design and Methods

Medline, Embase and Cochrane were searched up to February 2014 to identify population based studies reporting the proportion of people with type 2 diabetes experiencing hypoglycaemia or rate of events experienced. Two reviewers independently screened studies for eligibility and extracted data for included studies. Random effects meta-analyses were carried out to calculate the prevalence and incidence of hypoglycaemia.


46 studies (n = 532,542) met the inclusion criteria. Prevalence of hypoglycaemia was 45% (95%CI 0.34,0.57) for mild/moderate and 6% (95%CI, 0.05,0.07) for severe. Incidence of hypoglycaemic episodes per person-year for mild/moderate and for severe was 19 (95%CI 0.00, 51.08) and 0.80 (95%CI 0.00,2.15), respectively. Hypoglycaemia was prevalent amongst those on insulin; for mild/moderate episodes the prevalence was 50% and incidence 23 events per person-year, and for severe episodes the prevalence was 21% and incidence 1 event per person-year. For treatment regimes that included a sulphonylurea, mild/moderate prevalence was 30% and incidence 2 events per person-year, and severe prevalence was 5% and incidence 0.01 events per person-year. A similar prevalence of 5% was found for treatment regimes that did not include sulphonylureas.


Current evidence shows hypoglycaemia is considerably prevalent amongst people with type 2 diabetes, particularly for those on insulin, yet still fairly common for other treatment regimens. This highlights the subsequent need for educational interventions and individualisation of therapies to reduce the risk of hypoglycaemia.


Hypoglycaemia in type 2 diabetes is associated with a considerable cost and burden to the health service, with an estimated annual cost to the NHS of £39 million[1]. There can also be substantial consequences for the individual, with an increased risk of mortality and morbidity from severe episodes [24]. Hypoglycaemia significantly impacts on an individual’s quality of life, their employment, social interactions, and driving [57]. In addition to the direct effects of hypoglycaemia, there may be a substantial indirect impact on serious long-term health consequences from medication non-adherence and purposeful hyperglycaemia, due to fear and avoidance of hypoglycaemia [8].

A common cause of hypoglycaemia is iatrogenic [9]. In order to avoid long-term complications of type 2 diabetes, emphasis is placed on improving blood glucose control [911]. A recent meta-analysis revealed that intensive glycaemic control in people with type 2 diabetes can result in a 17% reduction in non-fatal myocardial infarction and a 15% reduction in coronary heart disease events [12]. To help achieve tight glycaemic control, people with type 2 diabetes are frequently placed on intensive treatment regimens, including earlier initiation of insulin. Intensive regimens and tighter glycaemic control have been shown to increase the risk of hypoglycaemia[1315]. However, the topic of glycaemic management and pharmacological treatments is becoming more complex. Newer therapies and more treatment combinations, are increasingly becoming available, with the aim of maximising glucose control without the increased risk of hypoglycaemia[9, 16].

In addition to treatment regimes, other currently identified potential risk factors for hypoglycaemia in type 2 diabetes include exercise [17], increased age [18], presence of co–morbidities [18], hypoglycaemia unawareness [18], dietary mistakes [19], excessive dieting [20] or weight loss, alcohol [21], number of years since diabetes diagnosis [22], and time since insulin initiated [23],

Hypoglycaemia prevalence in real world type 2 diabetes settings has been considered [3, 24], however, there has not been a systematic review and meta-analyses of the literature. Previously published systematic reviews that have considered hypoglycaemic episodes in type 2 diabetes have tended to focus on clinical trials of the safety and efficacy of a particular drug [15, 2530]. Clinical trials usually exclude participants at higher risk of hypoglycaemia, attract more motivated and selective participants, have a treat to target design and place participants on treatment regimens specifically for the study. Consequently, generalisability of findings to real world settings may be limited and hypoglycaemia prevalence and incidence in clinical trials may be lower than in clinical practice. Knowing the incidence of hypoglycaemia is important to provide insight into its impact both clinically and from a patient level. It enables the planning of resources, exploration of risk factors and design of interventions for prevention of hypoglycaemia. Additionally, the frequency and severity of hypoglycaemia is often used as a rationale for the use of newer treatments and as a clinical indicator for the choice of treatment patients are placed on.

To our knowledge, there have been no systematic reviews of the prevalence or incidence rates of hypoglycaemic events in type 2 diabetes in population based studies. This systematic review aimed to collate and evaluate the current literature reporting the prevalence (proportion of people) and the incidence (rate of episodes) of hypoglycaemia in a real world type 2 diabetes population.


Search Strategy and Study Selection

We searched electronic bibliographic databases Ovid Medline (including in-process and other non-indexed citations) and Embase from 1998 to February 2014 and Cochrane (issue 2, 2014), using a combination of keywords and MeSH terms with English and American spellings. The search terms used covered type 2 diabetes, hypoglycaemia prevalence and hypoglycaemia incidence. An example search strategy tailored for Ovid medline can be found in S1 Fig.

The primary aim of this systematic review was to explore the incidence and prevalence of hypoglycaemia in people with type 2 diabetes within a general population-based setting. We included observational studies where: 1) the study population were a defined general population sampled from either a defined geographical location, attendees at a primary, secondary or other healthcare centre, or people registered on a health service or health insurance database; 2) the study population (or sub-population) all had type 2 diabetes and were aged ≥ 18 years old; 3) they were published in English language; and 4) they were published as full papers. We additionally required studies to report the number of type 2 diabetes participants who had experienced ≥ 1 hypoglycaemic episode, the incidence of hypoglycaemic episodes experienced, or data to allow the calculation of one of these measures. We did not apply any restrictions relating to the classification, definition or measurement of how hypoglycaemia was utilised by studies. We excluded studies if: 1) they were pharmacological trials or the study methods involved any alteration to a participant’s treatment or care, either pharmacological or behavioural; 2) the majority of participants were pregnant, fasting, on a restrictive diet, or were selected on the basis of having a specific acute or chronic illness; 3) participants were selected on the basis of their hypoglycaemia history or recent initiation of treatment regime; 4) participants were sampled via an established survey/consumer panel or a consumer database; or 5) hypoglycaemia rates were solely reported over less than one week.

Following removal of duplicate publications, titles and abstracts were reviewed independently by two reviewers (CE, TR) in order to identify studies that met the inclusion criteria. Where it was unclear from the abstract whether the inclusion criteria were met, the full article was retrieved and reviewed. In instances where there were disagreements between reviewers, a third reviewer (AJD or KK) was consulted.

Data extraction

A data extraction form was developed and pilot tested, with adaptations made accordingly. Two reviewers (CE, TR) extracted independently from included studies.

Where data were available, we extracted the following for each study: 1) study details (including study design, year published and country); 2) population details (including sample source, mean age, ethnicity, mean HbA1c, sex,treatment regimens and cardiovascular desease or events); 3) methods used to measure hypoglycaemia (self-report questionnaires, prospective diaries, emergency department admission records, claims databases); 4) the time period hypoglycaemia was measured over; and 5) severity measured (definition given by the study). We extracted outcome data for the proportion of T2DM participants only who had experienced at least one hypoglycaemic episode (prevalence) and the incidence rate of episodes experienced (or data to calculate).

Quality assessment

We created a quality assessment tool using elements from the Effective Public Health Practice Project (EPHPP) Quality Assessment Tool for Quantitative Studies [31], the Cochrane Collaboration Risk of Bias Tool [32] and the Quality assessment tool for systematic reviews of Observational Studies (QATSO) [33], see S1 Table. Two reviewers (CE, TR) independently assessed all studies for methodological quality. Each criterion was given a score of ++, + or-, and an overall quality grade was assigned for each of the following; sample bias (sample source representative and described well, sampling method, eligibility criteria applied and described, sufficient sample response), data collection bias (data collection tool well described, measurement reliability) and confounding and explanatory factors considered.

Data synthesis

For analytic and descriptive purposes, we categorised and defined hypoglycaemia by severity of the episode measured within the study. Where possible, based on the description given in the paper, studies were categorised as either mild/moderate or severe. Mild/moderate hypoglycaemia was referred to by studies when no third party assistance was needed during a hypoglycaemic episode. Severe hypoglycaemia was used by studies when third party assistance was required. However, some studies did not specify a severity when gathering some or all of the hypoglycaemia data; these data were classified as “unspecified” hypoglycaemia. It was assumed that the “unspecified” data covered both severe and mild/moderate hypoglycaemia. Where a study reported data separately for more than one of the classifications (severe, mild/moderate and unclassified), we extracted and analysed the data separately. If a study reported zero hypoglycaemic events, we entered 0.001 into the meta-analysis to avoid the study being excluded [32]. Where two prevalence or incidence means were reported in a paper for different populations, the pooled mean was calculated [32]. Population treatment categories relating to hypoglycaemia prevalence and incidence, were based on the available data. Where possible, data were grouped into: insulin, with or without oral-glucose lowering therapies; sulphonylureas, non-insulin but with or without other oral-glucose lowering therapies; non-sulphonylureas, non-insulin and non sulphonylurea oral-glucose lowering therapies; or mixed oral-glucose lowering therapies, non-insulin but no further description given relating to type of oral-glucose lowering therapies. For studies where treatment categories were not mutually exclusive, if the whole population were pharmacologically controlled, data for insulin were subtracted from the overall data, where possible, and the remaining data were grouped in the mixed oral glucose lowering category.

For the primary outcome of interest (the proportion of people who had experienced hypoglycaemia), we conducted meta-analyses for mild/moderate episodes, severe episodes and unspecified episodes. We carried out sub-analyses within these categories by treatment regime of the population. We also carried out a sub-analysis for severe hypoglycaemia which specifically required emergency or medical assistance. We used meta-regression techniques to assess the following potential explanatory variables at the study-level: mean age, percentage of male participants, mean HbA1c level, and the time over which hypoglycaemia was measured.

We carried out further meta-analyses for the rate of hypoglycaemic episodes per person year, for the three categories of, mild/moderate, severe and unspecified. Again, analyses were stratified by therapy option.

When calculating confidence intervals for meta-analyses, negative values can be found. We decided to cap all confidence intervals at 0.00, due to it not being plausible to have a negative confidence interval for prevalence or incidence.

Heterogeneity was assessed using the I2 statistic. Due to high levels of heterogeneity, we used random-effects models throughout to calculate effect sizes. Publication bias was assessed with a funnel plot and the Egger test. This was carried out separately for studies reporting, mild/moderate, severe and unspecified events and separately for prevalence and incidence. Significance was set at P<0.05, all p-values are two-sided and 95% confidence intervals are quoted throughout. We performed all analyses in Stata version 13 (StataCorp, College Station, TX).


Identification of studies

Results relating to the identification process for eligible studies are summarised in Fig 1. Searches yielded 3348 citations, and 3063 unique titles and abstracts were screened for eligibility. Following full text retrieval of 285 potentially relevant papers, 239 were subsequently excluded, leaving 46 papers eligible for inclusion in the analyses.

Fig 1. Flowchart of selection of studies from search to final inclusion.

Summary of included studies

Descriptive characteristics including the definitions and descriptions of hypoglycaemia used bythe 46 studies included in the systematic review are summarised in Table 1. All studies were observational, with 27 being cross-sectional, 11 prospective, 6 retrospective and 2 used a mixed methods design of cross-sectional/prospective (n = 1) or cross-sectional/retrospective (n = 1). Papers included were published from 1998 to 2013 inclusive. The number of participants within each study ranged from 41 to 361,210. Studies were conducted in Europe (n = 25), North America (n = 14), Australia (n = 2), and Eastern/South East Asia (n = 5). Ethnicity was poorly reported; of the 9 (19.6%) studies that reported ethnicity, the proportion of non-white participants ranged from 17% to 100%. Population samples were obtained from either health clinic attendees (n = 21), diabetes registries/databases (n = 8), health insurance databases (n = 7), emergency department records (n = 4), community populations (n = 3) or pharmacy records (n = 2).

Table 1. Characteristics of studies included in the systematic review.

Over half of the studies reported only severe hypoglycaemia (n = 24), while 3 studies reported severe and mild/moderate. There were 6 studies which reported data on unspecified hypoglycaemia, 8 which reported data on unspecified hypoglycaemia along with severe, and 5 which reported data on unspecified, mild/moderate and severe. For studies reporting the prevalence of hypoglycaemia, the time period used for recall of previous hypoglycaemic episodes ranged from 1 month to 22 years. The majority of studies (n = 17) measured episodes over a period of 12 months. The measurement period for studies reporting the incidence of hypoglycaemia, ranged from 1 week to 22 years. The most frequently used was 12 months (9/21). A variety of methods/sources, either alone or in combination, were utilised to obtain relevant data. These included questionnaires (n = 25), emergency department records (n = 7), prospective diaries (n = 4), and claims databases (n = 3).

Study quality

A breakdown of study quality is presented in S1 Table. Most studies received a high quality grading for the consideration of data collection bias (42/46, 91.3%) and confounding and explanatory factors (43/46, 93.5%). However, under half of studies scored well for sample bias (15/46, 32.6%).

Prevalence of people who have experienced hypoglycaemia

Overall 46 studies involving 532,542 participants were included for the meta-analyses examining the prevalence (proportion of people who had experienced hypoglycaemia). Eight papers involving 4,083 participants, measuring over a period of 1 month to 10.5 months, reported mild/moderate hypoglycaemia. The pooled prevalence of people who had experienced hypoglycaemia was 0.45 (95% CI 0.34 to 0.57; Fig 2). In relation to those on insulin as a diabetes treatment regime, the prevalence was 0.52 (95% CI 0.37 to 0.67) compared with 0.33 (95% CI 0.24 to 0.42) for sulphonylureas. Data were not available to calculate any further treatment categories.

Fig 2. Forest plot showing the proportion of people experiencing mild/moderate hypoglycaemia in each study and the overall pooled estimate.

Boxes and horizontal lines represent proportion of people experiencing unspecified hypoglycaemia and the corresponding 95% CI, respectively, for each study. Size of box is proportional to weight of that study result. Diamonds represent the 95% CI for pooled estimates of effect and are centred on pooled hypoglycaemia incidence.

The meta-analysis for the prevalence of severe hypoglycaemia included 40 papers involving 528,310 participants, measuring over a period of 1 month to 22 years, with a pooled prevalence of 0.06 (95% CI 0.05 to 0.07; Fig 3). For people on insulin as a treatment regime, the prevalence was relatively higher at 0.21 (95% CI 0.16 to 0.25) when compared with regimens involving sulphonylurea (0.05 [95% CI 0.02 to 0.07]), non sulphonylurea therapies (0.05 [95% CI 0.03 to 0.07] and mixed oral glucose lowering therapies (0.05 [95% CI 0.02 to 0.07]); Table 2). In 14 studies involving 473,481 participants, severe hypoglycaemic episodes were restricted to only those requiring medical assistance (not any third party assistance), the overall pooled prevalence was 0.05 (95% CI 0.03 to 0.06). When these studies were excluded the overall pooled prevalence for severe hypoglycaemia was 0.08 (95% CI 0.06 to 0.10), see Table 2.

Fig 3. Forest plot showing the proportion of people experiencing severe hypoglycaemia in each study and the overall pooled estimate.

Boxes and horizontal lines represent proportion if people experiencing severe hypoglycaemia and 95% CI for each study. Size of box is proportional to weight of that study result. Diamonds represent the 95% CI for pooled estimates of effect and are centred on pooled hypoglycaemia incidence.

Table 2. Proportion of people experiencing hypoglycaemia and the number of episodes per patient-year by severity and treatment regime.

Eighteen papers involving 24,804 participants, measuring over 1 month to 1 year, reported data for the prevalence of unspecified type of hypoglycaemia (severity of hypoglycaemia experienced was not specified when data collecting, assumed to cover both mild/moderate and severe episodes), with a pooled prevalence of 0.43 (95% CI 0.36 to 0.50).

There was high heterogeneity shown between studies reporting mild/moderate (I2 = 98.0%), severe (I2 = 99.7%) unspecified (I2 = 99.4%) hypoglycaemia prevalence. A possible explanatory variable for this was the different time periods hypoglycaemia was measured over between studies. This along with other explanatory variables: year of publication, mean HbA1c, mean age, and percentage of male participants were carried out in a meta-regression on the prevalence of, mild/moderate, severe and unspecified hypoglycaemia. No variables were shown to be statistically significant, see Table 3.

Table 3. Meta-regression results showing the effect of study-level variables on the proportion of people experiencing each severity of hypoglycaemia.

Incidence of hypoglycaemic episodes per person-year

The pooled incidence of hypoglycaemic episodes per person-year for mild/moderate and severe hypoglycaemia was 19.03 (95% CI 0.00 to 51.08) and 0.80 (95% CI 0.00 to 2.15, Table 2), respectively. Unspecified hypoglycaemia episodes showed a pooled incidence of 27.78 (95% CI 0.00 to 58.20; per person–year). Those on insulin experienced 23.31 (95% CI 0.00 to 58.98) mild/moderate and 1.05 (95% CI 0.00 to 3.69) severe episodes per person-year. Data were not available to calculate any further treatment categories for mild/moderate, but a further incidence of 0.01 (95% CI 0.00 to 0.55) for those on sulphoylureas experiencing severe episodes was estimated.

Publication bias

There appeared to be publication bias for studies reporting prevalence of severe hypoglycaemia (p = 0.04), with studies reporting lower prevalence appearing to be missing from published literature (funnel plot can be found in S2 Fig). However, when analysis was carried out separating the study results relating to medical assistance only, from any third party assistance, there was no evidence of publication bias (p = 0.41 and; p = 0.36 respectively). No other significant publication bias was shown for prevalence (mild/moderate p = 0.09, unspecified p = 0.37) or for incidence (mild/moderate p = 0.06, severe p = 0.91, unspecified p = 0.06).


This review of 46 studies (n = 532,542) estimates that the prevalence (proportion of people) of hypoglycaemia is 45% for mild/moderate and 6% for severe in population-based studies of type 2 diabetes, and that on average an individual with T2DM experiences 19 mild/moderate episodes and 0.8 severe episodes per year. Hypoglycaemia is particularly prevalent amongst those on insulin (mild/moderate: prevalence = 52%, incidence = 23 events/ year; severe: prevalence = 21%, incidence = 1 event/year, yet still fairly common for treatment regimens that include sulphonylureas (mild/moderate: prevalence = 33%, incidence = 1.92 events/year; severe: prevalence = 5, incidence = 0.01 events/year. Severe hypoglycaemia prevalence was the same 5% for those on treatment regimens that did or did not include sulphonylureas.

Relationship to other literature

Previous literature focusing on hypoglycaemia in type 2 diabetes in a real-world setting is limited. One report of severe hypoglycaemia prevalence in clinical trials also considered severe hypoglycaemia prevalence in population based studies, but did not examine mild/moderate episodes nor provide pooled estimates [3]. Various systematic reviews have considered hypoglycaemia prevalence within randomised controlled trials involving people with type 2 diabetes [2528]. These tend to focus on individuals using insulin and their results indicate that severe hypoglycaemia prevalence is below 1%, which is substantially lower than our pooled estimate of 6% from population-based studies. Mild/moderate hypoglycaemia is reported less in clinical trials, though recent meta-analyses of clinical trials showed pooled prevalences of 10% (range 0.7%- 22%) for those on sulphonylureas [34], and 20% to 52% for those on insulin [3537]. Our estimate for sulphonylureas was higher at 33% (range 30%- 39%), but for insulin was similar, albeit at the higher end of the scale, at 52% (range 20%- 72%). Mild/moderate hypoglycaemia prevalence was fairly consistent across trials, as was the case in the population-based studies in this review. Randomised controlled trials often do not reflect real life situations as treatment regimens are more aggressively titrated than in standard clinical practice, and participants are often highly selected and generally do not include those at higher risk of hypoglycaemia. Therefore, the generalizability of results from trials is limited.

Strengths and Limitations

To our knowledge, this is the first systematic review and meta-analysis published which focuses on population-based studies of hypoglycaemia in people with type 2 diabetes. The review adhered to the Cochrane recommended standards for systematic reviews [32], involved robust methodology such as duplicate reviewing, and only included population-based studies which reported prevalence and/or incidence rates specifically for a type 2 diabetes population, making the results applicable to the type 2 diabetes general population. The results follow the American Diabetes Association Workgroups recommendations that both the proportion of patients experiencing hypoglycaemia and event rates for each severity are reported[38].

There was high heterogeneity between studies, which was not explained by any of the study level covariates considered. Therefore, the high heterogeneity is likely to be due to the very narrow confidence intervals associated with the proportion estimates, or to other study characteristics which were not measured, reported or extracted. Publication bias appeared to be a possibility for the published studies on severe hypoglycaemia. However, this appeared to be due to the different definitions of severe hypoglycaemia, since no such bias was present when studies that defined severe as requiring medical assistance were separated from those using the standard definition (requiring any third party assistance).

A strict consensus definition was not found across studies (Table 2). Due to the nature of a large proportion of the studies (questionnaire/interview/medical records), a biochemical definition was rarely used. It must be acknowledged that this does create some difficulty when collating and combining results across studies. For some studies this may have caused under or over reporting. Subsequently, the only way to define and analyse hypoglycaemia was by severity and whether or not third party assistance was required during an episode.

The way in which studies categorised different glucose lowering therapies also varied considerably, if indeed they reported hypoglycaemia by treatment categories at all. Therefore, it was only possible to use broader treatment categories in these analyses. It is important to note here also, that the event rate calculated for severe hypoglycaemia in those using sulphonylureas may appear particularly low. This is likely to be due to two of the three studies used to calculate this result only measuring “very severe” hypoglycaemia, those that specifically needed emergency treatment.

Additionally, studies varied considerably in length of time hypoglycaemia prevalence was related to; as a result a specific time period for prevalence could not be established. This was particularly the case for severe hypoglycaemia. However, a meta-regression was carried out for time hypoglycaemia was measured over on prevalence and it was not found to significantly affect the prevalence reported between studies.

Data from some studies could not be grouped by the severity of hypoglycaemia (mild/moderate or severe). These data were therefore labelled as “unspecified” and assumed to include any type of hypoglycaemic episodes. Nocturnal hypoglycaemia was not analysed due to the lack of reporting within studies.

Implications for practice

This review shows that hypoglycaemia is considerably prevalent amongst people with type 2 diabetes. We estimated that an individual with type 2 diabetes experiences one severe episode of hypoglycaemia per year on average. Severe episodes are a burden on both the individual and healthcare utilisation, due to their cost and the significant dangers that can result from an episode [1, 3]. The estimate of 19 mild/moderate episodes on average per year is also important. This quantity of mild/moderate episodes could substantially impact on work, social life and driving, as well as potentially decreasing general quality of life and increasing risk of severe events if left untreated [39, 40].

Our results highlight an urgent need for raising awareness within everyday clinical practice, particularly as prior evidence has suggested underreporting within this setting [41]. When considering treatment options, hypoglycaemia risk consideration should be incorporated through the individualisation of treatment regimens prescribed [9]. Blood glucose targets should also be individualised and in some cases a higher target may be optimal for the patient [42].

Educational programmes should be focused on successfully increasing knowledge of hypoglycaemia in relation to appropriate self-treatment methods, risk factors and predisposing symptoms [43], as this has previously been shown to be low in the type 2 diabetes population [44, 45].

Future directions

Research into mild episodes is more challenging than severe episodes in terms of reliability and access to data, with data collection methods limited to constant glucose monitoring, prospective diary recording, and retrospective recall. Constant glucose monitoring is the most reliable, but can be costly and so generally involve small populations and short data collection periods[46, 47]. Retrospective recall is convenient and least costly, though potentially less reliable.

In some studies (6/46, 13%) severity of hypoglycaemia was not specified when data were collected, and hypoglycaemia definitions also varied across studies (Tables 1 and 2). This could have led to errors in reporting, including participants interpreting the meaning of hypoglycaemia differently, with some perhaps only reporting severe episodes. Therefore, it is important that a standardised definition such as the widely used American Diabetes Association workgroup’s[38] is utilised across studies of hypoglycaemia.

More research is needed in relation to potential risks associated with particular glucose lowering therapies. Studies were inconsistent in how they reported treatment regimens, making it difficult to collate data. There is a lack of published population-based literature comparing sulphonylurea and non-sulphonylurea treatment regimens, including the newer glucose lowering therapies.


To our knowledge, this is the first systematic review and meta-analysis of the prevalence and incidence rates of hypoglycaemic events in population-based studies of people with type 2 diabetes. Hypoglycaemia is particularly prevalent amongst those on insulin, yet still fairly common for treatment regimens that do and do not include sulphonylureas. This highlights the subsequent need for educational interventions and individualisation of therapies to reduce the risk of hypoglycaemia. The current evidence also suggests that hypoglycaemia prevalence within clinical trials is likely to be an underestimate.

Supporting Information

S1 Fig. Search strategy for Ovid MEDLINE.


S2 Fig. Funnel plot of population based studies reporting severe hypoglycaemia prevalence in type 2 diabetes.



The authors acknowledge support from the National Institute for Health Research Collaboration for Leadership in Applied Health Research and Care–East Midlands (NIHR CLAHRC–EM), the Leicester Clinical Trials Unit and the NIHR Leicester-Loughborough Diet, Lifestyle and Physical Activity Biomedical Research Unit which is a partnership between University Hospitals of Leicester NHS Trust, Loughborough University and the University of Leicester.

Author Contributions

Conceived and designed the experiments: KK MJD CE AJD. Performed the experiments: CE TR AJD KK. Analyzed the data: CE DB LJG. Wrote the paper: CE AJD KK MJD TR DB LJG. Conducted the meta-analyses and meta-regression: CE. Made the final decisions regarding inclusion/exclusion of all papers: KK. Developed the search strategy, designed the data extraction and quality assessment tools: CE AJD. Conceived the idea for the review and is guarantor of the overall content: KK MJD. Selected and retrieved relevant papers, carried out extraction/checking of data and quality assessments: CE TR. Provided statistical support for conducting the meta-analyses and meta-regression with statistical support: DB LJG.


  1. 1. Hex N, Bartlett C, Wright D, Taylor M, Varley D. Estimating the current and future costs of Type 1 and Type 2 diabetes in the UK, including direct health costs and indirect societal and productivity costs. Diabetic medicine: a journal of the British Diabetic Association. 2012;29(7):855–62.
  2. 2. Feinkohl I, Aung PP, Keller M, Robertson CM, Morling JR, McLachlan S, et al. Severe Hypoglycemia and Cognitive Decline in Older People With Type 2 Diabetes: The Edinburgh Type 2 Diabetes Study. Diabetes care. 2014;37(2):507–15. pmid:24103900
  3. 3. Bloomfield HE, Greer N, Newman D, MacDonald R, Carlyle M, Fitzgerald P, et al. Predictors and Consequences of Severe Hypoglycemia in Adults with Diabetes—A Systematic Review of the Evidence. VA Evidence-based Synthesis Program Reports. Washington (DC)2012.
  4. 4. Zoungas S, Patel A, Chalmers J, de Galan BE, Li Q, Billot L, et al. Severe Hypoglycemia and Risks of Vascular Events and Death. New England Journal of Medicine. 2010;363(15):1410–8. pmid:20925543
  5. 5. Stargardt T, Gonder-Frederick L, Krobot KJ, Alexander CM. Fear of hypoglycaemia: defining a minimum clinically important difference in patients with type 2 diabetes. Health and quality of life outcomes. 2009;7:91. pmid:19849828
  6. 6. Vexiau P, Mavros P, Krishnarajah G, Lyu R, Yin D. Hypoglycaemia in patients with type 2 diabetes treated with a combination of metformin and sulphonylurea therapy in France. Diabetes, obesity & metabolism. 2008;10 Suppl 1:16–24.
  7. 7. Williams SA, Pollack MF, Dibonaventura M. Effects of hypoglycemia on health-related quality of life, treatment satisfaction and healthcare resource utilization in patients with type 2 diabetes mellitus. Diabetes research and clinical practice. 2011;91(3):363–70. pmid:21251725
  8. 8. Wild D, von Maltzahn R, Brohan E, Christensen T, Clauson P, Gonder-Frederick L. A critical review of the literature on fear of hypoglycemia in diabetes: Implications for diabetes management and patient education. Patient Education and Counseling. 2007;68(1):10–5. pmid:17582726
  9. 9. Inzucchi SE, Bergenstal RM, Buse JB, Diamant M, Ferrannini E, Nauck M, et al. Management of hyperglycaemia in type 2 diabetes: a patient-centered approach. Position statement of the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD). Diabetologia. 2012;55(6):1577–96. pmid:22526604
  10. 10. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. The Diabetes Control and Complications Trial Research Group. The New England journal of medicine. 1993;329(14):977–86. pmid:8366922
  11. 11. Excellence NIfHaC. Type 2 diabetes: The management of type 2 diabetes. London: National Institute for Health and Care Excellence: 2009.
  12. 12. Ray KK, Seshasai SR, Wijesuriya S, Sivakumaran R, Nethercott S, Preiss D, et al. Effect of intensive control of glucose on cardiovascular outcomes and death in patients with diabetes mellitus: a meta-analysis of randomised controlled trials. Lancet. 2009;373(9677):1765–72. pmid:19465231
  13. 13. Effect of intensive blood-glucose control with metformin on complications in overweight patients with type 2 diabetes (UKPDS 34). UK Prospective Diabetes Study (UKPDS) Group. Lancet. 1998;352(9131):854–65. pmid:9742977
  14. 14. Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). UK Prospective Diabetes Study (UKPDS) Group. Lancet. 1998;352(9131):837–53. pmid:9742976
  15. 15. Phung OJ, Scholle JM, Talwar M, Coleman CI. Effect of noninsulin antidiabetic drugs added to metformin therapy on glycemic control, weight gain, and hypoglycemia in type 2 diabetes. JAMA: the journal of the American Medical Association. 2010;303(14):1410–8. pmid:20388897
  16. 16. Schwarz PE. Report from the Congress of the American Diabetes Association (ADA): Orlando 2005—65th Annual Scientific Sessions in San Diego, CA, USA, June 10th-14th 2005. Experimental and clinical endocrinology & diabetes: official journal, German Society of Endocrinology [and] German Diabetes Association. 2005;113(8):475–9.
  17. 17. Banarer S, Cryer PE. Hypoglycemia in type 2 diabetes. The Medical clinics of North America. 2004;88(4):1107–16, xii-xiii. pmid:15308392
  18. 18. Henderson JN, Allen KV, Deary IJ, Frier BM. Hypoglycaemia in insulin-treated Type 2 diabetes: frequency, symptoms and impaired awareness. Diabetic medicine: a journal of the British Diabetic Association. 2003;20(12):1016–21. pmid:14632703
  19. 19. Miller CD, Phillips LS, Ziemer DC, Gallina DL, Cook CB, El-Kebbi IM. Hypoglycemia in patients with type 2 diabetes mellitus. Archives of internal medicine. 2001;161(13):1653–9. pmid:11434798
  20. 20. Murata GH, Duckworth WC, Shah JH, Wendel CS, Mohler MJ, Hoffman RM. Hypoglycemia in stable, insulin-treated veterans with type 2 diabetes: a prospective study of 1662 episodes. Journal of diabetes and its complications. 2005;19(1):10–7. pmid:15642485
  21. 21. Cryer PE, Davis SN, Shamoon H. Hypoglycemia in diabetes. Diabetes care. 2003;26(6):1902–12. pmid:12766131
  22. 22. Leese GP, Wang J, Broomhall J, Kelly P, Marsden A, Morrison W, et al. Frequency of severe hypoglycemia requiring emergency treatment in type 1 and type 2 diabetes: a population-based study of health service resource use. Diabetes care. 2003;26(4):1176–80. pmid:12663593
  23. 23. Donnelly LA, Morris AD, Frier BM, Ellis JD, Donnan PT, Durrant R, et al. Frequency and predictors of hypoglycaemia in Type 1 and insulin-treated Type 2 diabetes: a population-based study. Diabetic medicine: a journal of the British Diabetic Association. 2005;22(6):749–55.
  24. 24. Inkster B, Zammitt NN, Frier BM. Drug-induced hypoglycaemia in type 2 diabetes. Expert opinion on drug safety. 2012;11(4):597–614. pmid:22690846
  25. 25. Waugh N, Cummins E, Royle P, Clar C, Marien M, Richter B, et al. Newer agents for blood glucose control in type 2 diabetes: systematic review and economic evaluation. Health technology assessment. 2010;14(36):1–248. pmid:20646668
  26. 26. Shyangdan DS, Royle P, Clar C, Sharma P, Waugh N, Snaith A. Glucagon-like peptide analogues for type 2 diabetes mellitus. The Cochrane database of systematic reviews. 2011(10):CD006423.
  27. 27. Richter B, Bandeira-Echtler E, Bergerhoff K, Lerch CL. Dipeptidyl peptidase-4 (DPP-4) inhibitors for type 2 diabetes mellitus. The Cochrane database of systematic reviews. 2008(2):CD006739.
  28. 28. Swinnen SG, Simon AC, Holleman F, Hoekstra JB, Devries JH. Insulin detemir versus insulin glargine for type 2 diabetes mellitus. The Cochrane database of systematic reviews. 2011(7):CD006383.
  29. 29. Horvath K, Jeitler K, Berghold A, Ebrahim SH, Gratzer TW, Plank J, et al. Long-acting insulin analogues versus NPH insulin (human isophane insulin) for type 2 diabetes mellitus. The Cochrane database of systematic reviews. 2007(2):CD005613.
  30. 30. Rosenstock J, Dailey G, Massi-Benedetti M, Fritsche A, Lin Z, Salzman A. Reduced hypoglycemia risk with insulin glargine: a meta-analysis comparing insulin glargine with human NPH insulin in type 2 diabetes. Diabetes care. 2005;28(4):950–5. pmid:15793205
  31. 31. Website. EPHPP. Available from: Accessed 2015 May 7.
  32. 32. Higgins JPT GSe. Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration; 2011.
  33. 33. Wong WC, Cheung CS, Hart GJ. Development of a quality assessment tool for systematic reviews of observational studies (QATSO) of HIV prevalence in men having sex with men and associated risk behaviours. Emerging themes in epidemiology. 2008;5:23. pmid:19014686
  34. 34. Schopman JE, Simon AC, Hoefnagel SJ, Hoekstra JB, Scholten RJ, Holleman F. The incidence of mild and severe hypoglycaemia in patients with type 2 diabetes mellitus treated with sulfonylureas: a systematic review and meta-analysis. Diabetes/metabolism research and reviews. 2014;30(1):11–22. pmid:24030920
  35. 35. Bunck MC, Diamant M, Corner A, Eliasson B, Malloy JL, Shaginian RM, et al. One-year treatment with exenatide improves beta-cell function, compared with insulin glargine, in metformin-treated type 2 diabetic patients: a randomized, controlled trial. Diabetes care. 2009;32(5):762–8. pmid:19196887
  36. 36. Bergenstal R, Lewin A, Bailey T, Chang D, Gylvin T, Roberts V, et al. Efficacy and safety of biphasic insulin aspart 70/30 versus exenatide in subjects with type 2 diabetes failing to achieve glycemic control with metformin and a sulfonylurea. Current medical research and opinion. 2009;25(1):65–75. pmid:19210140
  37. 37. Gallwitz B, Bohmer M, Segiet T, Molle A, Milek K, Becker B, et al. Exenatide twice daily versus premixed insulin aspart 70/30 in metformin-treated patients with type 2 diabetes: a randomized 26-week study on glycemic control and hypoglycemia. Diabetes care. 2011;34(3):604–6. pmid:21285388
  38. 38. Seaquist ER, Anderson J, Childs B, Cryer P, Dagogo-Jack S, Fish L, et al. Hypoglycemia and Diabetes: A Report of a Workgroup of the American Diabetes Association and The Endocrine Society. J Clin Endocr Metab. 2013;98(5):1845–59. pmid:23589524
  39. 39. Brod M, Christensen T, Thomsen TL, Bushnell DM. The impact of non-severe hypoglycemic events on work productivity and diabetes management. Value in Health. 2011;14(5):665–71. pmid:21839404
  40. 40. Brod M, Pohlman B, Wolden M, Christensen T. Non-severe nocturnal hypoglycemic events: experience and impacts on patient functioning and well-being. Quality of Life Research. 2013;22(5):997–1004. pmid:22825805
  41. 41. Leiter LA, Yale J, Chiasson J, Harris SB, Kleinstiver P, Sauriol L. Assessment of the impact of fear of hypoglycemic episodes on glycemic and hypoglycemia management. Can J Diabetes. 2005;29(3):186–92.
  42. 42. Conditions. TNCCfC. Type 2 Diabetes: national clinical guideline for management in primary and secondary care(update). London: Royal College of Physicians 2008. pmid:21678628
  43. 43. Crasto W, Jarvis J, Khunti K, Skinner TC, Gray LJ, Brela J, et al. Multifactorial intervention in individuals with type 2 diabetes and microalbuminuria: the Microalbuminuria Education and Medication Optimisation (MEMO) study. Diabetes research and clinical practice. 2011;93(3):328–36. pmid:21640424
  44. 44. Browne D, Avery L, Turner B, Kerr D, Cavan DA. What do patients with diabetes know about their tablets? Diabetic medicine. 2000;17(7):528–31. pmid:10972583
  45. 45. Lund A, Knop FK. Worry vs. knowledge about treatment-associated hypoglycaemia and weight gain in type 2 diabetic patients on metformin and/or sulphonylurea. Current Medical Research & Opinion. 2012;28(5):731–6.
  46. 46. Monnier L, Wojtusciszyn A, Colette C, Owens D. The contribution of glucose variability to asymptomatic hypoglycemia in persons with type 2 diabetes. Diabetes technology & therapeutics. 2011;13(8):813–8.
  47. 47. Chico A, Vidal-Rios P, Subira M, Novials A. The continuous glucose monitoring system is useful for detecting unrecognized hypoglycemias in patients with type 1 and type 2 diabetes but is not better than frequent capillary glucose measurements for improving metabolic control. Diabetes care. 2003;26(4):1153–7. pmid:12663589