INTRODUCTION — Cystic fibrosis-related diabetes (CFRD) is a distinct form of diabetes mellitus that is an important complication of CF. It is different from either type 1 or type 2 diabetes mellitus but shares features of both [1]. The primary cause is a relative insulin deficiency related to destruction of pancreatic islets. Insulin resistance also may play a role, especially in association with acute exacerbations or chronic progression of pulmonary disease.

Development of CFRD is associated with worse lung function, poorer nutritional status, and more chest infections. In addition, longitudinal studies have demonstrated decreased survival (sixfold) in individuals with CFRD as compared with nondiabetic CF patients, with females at particularly high risk in some studies [2,3]. Importantly, insulin treatment improves lung function and nutritional status [4-6]. The addition of insulin treatment and glucose monitoring, however, adds substantially to the burden of treatment, and therapeutic targets may need to be modified to maximize benefit.

Several important management guidelines have been published. These provide recommendations on screening, diagnosis, and management. Available guidelines include those from the American Diabetes Association and Cystic Fibrosis Foundation (2010) [7,8], International Society for Pediatric and Adolescent Diabetes (2018) [9], Australian standards of care for CFRD (2014) [10], and United Kingdom Cystic Fibrosis Trust (2004) [11] (see 'Society guideline links' below). Key recommendations contained within these guidelines are discussed across various parts of this monograph.

The pathophysiology, diagnosis, and treatment of CFRD are discussed in this topic review. CF-associated lung disease is discussed in the following topic reviews:

●(See "Cystic fibrosis: Clinical manifestations of pulmonary disease".)

●(See "Cystic fibrosis: Overview of the treatment of lung disease".)

●(See "Cystic fibrosis: Treatment of acute pulmonary exacerbations".)

●(See "Cystic fibrosis: Antibiotic therapy for chronic pulmonary infection".)

●(See "Cystic fibrosis: Treatment with CFTR modulators".)

●(See "Cystic fibrosis: Management of advanced lung disease".)

Other aspects of CF care are also discussed separately:

●(See "Cystic fibrosis: Clinical manifestations and diagnosis".)

●(See "Cystic fibrosis: Genetics and pathogenesis".)

●(See "Cystic fibrosis: Overview of gastrointestinal disease".)

●(See "Cystic fibrosis: Assessment and management of pancreatic insufficiency".)

●(See "Cystic fibrosis: Nutritional issues".)

●(See "Cystic fibrosis: Hepatobiliary disease".)

EPIDEMIOLOGY AND NATURAL HISTORY — CFRD is the most common nonrespiratory complication of CF encountered among individuals school-aged and older, as shown in the figure (figure 1) [12].

Prevalence — Prevalence of CFRD increases markedly with age. The best prevalence data initially came from a CF center in Minnesota where routine annual screening using the oral glucose tolerance test (OGTT) was performed in patients six years and older since the early 1990s [13]. Prevalence of CFRD was 2 percent in children 10 years and younger, 19 percent in individuals 11 to 17 years, 40 percent in those ages 18 to 29 years, and 45 to 50 percent in those over age 30 years. A 2017 report from the Cystic Fibrosis Foundation patient registry documented CFRD in 5.3 percent of patients younger than 18 years and 31 percent of those over 18 years of age [14]. In a separate report from Germany and Austria, prevalence of CFRD was 11 percent at 20 years and 25 percent at 35 years [15]. This analysis did not include patients on glucocorticoids or who had had lung transplants; such patients would be expected to have higher risk for CFRD.

Awareness of CFRD has been increasing, and with increased awareness comes increased detection of CFRD. An Australian review found that reported incidences of CFRD had increased 10-fold between 2000 and 2008, from 2.0 to 22.1 per 1000 person-years [16]. This reported increase in CFRD diagnosis probably reflects better screening practices and longer lifespan of CF patients rather than a true increase in CFRD [16,17]. Some studies report substantially lower prevalence (eg, in the Irish CF registry, only 5.7 percent of children had CFRD [18]), but case ascertainment issues may compromise accuracy with registry-based data analysis.

Risk factors — CFRD is caused by progressive damage to the pancreas. As a result, the strongest risk factors are markers for CF-related pancreatic disease. The risk factors include:

●Pancreatic insufficiency – CFRD mainly occurs in people with exocrine pancreatic insufficiency. However, it is important to realize that CFRD has been reported in 6 percent of patients with relatively normal pancreatic exocrine function (pancreatic-sufficient) [19].

●Severe genotype – CFRD is more common in individuals with severe CF genotypes, including delta-F508 (pF508.del) homozygotes, which are associated with pancreatic insufficiency.

●Increased age – Risk for CFRD increases dramatically with age, as outlined above. (See 'Epidemiology and natural history' above.)

●Female gender – The prevalence of CFRD is higher among women, at least after 30 years of age [13]. Moreover, among patients with CFRD, women have reduced pulmonary function and survival as compared with men, and this may contribute toward the overall reduced lifespan observed for women with CF [2]. (See 'Mortality' below.)

●Lung function – CFRD is associated with greater decline in lung function (typically measured as forced expiratory volume in one second [FEV₁]) and more frequent pulmonary exacerbations [19,20]. In many cases, a marked decline in lung function precedes the diagnosis of CFRD and improves after insulin therapy, suggesting impaired pulmonary function is a consequence rather than a cause of CFRD [21,22]. (See 'Pulmonary function' below.)

●Impaired nutrition/growth – CFRD is more common among individuals with poor nutritional status. This may be explained in part by direct effects of CFRD on nutritional status. However, some of the association may be attributable to indirect effects of the severe genotype, pancreatic insufficiency, and other complications on nutritional status.

●Liver disease – CFRD is more common among individuals with CF-related liver disease. This is probably attributable to associations between CF-related liver disease and other CFRD risk factors, but a direct effect of CF-related liver disease on glucose metabolism is also possible.

Another proposed risk factor was poor sleep, which correlated with higher blood glucose and lower insulin sensitivity in adolescents with CF and dysglycemia [23].

PATHOPHYSIOLOGY

Mechanisms of impaired glucose tolerance — The primary abnormality predisposing to CFRD is slowly progressive insulin deficiency due to destruction of pancreatic tissue. Abnormal chloride channel function results in thick, viscous secretions, causing obstructive damage to the exocrine pancreas with progressive fibrosis and fatty infiltration [24]. Disruption and destruction of islet architecture leads to loss of endocrine beta, alpha, and pancreatic polypeptide cells. Data from a murine model suggest that beta cell loss and intra-islet inflammation rather than intrinsic islet dysfunction are the important mechanisms [25]. Islet interleukin-1 beta immunoreactivity may be an early contributing factor to this process [26].

The initial stages of the disease are characterized by insulin deficiency, manifested by impaired first-phase insulin secretion in response to stimulant agents [27-29]. Glucagon secretion is also impaired, reflecting destruction of the total islet (in contrast to type 1 diabetes) [9,30]. In later stages of the disease or during acute pulmonary exacerbations, CFRD also has a component of insulin resistance [31]. Indeed, insulin resistance is detectable in the earliest stages of CFRD, although to a lesser degree than impaired insulin secretion [32]. The insulin resistance may be caused by increases in growth hormone and cortisol, and increased catecholamines and inflammatory cytokines during acute pulmonary exacerbations. This mechanism may play a more prominent role in the development of CFRD in older compared with younger individuals with CF [33].

The early stage of CFRD is characterized by hyperglycemia in response to a glycemic load. In later stages, the disease tends to progress to fasting hyperglycemia (FH), at which point hemoglobin A1c (HbA1c) rises above the normal range [34]. Thus, postprandial hyperglycemia is the earliest and most sensitive marker of CFRD, and HbA1c has relatively low sensitivity but does become elevated in the later phases of the disease. (See 'Diagnosis' below.)

Secondary mechanisms contributing to diabetes in CF patients include malabsorption, liver dysfunction, and therapies such as glucocorticoids (typically used for treatment of allergic bronchopulmonary aspergillosis) and immunosuppressant therapy following lung transplantation (eg, tacrolimus).

Differences from type 1 diabetes mellitus — There are a number of important differences between CFRD and both type 1 and type 2 diabetes, which are summarized in the table (table 1) [35]. Importantly, CFRD is only rarely associated with islet autoantibodies or human leukocyte antigen (HLA) class II types that are seen in type 1 diabetes mellitus [36,37] (although classic type 1 diabetes mellitus has been described in CF patients) [37,38]. Thus, the beta-cell dysfunction seen in CFRD does not appear to have an autoimmune mechanism. Moreover, ketoacidosis rarely occurs in CFRD. Some overlap with type 2 diabetes mellitus is suggested by increased risk of CFRD in those with a family history of type 2 diabetes mellitus [39].

Several unique aspects of CF influence glucose metabolism and are not commonly encountered in other forms of diabetes. These include malnutrition, acute and chronic infection, glucagon deficiency, malabsorption, abnormal intestinal transit time, liver dysfunction, increased work of breathing and elevated energy expenditure, and exposure to some CF treatments that can precipitate glucose intolerance, such as glucocorticoids and immunosuppressants following lung/liver transplantation (eg, tacrolimus).

Clinical consequences of CFRD — There is a growing body of evidence showing that hyperglycemia and insulin insufficiency have detrimental impacts on clinical outcomes in patients with CF, even before formal diagnostic criteria for CFRD are met.

Pulmonary function — Several studies suggest detrimental effects of CFRD on both lung function and nutrition. A longitudinal study followed 152 CF subjects with varying degrees of glucose intolerance or CFRD but without FH [20]. At baseline, 45 percent of the patients had normal glucose tolerance (NGT), 39 percent had impaired glucose tolerance (IGT), and 16 percent had CFRD without FH. Forced expiratory volume in one second (FEV₁), forced vital capacity (FVC), and body mass index (BMI) were comparable among these groups. However, over the four-year follow-up, an overall decline in FEV₁ and FVC occurred. The highest rates of decline for FEV₁ and FVC were among the group with CFRD without FH. Subjects in the lowest quartile for insulin production at baseline experienced the highest rates of pulmonary function decline over time, suggesting a relationship between insulin deficiency and clinical deterioration. Similarly, data from the Swedish registry of CF subjects over the age of seven years described a greater rate of lung function decline in CFRD subjects compared with non-CFRD subjects [40]. An association between early abnormalities in lung function (ventilation inhomogeneity, or unevenness of gas mixing, as measured by multiple-breath washout) and insulin secretory defects (as measured by oral glucose tolerance test [OGTT]) has been reported [41].

Laboratory studies suggest that elevated glucose levels on the airway surface promote bacterial growth and cause an exaggerated but less effective inflammatory response, providing a mechanistic explanation for the observed association between CFRD and pulmonary function. One study demonstrated that subjects with CFRD had elevated airway glucose levels, as measured by nasal glucose monitoring, as compared with subjects without CFRD [42]. Moreover, in vitro studies demonstrated that increased glucose exposure enhanced bacterial growth of Staphylococcus aureus and Pseudomonas aeruginosa, with significant changes occurring at far lower glucose levels than those documented in CF subjects. A separate study in a mouse model of CF showed that airway hyperglycemia was associated with impaired ability to clear bacteria from the lung, despite enhanced recruitment of neutrophils to the airways [43]. Several other effects of hyperglycemia have been reported: increased lactate generation and efflux, which acidify the airway surface liquid [44]; a deleterious effect on transepithelial ion transport and epithelial repair functions [45]; and proinflammatory changes with associated impairment of airway epithelial cell potassium channel function, which is important in mucus clearance [46]. The airway glucose barrier is regulated by insulin and is dysfunctional in CF [47].

CF patients undergoing lung transplantation have more complications and a higher mortality rate if they have preceding CFRD as compared with those without CFRD [48,49]. It is hoped that with lower targets and earlier treatment of diabetes, these differences in outcome will diminish.

Nutritional status — CFRD is associated with poor nutritional status and, particularly, with a decline in nutritional status prior to its diagnosis. The nutritional impact of CFRD can be explained by the fact that insulin is a potent anabolic hormone that plays an important role in maintaining body weight and lean body mass. Insulin deficiency promotes a catabolic state with detrimental effects on nutritional outcomes. As an example, patients with CFRD (without FH) experienced a decline in BMI of 0.3±0.21 units during an observation period, and this was reversed after the initiation of insulin therapy [50]. Detrimental effects of CFRD also have been shown in milder CFRD categories. In a cohort of CF subjects with NGT (as defined by the OGTT), subclinical deficiencies of insulin secretion were detected by measuring the area under the curve for insulin secretion over the 120 minutes of the OGTT, and greater impairment in insulin secretion was associated with the lowest BMI [51]. Similarly, another study demonstrated that abnormalities in glucose homeostasis prior to the development of overt CFRD were associated with preceding declines in weight Z-score [52].

Vascular complications — Multiple reports have demonstrated that CFRD is associated with significant microvascular complications, including retinopathy, neuropathy, and nephropathy, which depend on disease duration and glycemic control. Diabetic gastroenteropathy including gastroparesis also occurs in CFRD and may contribute to gastrointestinal symptoms. Because the understanding and management of CFRD is still rapidly evolving, the precise prevalence of these complications and their responsiveness to treatment are not well established.

As an example, in a large cohort of patients with CFRD, peripheral neuropathy and symptoms consistent with diabetic gastroenteropathy were each seen in 52 percent of patients [53]. This frequency of neuropathy and diabetic gastroenteropathy was similar to that in patients with longstanding type 1 or type 2 diabetes. Among those with FH, 16 percent had retinopathy and 14 percent had nephropathy (indicated by moderately elevated albuminuria). These frequencies were somewhat lower than for type 1 diabetes, but the complications tended to appear at lower levels of HbA1c [53,54]. Separate case reports have described severe retinopathy in individuals with CFRD and poor glycemic control, with proliferative retinopathy complicated by blindness despite laser therapy to the retina [55]. Other reports from CF clinics have also described peripheral neuropathy, retinopathy, and nephropathy, which were generally associated with poor glycemic control and arose after 10 years of diabetes duration [56-58]. Microvascular complications have also been documented after short diabetes duration [59]. Antibiotic use also may contribute to the risk of peripheral neuropathy. Regular screening for microvascular complications is recommended. (See 'Monitoring for complications' below.)

Macrovascular disease also occurs in CFRD, although it is uncommon. Myocardial disease has been found at postmortem examinations [60], and symptomatic myocardial infarction has been described in a few case reports [61,62]. Noninvasive measures of arterial stiffness, as a marker of large vessel disease, are higher in patients with CF and CFRD as compared with non-CF controls, without increases in blood pressure [63], supporting the notion that CFRD increases the risk for cardiovascular disease.

Mortality — CFRD has been associated with increased mortality, especially in female subjects, and this detrimental effect is greatly attenuated by early diagnosis and treatment. One report described the clinical course of 872 CF patients treated at a single center in Minnesota during three consecutive intervals: 1992 to 1997, 1998 to 2002, and 2003 to 2008 [13]. Mortality in patients with CFRD significantly decreased over time, as did overall mortality. In female subjects, the mortality rate halved from 6.9 to 3.2 deaths per 100 patient-years between the first and last study periods; in male subjects, the mortality rate dropped from 6.5 to 3.8 deaths per 100 patient-years. An analysis of patient data from 2008 to 2012 found that the mortality rate of individuals with CFRD was still 3.5 times greater than those without CFRD, suggesting that aggressive CFRD screening and management has yet to show the expected benefit in mortality rates of older CF subjects (>30 years of age) [64]. Thus, early diagnosis and effective treatment for patients with CFRD eliminated the gender gap in mortality and also narrowed the gap in mortality between CF patients with and without diabetes.

SCREENING — As outlined above, CFRD is often clinically silent but increases morbidity and mortality, and these effects can be attenuated by early diagnosis and treatment. Therefore, rigorous screening for the disorder is warranted.

When to screen — We recommend annual screening for CFRD in all patients with CF beginning by 10 years of age, consistent with guidelines from the American Diabetes Association, Cystic Fibrosis Foundation, Pediatric Endocrine Society (PES), European Society for Clinical Nutrition and Metabolism (ESPEN) and International Society for Pediatric and Adolescent Diabetes [8,9,65,66].

It is possible that screening will be extended to younger children in the future; some centers already begin screening around six years of age because abnormal glucose tolerance at this age predicts early progression to CFRD [67]. The emergence of abnormal glucose tolerance may occur even earlier and was detected in almost 40 percent of CF subjects between three months and five years of age at one CF center [68]. Screening should be performed during a period of baseline health (ie, not during a pulmonary exacerbation), using an oral glucose tolerance test (OGTT), as discussed below, or possibly continuous glucose monitoring (CGM). (See 'Diagnosis' below.)

In addition to annual screening for all patients, screening is suggested in the following situations [8]. Interpretation of the results is discussed below. (See 'Other diagnostic tests' below.)

●Patients with a significant acute pulmonary exacerbation (eg, requiring intravenous antibiotics or systemic glucocorticoids) – Screen by measuring fasting and postprandial blood glucose concentrations for the first 48 hours of treatment.

●Patients with enteral tube feedings – Screen by measuring blood glucose concentrations midway through and immediately after a feed, at the time of gastrostomy feeding initiation and monthly thereafter.

●Pregnancy – Screen with OGTT prior to conception (if possible), at the end of both the first and second trimesters, and again 6 to 12 weeks after delivery [8]. Women with CF are at increased risk for gestational diabetes, but positive results of testing during pregnancy are considered to be gestational diabetes rather than CFRD. CFRD is diagnosed only if the diabetes persists after delivery.

Screening tests

Oral glucose tolerance test — The best test for screening and diagnosis of CFRD is the OGTT. Interpretation of the results is outlined in the table (table 2) and discussed below. (See 'Interpretation of the oral glucose tolerance test' below.)

Other tests

●Not recommended – Tests not recommended for screening purposes (due to low sensitivity) include:

•Hemoglobin A1c (HbA1c) – If HbA1c is measured, a value ≥6.5 percent is consistent with a diagnosis of CFRD, as for other types of diabetes. This threshold has low sensitivity for detecting CFRD, with only 16 percent of CF patients having elevated HbA1c values at the time of CFRD diagnosis [69]. The possibility of screening with HbA1c using a lower threshold of 5.8 percent has been proposed [70], but this test still performs poorly for diagnosis in comparison with OGTT [71]. Even lower thresholds (5.5 percent) may be appropriate, with sensitivity of 91 to 95 percent compared with the OGTT [72,73].

•Clinical symptoms – Symptoms are not a good method for screening for CFRD. In a population of pediatric CF patients in Toronto, only 2.7 percent were clinically recognized as having CFRD, but on OGTT testing of asymptomatic adolescents (aged 10 to 18 years), 17 percent were found to have impaired glucose tolerance (IGT) and 13 percent had CFRD (without fasting hyperglycemia [FH]) [74]. In this cohort, abnormal glucose tolerance was almost exclusively found in those with pancreatic insufficiency and severe (class 1 to 3) variants in the CFTR gene (cystic fibrosis transmembrane conductance regulator).

●Possibly useful – Tests that may be useful for identifying early stages of abnormal glucose intolerance (high sensitivity, clinical utility unclear) include:

•CGM – CGM has not been formally established as a screening procedure for CFRD, but studies suggest that this method can identify early abnormalities in glucose tolerance that are not detected by OGTT [72,75-77], as illustrated by the following examples:

-One study compared CF subjects with normal OGTT results and normal body weight with age-matched controls (mean age 27 years, mean body mass index [BMI] 22.3) [78]. Compared with controls, the individuals with CF had postprandial hyperglycemia. On OGTT, plasma glucose values at 30, 60, and 90 minutes after the glucose load were higher than in controls. On CGM, the mean glucose was higher (106 versus 92 mg/dL [5.9 versus 5.1 mmol/L]) and a larger proportion had glucose excursions into "diabetic range" over 200 mg/dL (33 versus 5 percent). The highest value during OGTT for the group was at 60 minutes: 160 versus 112 mg/dL (8.9 versus 6.2 mmol/L).

-The clinical significance of abnormalities detected by CGM remain unclear, but data are emerging that abnormalities in CGM glucose values are associated with adverse clinical outcomes. Among patients with CF and normal OGTT results, those with higher maximum glucose values during CGM monitoring (≥11 mmol/L [200 mg/dL]) had worse pulmonary function and more P. aeruginosa colonization compared with those with lower maximum glucose values (<11 mmol/L) [51]. In participants aged 10 to 18 years, positive correlations between CGM abnormalities and the rate of lung function decline have been reported [79]. In adults, CGM glucose levels >200 mg/dL predicted the development of later CFRD [80]. In children under six years of age, transient glucose excursions are associated with P. aeruginosa infection and pulmonary inflammation compared with no glucose excursions [81].

The optimal CGM metric to use also remains unclear. Examples of metrics explored are minimum daytime glucose; interquartile range; peak glucose; excursions >140 mg/dL/day (>7.8 mmol/L); percent of time within the acceptable range or, alternatively, time >140 mg/dL (>7.8 mmol/L); and standard deviation and mean amplitude of glycemic excursions [23,79,82].

The clinical significance of these abnormalities detected by CGM needs further clarification. CGM has been validated for use in children and adolescents with CF [83-85] and is a useful tool for managing insulin therapy in those with CFRD [72] (see 'Continuous glucose monitoring' below). At present, CGM should be considered in patients whose clinical course suggests the possibility of CFRD but the OGTT result is normal. Formal glucose thresholds identifying risk of CFRD have been proposed but require further validation [80]. In addition, the beneficial effect of treating CGM-detected abnormalities remains unclear, although one retrospective study suggested a beneficial effect of insulin on lung function and weight outcomes in adult subjects with CGM evidence of hyperglycemia [86].

•Insulin concentrations during OGTT – Measurements of insulin concentrations during the OGTT can detect early abnormalities of insulin secretion, which is the hallmark of CFRD. This was shown in a study that compared insulin secretion during a standard OGTT in individuals with CF and normal or IGT (as determined by plasma glucose measurements during the OGTT), compared with individuals without CF [87,88]. In the individuals with CF, the peak insulin concentrations were lower and occurred later (insulin peak two hours after glucose load, compared with 30 to 60 minutes in the control). Glucagon responses to stimulation by intravenous arginine were blunted, suggesting defects in function of alpha as well as beta cells of the pancreas. At present, such measures of insulin secretion are not part of the formal diagnostic criteria but may give further information in those with suspected CFRD.

DIAGNOSIS — The oral glucose tolerance test (OGTT) is the test of choice for diagnosis of CFRD and prediabetic stages of the disease. It is performed using a glucose load of 1.75 g/kg body weight (up to a maximum of 75 grams). Plasma glucose concentrations should be measured in the fasting state and both one and two hours after the glucose load. The one-hour measurement is important to detect indeterminate glycemia, an earlier stage of abnormal glucose tolerance, in which this midpoint value is elevated but the two-hour level is not.

Interpretation of the oral glucose tolerance test — CFRD represents one end of a spectrum of abnormalities of glucose tolerance in cystic fibrosis (CF). This spectrum is reflected in the diagnostic categories used for interpretation of the OGTT (table 2):

●Normal glucose tolerance (NGT) – Fasting plasma glucose <100 mg/dL (5.6 mmol/L) and two-hour plasma glucose <140 mg/dL (7.8 mmol/L)

●Impaired fasting glucose (IFG) – Fasting plasma glucose 100 to 125 mg/dL (5.6 to 7 mmol/L)

●Indeterminate glycemia (INDET) – One-hour plasma glucose peak >200 mg/dL (11.1 mmol/L), but normal two-hour plasma glucose and fasting plasma glucose

●Impaired glucose tolerance (IGT) – Two-hour plasma glucose 140 to 200 mg/dL (7.8 to 11.1 mmol/L)

●CFRD:

•CFRD without fasting hyperglycemia (FH) – Two-hour plasma glucose ≥200 mg/dL (11.1 mmol/L) or fasting plasma glucose <126 mg/dL (7.0 mmol/L)

•CFRD with FH – Two-hour plasma glucose ≥200 mg/dL (11.1 mmol/L) or fasting plasma glucose ≥126 mg/dL (7.0 mmol/L)

The categories of IGT and CFRD are defined by the same plasma glucose thresholds that are used for type 1 and type 2 diabetes. The INDET category is unique to CF and is characterized by mild, transient postprandial hyperglycemia, with a normal baseline plasma glucose concentration, midpoint plasma glucose peak of >11.1 mmol/L, and return to normal by the two-hour timepoint [89]. The two categories of abnormal glucose tolerance (INDET and IGT) are important risk factors for progression to overt diabetes [34]. In one study, subjects classified as INDET on OGTT were 10 times more likely to develop CFRD over the study period compared with subjects classified as having IFG or NGT [90]. Patients with CFRD are often subcategorized by the absence or presence of FH because FH usually represents a more advanced stage of disease and is associated with microvascular complications [8]. However, because treatment is beneficial for patients with and without FH, distinguishing between these groups may not affect clinical management [8,9,50].

The plasma glucose threshold for the diagnosis of CFRD is based on the plasma glucose concentrations at which intervention has been shown to prevent microvascular and macrovascular complications in type 1 or type 2 diabetes. The prediabetic categories (INDET and IGT) are included in the classification system for CF patients in recognition that the plasma glucose threshold that corresponds to pulmonary deterioration and increased mortality is not yet known. As an example, beta cell secretory defects have been described in subjects with pancreatic insufficiency and one-hour OGTT glucose ≥155 mg/dL (8.6 mmol/L) [91]. The clinical significance of these categories will become more apparent as comprehensive data become available, describing the natural history of CFRD and its pulmonary, nutritional, and microvascular complications. (See 'Clinical consequences of CFRD' above.)

An additional category has been proposed: a one-hour glucose-stimulated value above 140 mg/dL (7.8mmol/L) on OGTT, based on publications describing early signs of insulin defects in the pathogenesis of CFRD [32,52]. This pattern indicates subtle impairment in insulin sensitivity and secretion, which becomes more pronounced with IGT, INDET and CFRD. This category has not yet been included in guidelines.

A single abnormal OGTT may not indicate the need for treatment but identifies patients with abnormal glucose tolerance (INDET or IGT) and the need for close monitoring. Close monitoring may include repeating the OGTT sooner than the planned annual screen, intensive monitoring of blood glucose during pulmonary infections, and attention to changes in nutritional or pulmonary status, which often signal the onset of CFRD. While there is a general trend from INDET to frank CFRD with FH, fluctuation between severity stages may occur. The OGTT result may vary over time, as illustrated in a four-year study that showed deterioration of glucose tolerance in 22 percent and improvement of glucose tolerance in 18 percent [20]. In another prospective study, 58 percent of those with IGT had normal subsequent OGTT results [69]. These changes in diagnostic categories may be due to acute effects of clinical status on insulin resistance [92].

IFG is defined as fasting plasma glucose between 100 and 125 mg/dL (5.6 to 7.0 mmol/L). IFG is not generally a diagnostic category in the CFRD spectrum. However, a registry study from Germany and Austria found that IFG was present in 25 percent of individuals with CF at 14 years of age and was more common than IGT at all ages [15].

Other diagnostic tests — As an alternative to diagnosing CFRD based on results of an OGTT, other tests may qualify for the diagnosis under certain circumstances [8]:

●CFRD may be diagnosed during a period of stable baseline health if a patient meets other standard criteria for diagnosis of diabetes on two occasions (hemoglobin A1c [HbA1c] ≥6.5 percent, fasting plasma glucose ≥126 mg/dL [7 mmol/L]) or classical symptoms of diabetes in the presence of a random plasma glucose concentration ≥200 mg/dL (11.1 mmol/L). However, it should be noted that HbA1c and fasting plasma glucose have low sensitivity for detecting CFRD, so normal values do not exclude the diagnosis.

●In patients on enteral tube feedings, CFRD may be diagnosed if mid- or post-feeding plasma glucose levels are ≥200 mg/dL (11.1 mmol/L) on two separate days.

●In CF patients with acute illness (eg, those hospitalized with a pulmonary exacerbation), the diagnosis of CFRD can be made if fasting plasma glucose is ≥126 mg/dL (7 mmol/L) or postprandial plasma glucose is ≥200 mg/dL, persistently for more than 48 hours. (See 'Screening tests' above.)

TREATMENT

Insulin therapy — We recommend treatment with insulin therapy for all patients with CFRD. The insulin dose, regimen, and glycemic targets differ from those used for type 1 and type 2 diabetes, as outlined below. General principles for CFRD management are:

●Give as much insulin as can be safely tolerated to eliminate the catabolic effects of CFRD

●Maintain hemoglobin A1c (HbA1c) as low as possible

•HbA1c target is <7 percent and ideally in the lower part of the normal range (<5.5 percent)

•An increase in HbA1c above the baseline level (prior to insulin therapy) suggests that insufficient insulin is being given

●Test blood glucose preprandially (fasting) and postprandially (approximately one hour after eating)

•If preprandial hypoglycemia is detected, reduce basal insulin dose

•If postprandial hyperglycemia is detected, increase basal insulin dose or add prandial insulin to maximize benefits for the lung

When to intervene — Insulin therapy is the recommended treatment for CFRD, both with or without fasting hyperglycemia (FH). Insulin therapy may also be considered on a case-by-case basis for subjects with evidence of milder dysglycemia (eg, indeterminate glycemia [INDET] or impaired glucose tolerance [IGT]) if this appears to be causing polyuria, polydipsia, nutritional decline, or deterioration in lung function. Although insulin treatment has important benefits, it also adds substantially to the burden of cystic fibrosis (CF) treatment. The treatment burden is an important factor when deciding whether to start treatment for patients with INDET or IGT and when choosing among insulin regimens.

For patients with established CFRD, we recommend initiating insulin therapy because this clearly has beneficial effects on nutrition and probably improves pulmonary function and survival. In a randomized trial, adult patients with CFRD without FH gained 0.39±0.21 body mass index (BMI) units during one year of insulin therapy, which reversed the BMI decline that they had experienced during the year prior to treatment [50] (see 'Nutritional status' above). Among the insulin-treated subjects, there was a trend toward slower rate of decline in pulmonary function compared with the rate of decline prior to treatment; insulin-treated subjects also tended to have a slower overall decline in pulmonary function compared with placebo-treated subjects, but neither of these findings reached statistical significance [50]. Multiple small observational studies with one to three years follow-up confirm improvements in BMI after initiation of insulin therapy and also document improvements in pulmonary function, as measured by forced expiratory volume in one second (FEV₁), or reductions in acute pulmonary infections [5,93-96]. As an example, among 42 young adults with CFRD, insulin therapy slowed the annual rate of decline for FEV₁, delaying the decline by an average of 34 months [5]. CFRD-associated mortality was described in an observational study from a single center and documented progressively lower mortality rates over a 20-year period, which narrowed the gap in mortality between CF patients with and without diabetes and eliminated the gender gap in mortality [13]. Because the center progressively improved protocols for early identification and treatment of CFRD, this study provides indirect evidence that insulin therapy reduces CFRD-associated mortality. (See 'Mortality' above.)

For patients with INDET or IGT, we suggest insulin therapy for selected patients. In our practice, we initiate insulin therapy for selected patients with INDET or IGT if there is an acute decline in nutritional status or lung function, or symptoms of hyperglycemia (polyuria or polydipsia). This practice is based on limited evidence from small observational studies:

●A study from the United Kingdom described clinical benefit from insulin treatment in four patients with normal oral glucose tolerance tests (OGTTs) but elevated random blood glucose (200 to 325 mg/dL, 11.1 to 18 mmol/L), associated with deteriorating respiratory function and weight loss [21]. After three months of insulin therapy at low doses (total of 6 to 12 units daily), each patient had improved body weight and lung function, as measured by spirometry.

●Two open-labelled studies from Naples described the effects of early use of basal insulin in CF with dysglycemia and CFRD:

•Four patients with intermittent CFRD (ie, who previously required insulin only during acute pulmonary exacerbations) were treated with insulin glargine (0.3 units/kg/day) [97]. The number of lung infections decreased by approximately 50 percent compared with baseline (from 2.75±0.50 to 1.25±0.5), although there was no change in HbA1c or BMI. In a control group with intermittent CFRD who were not treated with basal insulin, there was no change in the frequency of lung infection. A separate group of four patients with established CFRD previously treated with prandial insulin also experienced a decrease in the frequency of lung infection after basal insulin was added to their regimen.

•A second study from the same group evaluated the effect of insulin glargine given for 12 months to 22 pediatric patients with CF (mean age 12.4 years, range 2.6 to 19 years) and various degrees of dysglycemia: nine with CFRD, nine with IGT, and four with abnormal results of continuous glucose monitoring (CGM) but with normal OGTT [93]. All subjects were treated with 0.2 units/kg of insulin glargine before breakfast and titrated to a mean dose of 0.23 units/kg/day, aiming for blood glucose levels between 70 and 140 mg/dL (3.9 to 7.7 mmol). Those who were most underweight showed an improvement in BMI Z-score compared with pretreatment values. For the whole group, there was improvement in lung function, with an 8.8 percent increase in FEV₁ and a 42 percent decrease in pulmonary infections.

●In a pediatric study from Sydney, Australia, 12 patients with early insulin deficiency (most in the INDET category) and six with CFRD were treated with basal insulin [98]. Insulin detemir was begun at 0.1 units/kg/day and titrated to a mean of 0.13 units/kg/day, aiming for random blood glucose levels between 72 and 144 mg/dL (4 to 8 mmol/L). Benefits were seen in both nutritional and lung function outcomes, with reversal of weight loss that had occurred during the 12 months prior to insulin therapy, and improvement in lung function. In the overall group, FEV₁ improved by 3.7±10.6 percent; in the subgroup with early insulin deficiency, FEV₁ improved by 5.3±11.5 percent.

Suggested protocol — During clinical stability, patients with CFRD typically require 0.5 to 0.8 units/kg/day [9]. Insulin doses should be adjusted up to the maximum that can be safely tolerated to eliminate the catabolic effects of CFRD [9]. Considerably higher insulin doses may be required during acute pulmonary exacerbations or other stress conditions. There is considerable variation in dosing in the literature, with a mean dose of 0.8 units/kg reported from Germany and Austria [99]. The reason for higher insulin dose in this cohort is not clear, as mean HbA1c was 7.1 percent in adolescents less than 21 years, comparable with the 6.9 percent achieved in another large cohort [100].

At our institution, we vary the insulin regimen depending on the patient's characteristics and response (table 3A-B):

●Early in the course of disease, a single daily basal insulin injection usually can achieve desired outcomes and has lower treatment burden than regimens that include multiple daily insulin injections. In our practice, we initiate insulin therapy with a long-acting insulin analog (such as insulin detemir in the morning) at 0.1 units/kg. We then increase the dose by increments of 0.1 unit/kg, aiming for postprandial blood glucose <140 mg/dL (7.8 mmol/L) without hypoglycemia.

•If hypoglycemia (blood glucose <70 mg/dL [<3.9 mmol/L]) occurs prior to the midday meal, then we split the dose of detemir to two-thirds in the morning and one-third in the evening, and increase the dose.

•If fasting hypoglycemia develops and postprandial hyperglycemia persists, then we add a prandial bolus prior to meals while reducing the dose of basal insulin. The prandial bolus is a rapid-acting insulin, with a starting dose of 0.5 to 1 unit per 15 grams carbohydrate.

●Alternatively, basal insulin can be started at a dose of 0.1 to 0.2 units/kg together with prandial insulin 0.5 to 1 unit per 15 grams carbohydrate. In addition to prandial insulin, correction for ambient hyperglycemia is performed with rapid-acting insulin using a starting dose (also referred to as a correction factor) of 1 unit to lower blood glucose by 3 to 6 mmol/L (54 to 108 mg/dL) to a target of 5 mmol/L (90 mg/dL).

●During hospitalizations and when glucocorticoids are used, the starting dose for basal insulin is generally 0.2 units/kg. Most patients also require prandial insulin during intercurrent infections.

●For those on overnight feeds, insulin is given in the evening, using a long-acting insulin (eg, detemir) or a combination of short-acting and intermediate-acting insulin (eg, regular insulin and NPH [neutral protamine hagedorn]). Blood glucose should be checked at four to five hours after commencement and at the end of feeds to titrate the dose. Alternately, intermediate-acting insulin can be given mixed with short-acting insulin before feeds.

●Continuous subcutaneous insulin infusion pumps can provide better matching of carbohydrate-to-insulin requirements, especially during overnight feeds. Square wave boluses and dual wave boluses can help control blood glucose levels. Substantially more insulin is needed for meals than for basal insulin requirements. For overnight feeds, an increase in the rate of basal insulin delivery over the period of feeding is a reasonable alternative to square wave boluses. This approach permits administration of correction boluses at the start of feeds, which otherwise cannot be programmed with a compound carbohydrate (meal) bolus.

For surgical procedures such as port insertions, insulin usually can be withheld without risk of significant hyperglycemia. CFRD is rarely associated with ketoacidosis, so risk of developing ketosis during periods without insulin is very low.

Diet and drinks — Most patients with CF benefit from dietary strategies to increase caloric intake, including calorie-rich beverages. However, for patients with CFRD or abnormal glucose tolerance (INDET or IGT), beverages with high carbohydrate loads such as carbonated soft drinks and juices will add to the glycemic burden and increase insulin needs. Therefore, patients with any degree of abnormal glucose tolerance are optimally managed with a diet with low glycemic load, including avoidance of high-sugar beverages [101]. In CFRD subjects where glycemic control remains challenging despite efforts to optimize insulin, strategies such as increased intake of low glycemic index carbohydrate, reduction of foods with high glycemic index (eg, soft drinks), and distribution of carbohydrate evenly throughout the day may help optimize blood glucose control without compromising total caloric intake.

Other than these considerations for glycemic load, dietary goals are identical to those for all patients with CF, including [8,35,102]:

●Liberal intake of total energy (calories) – Typically 120 to 150 percent or more of the estimated energy requirement for age and sex, targeting BMI ≥50^th percentile for age and sex. Recommended intakes of energy and key nutrients for children without CF are summarized in the table, for comparison (table 4). (See "Dietary history and recommended dietary intake in children", section on 'Energy needs'.)

●Liberal intake of fat – eg, 40 percent of total energy.

●Moderate proportion of carbohydrates – eg, 45 to 50 percent of total energy (ideally with low glycemic load for patients with CFRD or abnormal glucose tolerance).

●High intake of protein – eg, 1.5 to 2 times the dietary reference intake for age, including in patients with nephropathy.

●High intake of salt (unrestricted intake due to increased requirement).

These dietary goals differ from those for patients with type 1 or type 2 diabetes mellitus, for whom dietary recommendations restrict total energy, fat, protein, and sodium. Nutritional management of patients with CF is discussed in detail in a separate topic review. (See "Cystic fibrosis: Nutritional issues".)

Glycemic targets and adjustment of insulin

Self-monitoring of blood glucose — Patients with CFRD on insulin therapy should ideally perform self-monitoring of blood glucose (SMBG; via "fingerstick") at least three times a day, as is recommended for all individuals with diabetes [8,9]. Less frequent but staggered SMBG monitoring to provide an overall picture of glucose control may be suitable in selected patients.

●Preprandial blood glucose – SMBG should be performed prior to meals (fasting), especially during periods of illness or changes in insulin dosing, and routinely during periods of baseline health. Target fasting blood glucose is 70 to 90 mg/dL (3.9 to 5 mmol/L) when on basal insulin. If preprandial hypoglycemia is detected, the basal insulin dose should be decreased. Preprandial testing is important because spontaneous hypoglycemia is common among individuals with CF, even in the absence of insulin therapy [103,104].

CF patients are prone to hypoglycemia in the fasting state because of malnutrition and/or increased energy needs, and also to postprandial hypoglycemia (reactive hypoglycemia), due to delayed and disordered insulin secretion. They tend to have reduced glucagon response to hypoglycemia but normal catecholamine response [8]. Patients on insulin therapy should be educated about the risks and management of hypoglycemia, including the use of glucagon.

●Postprandial blood glucose – SMBG also should be performed one to two hours after meals. If hyperglycemia is detected, the basal insulin dose should be increased or a prandial dose of rapid-acting insulin should be added; the goal is to maintain postprandial blood glucose in the normal range (<140 mg/dL [7.8 mmol/L]). More frequent SMBG may be needed after changes in insulin dose or during acute illness. Postprandial testing is important to maximize the benefit of insulin therapy on pulmonary function, and the optimal time point for testing may be one hour after meals. This was suggested by a study of young adolescents with CF in good baseline health, in which the one-hour glucose excursion on OGTT was negatively associated with lung function; this was not the case for the two-hour stimulated value, which is typically used to make the diagnosis of diabetes [105].

Hemoglobin A1c — Although HbA1c is not recommended as a screening test for CFRD, it is helpful in monitoring treatment and should be measured every three months in patients on insulin treatment [8,9]. The goals and glycemic targets for CFRD differ from those used for type 1 diabetes:

●For patients with CFRD, we suggest trying to maintain HbA1c as low as possible, ideally in the lower part of the normal range (eg, HbA1c <5.5 percent). This target is selected to optimize lung function and reduce pulmonary exacerbations. At diagnosis, patients with CFRD often have HbA1c in the nondiabetic range (<6 percent) or in the borderline dysglycemic range (6 to 6.5 percent). In part, this is because HbA1c is affected by the shorter lifespan of red blood cells in chronic disease. If the HbA1c increases above the baseline level (measured prior to insulin therapy), the insulin dose usually should be increased. In one study, good CFRD control (defined in this study as HbA1c <7 percent) was not associated with fewer pulmonary infections, but it is possible that a benefit would have been found if a more stringent HbA1c threshold had been used [106].

●By contrast, in type 1 diabetes, treatment goals are to achieve HbA1c <7 percent for most adolescents and adults [107,108]. This is because the primary goals of treatment for type 1 diabetes are to reduce risks for microvascular complications and to avoid acute episodes of hyper- and hypoglycemia.

Continuous glucose monitoring — CGM provides frequent measurements of glycemia and insight about insulin needs and responsiveness. In theory, if CGM helps to avoid blood glucose excursions above the normal range, this may improve lung function. In clinical practice, CGM can be useful for patients whose CFRD is difficult to manage through standard approaches, such as those with poor growth despite insulin therapy. CGM has been validated in patients with CF and is comparable with non-CF diabetic individuals [85]. CGM measures interstitial glucose and reflects glucose trends but not absolute plasma glucose levels.

Monitoring for complications — Patients with CFRD should undergo routine monitoring for complications, as outlined in the table (table 5) and detailed below.

●Microvascular complications – CFRD leads to chronic microvascular complications, including retinopathy, nephropathy, and neuropathy. The frequency of these complications depends on disease duration and glycemic control and appears to be similar to that for type 1 diabetes, although in CFRD, they tend to appear at lower levels of HbA1c. (See 'Vascular complications' above.)

Annual screening for microvascular complications is recommended, beginning five years after the diagnosis of CFRD or at the time that CFRD with FH is first diagnosed (whichever is earlier) [8]. Screening procedures are the same as for patients with type 1 diabetes and consist of dilated eye examination for retinopathy, urine albumin:creatinine ratio (spot specimen), and sensory examination of the feet for peripheral neuropathy. (See "Complications and screening in children and adolescents with type 1 diabetes mellitus", section on 'Vascular complications'.)

●Hypertension – Hypertension is a concern in CFRD primarily because it may contribute to nephropathy. Measurement of blood pressure is recommended at all routine visits [8]. If hypertension is present, it should be treated as recommended for other people with diabetes, usually with an angiotensin-converting enzyme (ACE) inhibitor or angiotensin II receptor blocker (ARB). (See "Complications and screening in children and adolescents with type 1 diabetes mellitus", section on 'Hypertension'.)

●Dyslipidemia – Hyperlipidemia and macrovascular disease are rare in patients with CF except in those with pancreatic exocrine sufficiency (which is particularly uncommon among patients with CFRD). Annual screening with a lipid panel is recommended for patients with CFRD and pancreatic sufficiency, or with risk factors for cardiovascular disease (post-transplantation, obesity, or family history of early cardiovascular disease) [8].

Oral hypoglycemic agents — Use of oral hypoglycemic agents to augment insulin production have been largely unsuccessful. In a randomized trial, treatment with repaglinide yielded transient improvement in IGT during the first six months of therapy but was not sustained at 12 months, whereas insulin therapy had an ongoing positive effect on weight gain [50]. These findings may reflect the fact that oral hypoglycemic agents have limited ability to stimulate the failing beta cell mass in patients with CFRD. A subsequent multicenter randomized trial comparing insulin and repaglinide therapy over two years showed no differences in HbA1c, BMI, or lung function. There were no differences in adverse events between the two treatment strategies [109]. While this does suggest some utility of repaglinide, the results of the study should be interpreted with caution, as outlined by the authors of the International Society for Pediatric and Adolescent Diabetes guideline [9]. Limitations of the study included high dropout rates, considerable variation in insulin dose administered across centers, prolonged recruitment period across a large number of centers, and, finally, the fact that there was no improvement in either treatment group. Oral diabetes agents are not currently recommended in CFRD [9].

Effect of CFTR modulator treatments — A new class of drugs known as cystic fibrosis transmembrane conductance regulator (CFTR) modulators target the underlying defect in CFTR function [110]. Emerging evidence suggests that these drugs can have wide-ranging effects on important CF outcomes. Their effect on CFRD has not been fully established, but preliminary results suggest possible benefits [111,112]. One observational study described reduced insulin requirements with ivacaftor treatment in adult patients with ivacaftor-sensitive CFTR mutations [113].

Details of CFTR modulator therapy, including selection based on genotype and efficacy for other CF-related outcomes, are discussed in a separate topic review. (See "Cystic fibrosis: Treatment with CFTR modulators".)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Cystic fibrosis".)

SUMMARY AND RECOMMENDATIONS

●Cystic fibrosis-related diabetes (CFRD) is the most common complication of cystic fibrosis (CF) encountered among individuals school-aged and older (figure 1). Its prevalence among people with CF increases markedly with age, rising to almost 50 percent in those over age 30 years. Other risk factors for developing CFRD include severe genotype (eg, pF508.del homozygotes), pancreatic insufficiency, and female sex. (See 'Epidemiology and natural history' above.)

●The primary abnormality predisposing to CFRD is slowly progressive insulin deficiency due to fibrosis and atrophy of pancreatic tissue. Acute pulmonary exacerbations tend to further impair glucose tolerance and may trigger CFRD because inflammatory cytokines and catecholamines promote insulin resistance. (See 'Differences from type 1 diabetes mellitus' above.)

●CFRD is associated with clinically important declines in pulmonary function and nutritional status, and with increased mortality. These effects can be attenuated or reversed by treatment with insulin. (See 'Clinical consequences of CFRD' above.)

●All patients with CF should undergo annual screening for CFRD, beginning by 10 years of age. Screening should be performed with an oral glucose tolerance test (OGTT), during a period of baseline health. Measurements of hemoglobin A1c (HbA1c) should not be used for screening, because this test has low sensitivity for CFRD. In addition to annual screening described above, monitoring of blood glucose is suggested during acute pulmonary exacerbations and for patients on enteral tube feedings. Women with CF who are pregnant or planning pregnancy warrant increased vigilance and screening for CFRD. (See 'Screening' above.)

●The OGTT is interpreted as follows (table 2) (see 'Interpretation of the oral glucose tolerance test' above):

•Indeterminate glycemia (INDET) – One-hour plasma glucose ≥200 mg/dL (11.1 mmol/L) and two-hour plasma glucose <140 mg/dL

•Impaired fasting glucose (IFG) – Fasting plasma glucose 100 to 125 mg/dL (5.6 to 7 mmol/L)

•Impaired glucose tolerance (IGT) – Two-hour plasma glucose 140 to 200 mg/dL (7.8 to 11.1 mmol/L)

•CFRD – Two-hour plasma glucose ≥200 mg/dL (11.1 mmol/L), with or without fasting hyperglycemia (FH; where FH is plasma glucose >126 mg/dL [7 mmol/L])

●We recommend treatment with insulin for all patients with CFRD rather than dietary management or no treatment (Grade 1B). We also suggest insulin treatment for selected patients with milder dysglycemia (eg, INDET or IGT) if this appears to be causing polyuria, polydipsia, nutritional decline, or deterioration in lung function (Grade 2C).

Although insulin treatment has important benefits, it also adds substantially to the burden of CF treatment. The treatment burden is an important factor when deciding whether to start treatment for patients with INDET or IGT and when choosing among insulin regimens (see 'When to intervene' above). A role for oral hypoglycemic agents in the treatment of CFRD has not been established. (See 'Oral hypoglycemic agents' above.)

●For CFRD, the insulin dose, regimen, and glycemic targets differ from those used for type 1 and type 2 diabetes. General principles for CFRD management are to give as much insulin as can be safely tolerated to eliminate the catabolic effects of CFRD, maintain HbA1c as low as possible (ideally <5.5 percent), and minimize postprandial hyperglycemia while avoiding episodes of hypoglycemia. The insulin regimen varies depending on the patient's characteristics and response (table 3A). (See 'Suggested protocol' above and 'Glycemic targets and adjustment of insulin' above.)

●Monitoring during insulin therapy includes self-monitoring of blood glucose (SMBG) via fingerstick at least three times daily (or continuous glucose monitoring [CGM]) and laboratory measurement of HbA1c every three months, with adjustment of the insulin regimen to reach targets (table 5). In addition, patients with CFRD should undergo annual screening for microvascular complications, beginning five years after diagnosis of CFRD. (See 'Glycemic targets and adjustment of insulin' above and 'Monitoring for complications' above.)