Reviews/Commentaries/ADA Statements COMMENTARY

Recommendations for Management of During Ramadan Update 2010

1 8 MONIRA AL-AROUJ, MD MAHMOUD ASHRAF IBRAHIM, MD less, many patients with diabetes insist on 2 9 SAMIR ASSAAD-KHALIL, MD, PHD DAVID KENDALL, MD fasting during Ramadan, thereby creating 3 10 JOHN BUSE, MD, PHD SUHAIL KISHAWI, MD 4 11 a medical challenge for themselves and IBTIHAL FAHDIL, MD, PHD ABDULRAZZAQ AL-MADANI, MD 5 1 their health care providers. It is increas- MOHAMED FAHMY, MD, PHD ABDULLAH BEN NAKHI, MD 6 12 ingly important that medical profession- SHERIF HAFEZ, MD, FACP KHALED TAYEB, MD 7 13 als be aware of potential risks associated MOHAMED HASSANEIN, FRCP ABRAHAM THOMAS, MD with fasting during Ramadan and with ap- proaches to mitigate those risks. These is- sues are rapidly becoming global issues, ince our last publication about dia- adult Muslims. The high global preva- not only in Indonesia, Pakistan, and the betes and fasting during Ramadan lence of type 2 diabetes—6.6% among Middle East, but also in North America, S (1), we have received many inquires adults age 20–79 years (5)—coupled Europe, and Oceania. and comments concerning important is- with the results of the population-based Although recommendations for man- sues that were not discussed in the previ- Epidemiology of Diabetes and Ramadan agement of diabetes in patients who elect ous document, including the voluntary 1422/2001 (EPIDIAR) study, which dem- to fast during Ramadan were proposed in 1- to 2-day fasts per week that many Mus- onstrated among 12,243 people with di- 1995 at a conference in Casablanca (7), lims practice throughout the year, as well abetes from 13 Islamic countries that our previous document was prompted by ϳ as the effect of prolonged fasting (more 43% of patients with type 1 diabetes the EPIDIAR study (6). The purpose of ϳ than 18 h a day) in regions far from the and 79% of patients with type 2 diabe- this review is to evaluate new data that equator during Ramadan when it occurs tes fast during Ramadan (6), lead to the has emerged since the publication of in summer—a phenomenon expected to estimate that worldwide more than 50 the 2005 article and to refine our affect millions worldwide for the next million people with diabetes fast during recommendations. 10–15 years. Since 2005, there have been Ramadan. In this revised document, we con- substantial additions to the literature, in- Ramadan is a lunar-based month, and tinue to avoid use of the terms “indica- cluding two studies examining the effect its duration varies between 29 and 30 tions” or “contraindications” for fasting of structured education and support for days. Muslims who fast during Ramadan because fasting is a spiritual issue for safe fasting, both of which had promising must abstain from eating, drinking, use of which patients make their own decision results (2,3). In addition, new medica- oral medications, and smoking from pre- after receiving appropriate advice from tions, such as the incretin-based thera- dawn to after sunset; however, there are religious teachings and from health care pies, have been introduced with less risk no restrictions on food or fluid intake be- providers. However, we emphasize that for hypoglycemia. tween sunset and dawn. Most people con- fasting, especially among patients with According to a 2009 demographic sume two meals per day during this type 1 diabetes with poor glycemic con- study, Islam has 1.57 billion adherents, month, one after sunset and the other be- trol, is associated with multiple risks. making up 23% of the world population fore dawn. ϳ of 6.8 billion, and is growing by 3% per Fasting is not meant to create exces- SUMMARY OF MAJOR year (4). Fasting during Ramadan, a holy sive hardship on the Muslim individual CHANGES AND UPDATES — The month of Islam, is a duty for all healthy according to religious tenets. Neverthe- current report: ●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●● ● From the 1Dasman Diabetes Institute, Dasman, Kuwait; the 2Department of Internal Medicine, Unit of Addresses the voluntary 1- to 2-day Diabetes & Metabolism, Alexandria Faculty of Medicine, Alexandria, Egypt; the 3Department of Medicine, fasts per week that many Muslims prac- University of North Carolina School of Medicine, Chapel Hill, North Carolina; the 4Eastern Mediterranean tice throughout the year 5 Office of the World Health Organization, Cairo, Egypt; the Department of Internal Medicine and Endo- ● Discusses the effect of prolonged fast- crinology, Ain Shams University, Cairo, Egypt; the 6Department of Internal Medicine and Diabetes, Cairo University, Cairo, Egypt; 7Consultant Diabetes and Endocrinology, North Wales, U.K.; the 8EDC, Center ing (more than 18 h a day) in regions far for Diabetes Education, McDonough, Georgia; the 9American Diabetes Association, Alexandria, Virginia; from the equator during Ramadan the 10Ministry of Health, Palestinian National Authority, Shifa Hospital, Gaza, Palestine; 11Dubai Hospital, when it occurs in summer (a phenom- Dubai, United Arab Emirates; 12Al-Nour Hospital, Mekkah, Saudi Arabia; and the 13Division of Endo- enon expected to affect millions of peo- crinology, Diabetes, and Bone and Mineral Disorders, Department of Medicine, Henry Ford Hospital, ple world-wide for the next 10–15 Detroit, Michigan. Corresponding author: Mahmoud Ibrahim, [email protected]. years) The document and specific recommendations were developed in collaboration with members of the Amer- ● Reviews additional and novel literature, ican Diabetes Association (ADA). This report represents the collective analysis, evaluation, and opinion of including studies examining the effect the authors at the time of publication and does not represent the official position of ADA. of structured education and support for DOI: 10.2337/dc10-0896 © 2010 by the American Diabetes Association. Readers may use this article as long as the work is properly safe fasting ● cited, the use is educational and not for profit, and the work is not altered. See http://creativecommons. Provides additional information in- org/licenses/by-nc-nd/3.0/ for details. cluding the clinical use of new medica-

care.diabetesjournals.org DIABETES CARE, VOLUME 33, NUMBER 8, AUGUST 2010 1895 Management of diabetes during Ramadan

tions with an emphasis on those with In individuals without diabetes, the Table 1—Major risks associated with fasting lesser risk for hypoglycemia, such as in- processes described above are regulated in patients with diabetes cretin-based therapies by a delicate balance between circulating ● Hypoglycemia Addresses safety information and the levels of and counterregulatory Hyperglycemia use and limitations of existing medica- hormones that help maintain glucose Diabetic ketoacidosis tions such as concentrations in the physiological range. Dehydration and thrombosis ● Addresses the growing global scope of In patients with diabetes, however, glu- the challenge of diabetes and fasting cose homeostasis is perturbed by the un- during Ramadan—more than 50 mil- derlying pathophysiology and often by RISKS ASSOCIATED WITH lion people with diabetes will fast dur- pharmacological agents designed to en- FASTING IN PATIENTS ing Ramadan in 2010 hance or supplement insulin secretion. In WITH DIABETES — Fasting during patients with type 1 diabetes, glucagon Ramadan has been uniformly discour- PATHOPHYSIOLOGY OF secretion may fail to increase appropri- aged by the medical profession for pa- FASTING — Insulin secretion, which ately in response to hypoglycemia. Epi- tients with diabetes. In keeping with this, a large epidemiological study conducted promotes the storage of glucose in liver nephrine secretion is also defective in in 13 Islamic countries on 12,243 diabetic and muscle as glycogen, is stimulated by some patients with type 1 diabetes be- individuals who fasted during Ramadan feeding in healthy individuals. During cause of a combination of autonomic neu- showed a high rate of acute complications fasting, circulating glucose levels tend to ropathy and defects associated with fall, leading to decreased secretion of in- (6). However, a few studies on this topic recurrent hypoglycemia (8). In patients using relatively small groups of patients sulin. Concurrently, levels of glucagon with severe insulin deficiency, a pro- and catecholamines rise, stimulating the suggest that complication rates may not longed fast in the absence of adequate in- breakdown of glycogen, and at the same be significantly increased (14–18). Some sulin can lead to excessive glycogen time gluconeogenesis is augmented (8). of the major potential complications asso- As fasting becomes protracted for more breakdown and increased gluconeogene- ciated with fasting in patients with diabe- than several hours, glycogen stores be- sis and ketogenesis, leading to hypergly- tes are outlined in Table 1. come depleted, and the low levels of cir- cemia and ketoacidosis. Patients with culating insulin allow increased fatty acid type 2 diabetes may suffer similar pertur- Hypoglycemia release from adipocytes. Oxidation of bations in response to a prolonged fast; Decreased food intake is a well-known fatty acids generates ketones that can be however, ketoacidosis is uncommon, and risk factor for the development of hypo- used as fuel by skeletal and cardiac mus- the severity of hyperglycemia depends on glycemia (19). It has been estimated that cle, liver, kidney, and adipose tissue, thus the extent of insulin resistance and/or hypoglycemia accounts for 2–4% of mor- sparing glucose for continued utilization deficiency. tality in patients with type 1 diabetes (20). by brain and erythrocytes. In a recent study, normal volunteers There are no reliable estimates concern- The transition from the fed state were subjected to intermittent 20-h fasts ing the contribution of hypoglycemia to through brief fasting and into prolonged every 2nd day for 15 days while maintain- mortality in type 2 diabetes; however, it is starvation is mediated by a series of com- ing body weight. Plasma free fatty acid felt that hypoglycemia is an infrequent plex metabolic, hormonal, and glu- and ␤-hydroxybutyrate concentrations cause of death in this group of patients. coregulatory mechanisms. Felig (9) increased after 20 h of fasting, confirming Rates of hypoglycemia are several-fold conveniently divided the transition from a that the subjects were fasting. Insulin- lower in patients with type 2 compared fed to a fasted state into three stages: mediated whole-body glucose rates in- with type 1 diabetes (6), and rates are creased and insulin-induced inhibition of even lower in patients with type 2 diabe- 1) the postabsorptive phase, 6–24 h after lipolysis in adipose tissue was more tes treated with oral agents (21). beginning fasting prominent after than before the interven- The effect of fasting during Ramadan tion. After the 20-h fasting periods, on rates of hypoglycemia in patients with 2) the gluconeogenic phase, from 2–10 diabetes is not known with certainty. The plasma adiponectin was increased com- days of fasting largest dataset is the recent EPIDIAR pared with the basal levels before and af- 3) the protein conservation phase, be- study (6), which showed that fasting dur- ter the intervention. This experiment was yond 10 days of fasting. ing Ramadan increased the risk of severe the first to show in humans that intermit- hypoglycemia (defined as hospitalization Although most religious fasts seldom ex- tent fasting increases insulin-mediated due to hypoglycemia) some 4.7-fold in ceed 24 h, the variability of the duration glucose uptake rates, compatible with patients with type 1 diabetes (from 3 to 14 Ϫ Ϫ of every phase may lead to different phys- the thrifty gene concept (12). Limited events 100 people 1 month 1) and iological responses to fasting. This vari- human data suggests higher HDL cho- 7.5-fold in patients with type 2 diabetes Ϫ ability may explain the feasibility of lesterol and lower triacylglycerol con- (from 0.4 to 3 events 100 people 1 Ϫ prolonged fast even in subjects with type centrations but no effect on blood month 1). The incidence of severe hypo- 1 diabetes in some studies (11). After an pressure from fasting. In terms of cancer glycemia was probably underestimated in overnight fast, the average rate of glucose risk, there is no human evidence to date this study because events requiring assis- utilization by a healthy human is ϳ7 g per of the effects of fasting. However, ani- tance from a third party without the need hour. By extrapolation, the 70–80 g of mal studies found decreases in lym- for hospitalization were not included. Al- glycogen present in the liver can provide phoma incidence, longer survival after though the average A1C in these patients glucose to the brain and peripheral tissues tumor inoculation, and lower rates of at the beginning of Ramadan was not for about 12 h (10). proliferation of several cell types (13). given, it is unlikely that the patients in this

1896 DIABETES CARE, VOLUME 33, NUMBER 8, AUGUST 2010 care.diabetesjournals.org Al-Arouj and Associates study had good glycemic control. Severe osmotic diuresis, further contributing Table 2— Categories of risk in patients with hypoglycemia was more frequent in pa- to volume and electrolyte depletion. type 1 or type 2 diabetes who fast during tients in whom the dosage of oral hypo- Orthostatic hypotension may develop, Ramadan glycemic agents or insulin were changed especially in patients with preexisting Very high risk and in those who reported a significant autonomic neuropathy. Syncope, falls, Severe hypoglycemia within the 3 months change in their lifestyle (6). injuries, and bone fractures may result prior to Ramadan from hypovolemia and the associated hy- A history of recurrent hypoglycemia Hyperglycemia potension. In addition, contraction of the Hypoglycemia unawareness Long-term morbidity and mortality stud- intravascular space can further exacerbate Sustained poor glycemic control ies in people with diabetes, such as the the hypercoagulable state that is well Ketoacidosis within the 3 months prior to Diabetes Control and Complications Trial demonstrated in diabetes (23). Increased Ramadan (DCCT) and the UK Prospective Diabetes blood viscosity secondary to dehydration Type 1 diabetes Study (UKPDS), demonstrated the link may enhance the risk of thrombosis and Acute illness among hyperglycemia, microvascular stroke (24). A report from Saudi Arabia Hyperosmolar hyperglycemic coma within complications, and possibly macrovascu- suggested an increased incidence of reti- the previous 3 months lar complications (19,22). However, nal vein occlusion in patients who fasted Performing intense physical labor there is no information linking repeated during Ramadan (25). However, hospital- Pregnancy yearly episodes of short-term hyperglyce- izations due to coronary events or stroke Chronic dialysis mia (e.g., 4-week duration) and diabetes- were not increased during Ramadan (26). High risk related complications. Control of There are no data concerning the effect of Moderate hyperglycemia (average blood glycemia in patients with diabetes who fasting on mortality in patients with or glucose 150–300 mg/dl or A1C fasted during Ramadan has been reported without diabetes. 7.5–9.0%) to deteriorate, improve, or show no Renal insufficiency MANAGEMENT change (21–25). The extensive EPIDIAR — It is worth re- Advanced macrovascular complications study showed a fivefold increase in the emphasizing that fasting for patients with Living alone and treated with insulin or incidence of severe hyperglycemia (re- diabetes represents an important personal quiring hospitalization) during Ramadan decision that should be made in light of Patients with comorbid conditions that in patients with type 2 diabetes (from 1 to guidelines for religious exemptions and Ϫ1 Ϫ1 present additional risk factors 5 events 100 people month ) and after careful consideration of the associ- Old age with ill health an approximate threefold increase in the ated risks in consultation with health care Treatment with drugs that may affect incidence of severe hyperglycemia with or providers. Most often, the medical recom- mentation without ketoacidosis in patients with type mendation will be to not undertake fast- Moderate risk 1 diabetes (from 5 to 17 events 100 peo- ing. However, patients who insist on Ϫ1 Ϫ1 Well-controlled diabetes treated with ple month ) (6). Hyperglycemia fasting need to be aware of the associated short-acting insulin secretagogues may have been due to excessive reduction risks of the fasting experience and tech- Low risk in dosages of medications to prevent hy- niques to decrease this risk. Patients may Well-controlled diabetes treated with poglycemia. Patients who reported an in- be at higher or lower risk for fasting- lifestyle therapy, , , crease in food and/or sugar intake had related complications depending on the thiazolidinediones, and/or incretin- significantly higher rates of severe hyper- number and extent of their risk factors. based therapies in otherwise healthy glycemia (6). Conditions associated with “very high,” patients “high,” “moderate,” and “low” risk for ad- Note: This classification is based largely on expert Diabetic ketoacidosis verse events in patients with type 1 or opinion and not on scientific data derived from clin- Patients with diabetes, especially those type 2 diabetes who decide to fast during ical studies. with type 1 diabetes, who fast during Ra- Ramadan are listed in Table 2. madan, are at increased risk for develop- ment of diabetic ketoacidosis, particularly General considerations pared with other times of the year. Most if their diabetes is poorly controlled be- Several important issues deserve special health problems are likely to arise from fore Ramadan (6). In addition, the risk for attention. inappropriate diet or as a consequence of diabetic ketoacidosis may be further in- Individualization. Perhaps the most over-eating and insufficient sleep. There- creased due to excessive reduction of in- crucial issue is the realization that care fore, the diet during Ramadan for people sulin dosages based on the assumption must be highly individualized and that with diabetes should not differ signifi- that food intake is reduced during the the management plan will differ for each cantly from a healthy and balanced diet. month. specific patient. The nutritional advice should be tailored Frequent monitoring of glycemia. It is to their special needs and medical prob- Dehydration and thrombosis essential that patients have the means to lems. It should aim at maintaining a con- Limitation of fluid intake during the fast, monitor their blood glucose levels multi- stant body mass. In most studies, 50– especially if prolonged, is a cause of dehy- ple times daily. This is especially critical 60% of individuals who fast maintain dration. The dehydration may become se- in patients with type 1 diabetes and in their body weight during the month, vere as a result of excessive perspiration in patients with type 2 diabetes who require while 20–25% either gain or lose weight hot and humid climates and among indi- insulin. (6). The common practice of ingesting viduals who perform hard physical labor. Nutrition. During Ramadan there is a large amounts of foods rich in carbohy- In addition, hyperglycemia produces an major change in the dietary pattern com- drates and fats, especially at the sunset care.diabetesjournals.org DIABETES CARE, VOLUME 33, NUMBER 8, AUGUST 2010 1897 Management of diabetes during Ramadan meal, should be avoided. Because of the glycemia during Ramadan as a reflection glycemia. The educational program delay in digestion and absorption, inges- of social habits encountered during the should include advice on the timing and tion of foods containing “complex” carbo- month. intensity of physical activity during fast- hydrates (slow digesting foods) may be ing. Certainly, it is important that use of advisable at the predawn meal, which Ramadan-focused structured diabetes-related medications and their should be eaten as late as possible before diabetes education potential risk during fasting are also the start of the daily fast. It is also recom- The role of structured education for pa- discussed. mended that fluid intake be increased tients is well established in the manage- A well-trained health care profes- during nonfasting hours. ment of diabetes. This should be sional should be able to deliver all these Exercise. Normal levels of physical ac- extended to Ramadan-focused diabetes components to people with diabetes ei- tivity may be maintained. However, ex- education. Many Muslims with diabetes ther individually or in a group session at cessive physical activity may lead to a are very passionate about fasting during diabetes centers, primary health care cen- higher risk of hypoglycemia and should Ramadan. This passion is a golden oppor- ters, local mosques, and/or community be avoided, particularly during the few tunity to empower people with diabetes centers. The ability to deliver this educa- hours before the sunset meal. Quite com- for better management of their diabetes, tional program in a simple, structured monly, multiple prayers are performed af- not only during Ramadan but also method and in the patients’ own language ter the sunset meal; this generally involves throughout the year. However, many is a distinct advantage, especially in mul- repeated cycles of rising, kneeling, and health care professionals find that they are tilingual communities. Certainly, many of bowing and should be considered a part unable to give the appropriate medical the components of the program will help of the daily exercise program. In some advice due to lack of knowledge about the empower people with diabetes to take patients with poorly controlled type 1 optimum management of diabetes while better care of themselves throughout the diabetes, exercise may lead to severe fasting. Indeed, often people with diabe- year. hyperglycemia. tes feel that there is lack of harmony be- An example of such an educational Breaking the fast. All patients should tween the medical and the religious program, which involves increasing understand that they must always and im- advice they receive. Hence, a Ramadan- awareness and training for health care mediately end their fast if hypoglycemia focused diabetes educational program professionals and the local community (blood glucose of Ͻ60 mg/dl [3.3 mmol/ should ideally include three components: about diabetes and Ramadan through a l]) occurs because their blood glucose structured education program, was con- may drop further if they delay treatment. 1) An awareness campaign aimed at peo- ducted in 2007 in the U.K. for a group of The fast should also be broken if blood ple with diabetes, health care profes- 111 people with type 2 diabetes (2). glucose reaches Ͻ70 mg/dl (3.9 mmol/l) sionals, the religious and community Though the group excluded people in the first few hours after the start of the leaders as well as the general public treated with insulin, secretagogues were fast, especially if insulin, 2) Ramadan-focused structured educa- used in over 90% of the people. Individ- drugs, or are taken at pre- tion for health care professionals ualized medication dose adjustments dawn. Finally, the fast should be broken if 3) Ramadan-focused structured educa- were suggested to all patients. At the end blood glucose exceeds 300 mg/dl (16.7 tion for people with diabetes. of Ramadan, when compared with a con- mmol/l). Patients should avoid fasting on trol group comprising those who did not “sick days.” Raising the general awareness of Ramadan participate in the Ramadan-focused dia- and diabetes should strengthen the har- betes education, those who received such Pre-Ramadan medical assessment mony between medical and religious ad- education exhibited a nearly 50% reduc- All patients with diabetes who wish to fast vice. This is of particular importance in tion in hypoglycemic event rates despite during Ramadan should prepare by un- non-Muslim countries where poor com- fasting, whereas those in the control dergoing a medical assessment and en- munication and understanding between group had experienced a fourfold in- gaging in a structured education program these communities is commonplace. crease in the rate of hypoglycemic events to undertake the fast as safely as possible. Greater understanding regarding the reli- from baseline during fasting. It is impor- This assessment should take place 1–2 gious context and perspective regarding tant to note that this occurred while gly- months before Ramadan. Specific atten- Ramadan and the act of fasting, as well as cemic control was maintained at the same tion should be devoted to patients’ overall the potential risks and the medical op- level for 12 months (2). Furthermore, the well-being and to the control of their gly- tions to achieve a safer outcome for those group that received structured education cemia, blood pressure, and lipids. Appro- who wish to fast, is critical for all parties. lost a small amount of weight compared priate blood studies should be ordered Health care professionals should be to weight gain in the control group (27). and evaluated. Specific medical advice trained to deliver a structured patient ed- must be provided to individual patients ucation program that includes a better Management of patients with type 1 concerning the potential risks they must understanding of fasting and diabetes, in- diabetes accept if they decide to fast. During this dividual risk quantification, and options Fasting at Ramadan carries a very high assessment, necessary changes in diet or to achieve safer fasting. This includes the risk for people with type 1 diabetes. This medication regimen should be made so importance of glucose monitoring during risk is particularly exacerbated in poorly that the patient initiates fasting while on a fasting and nonfasting hours, when to controlled patients and those with limited stable and effective program. This assess- stop the fast, meal planning to avoid hy- access to medical care, hypoglycemic un- ment should also extend to those who poglycemia and dehydration during pro- awareness, unstable glycemic control, or do not wish to fast because they often are longed fasting hours, and the appropriate recurrent hospitalizations. In addition, exposed to the risk of hypo- and hyper- meal choices to avoid postprandial hyper- the risk is also very high in patients who

1898 DIABETES CARE, VOLUME 33, NUMBER 8, AUGUST 2010 care.diabetesjournals.org Al-Arouj and Associates are unwilling or unable to monitor their pealing alternative strategy, but at a sub- Evaluated for Cardiovascular Outcomes blood glucose levels several times daily. It stantially greater expense. Compared and Regulation of Glycaemia in Diabetes is currently recommended that treatment with those who did not fast during Ra- (RECORD) study, which failed to demon- regimens aimed at intensive glycemia madan, patients with type 1 diabetes on strate either harm or benefit. Neverthe- management be used in patients with di- insulin pump therapy who fasted showed less, most perceive a relative advantage of abetes. The DCCT and its follow up, the a slight improvement in A1C (3). compared with Epidemiology for Diabetes Interven- vis-a`-vis lipid effects. A practical issue of tions and Complications (EDIC) study, Management of patients with type 2 significant importance with respect to the demonstrated that intensive glycemia diabetes utility of glitazones in periods of fasting management is protective against mi- Diet-controlled patients. In patients such as Ramadan is that these agents re- crovascular and perhaps macrovascular with type 2 diabetes who are well con- quire 2–4 weeks to exert substantial an- complications and that the benefits are trolled with lifestyle therapy alone, the tihyperglycemic effects. Therefore, these long lasting (19,28). Glycemic control at risk associated with fasting is quite low. agents cannot be quickly substituted for near-normal levels requires use of multi- However, there is still a potential risk for agents associated with hypoglycemia dur- ple daily insulin injections (three or more) occurrence of postprandial hyperglyce- ing periods of fasting (31). or use of continuous subcutaneous insu- mia after the predawn and sunset meals if Sulfonylureas. It has been suggested that lin infusion through pump therapy. Close patients overindulge in eating. Distribut- this group of drugs is unsuitable for use monitoring and frequent insulin dose ad- ing calories over two to three smaller during fasting because of the inherent risk justments in this setting are essential to meals during the nonfasting interval may of hypoglycemia. However, severe or fatal achieve optimal glycemic control and help prevent excessive postprandial hy- hypoglycemia is a relatively rare compli- avoid hypo- or hyperglycemia in patients perglycemia. Physical activity may be cation of sulfonylurea use. Nevertheless, with type 1 diabetes. modified in its intensity and timing, e.g., ϳ their use should be individualized with Some patients with type 1 diabetes 2 h after the sunset meal. caution. Use of is rela- prefer to fast at Ramadan, and most of Patients treated with oral agents. The tively contraindicated during Ramadan them change their insulin regimens im- choice of oral agents should be individu- because of the possibility of prolonged mediately before, during, and a few days alized. In general, agents that act by in- and unpredictable hypoglycemia. Simi- after this month. However, very few stud- creasing insulin sensitivity are associated larly, it has been suggested that glyburide ies have documented the safety and/or ef- with a significantly lower risk of hypogly- or may be associated with a ficacy of different insulin regimens in type cemia than compounds that act by in- higher risk of hypoglycemia than other 1 diabetic patients who fast during the creasing insulin secretion. second-generation sulfonylureas, specifi- month of Ramadan. The current under- Metformin. Patients treated with met- cally , , and standing is that the basal-bolus regimen is formin alone may safely fast because the (32,33). Finally, it should be noted that the preferred protocol of management. It possibility of severe hypoglycemia is min- the sulfonylureas glyburide (gliben- is thought to be safer, with fewer episodes imal. However, perhaps the timing of the of hyper- and hypoglycemia. A frequently doses should be modified to provide two- clamide) and gliclazide MR have played a used option is once- or twice-daily injec- thirds of the total daily dose with the sun- central role in the long-term outcome tions of intermediate or long-acting insu- set meal and the other third before the studies UKPDS and ADVANCE (Action in lin along with premeal rapid-acting predawn meal. Diabetes and Vascular Disease: Preterax insulin. It is unlikely that other regimens, Glitazones. The or gli- and Diamicron MR Controlled Evalua- including one or two injections of inter- tazone agents (pioglitazone and rosiglita- tion), both of which demonstrated micro- mediate-, long-acting, or premixed insu- zone) are not independently associated vascular benefits and at least trends lin, would provide adequate insulin with hypoglycemia, though they can am- toward improvements in cardiovascular therapy. A recent small study with insulin plify the hypoglycemic effects of sulfonyl- disease without evidence of excess mor- glargine suggests the relative safety and ureas, glinides, and insulin. However, tality (34). Additional studies on the use efficacy of this agent in 15 relatively well- they are associated with weight gain and of sulfonylureas in patients who fast dur- controlled patients with type 1 diabetes anecdotally can be associated with in- ing Ramadan are needed before strong who fasted for 18 h and experienced a creased appetite. The longstanding con- recommendations on their utility can be minimal decline in mean plasma glucose cerns regarding cardiovascular safety, made. Nevertheless, because of their from 125 to 93 mg/dl with only two epi- caused by the increased frequency of worldwide use and relatively low cost, sodes of mild hypoglycemia (29). An- heart failure, continue despite greater un- these agents may be used in Ramadan, other study in patients with type 1 derstanding that the mechanism of this though with caution. diabetes using , lispro, or adverse effect seems to be related to renal Short-acting insulin secretagogues. aspart instead of in com- tubular sodium and water reabsorption Members of this group ( and bination with intermediate-acting insulin and not to an intrinsic affect on cardiac ) are useful because of their injected twice a day led to improvement contractility. More recently, apprehen- short duration of action. They could be in postprandial glycemia and was associ- sion has emerged regarding reports of in- taken twice daily before the sunset and ated with fewer hypoglycemic events creased frequency of macular edema and predawn meals. One study in patients (30). Clinical studies with other types of of bone fractures, particularly in post- with type 2 diabetes who fasted showed insulin in multiple daily injection regi- menopausal women. The recent contro- that use of repaglinide was associated mens during fasting are limited. versy regarding the cardiovascular safety with less hypoglycemia compared with Continuous subcutaneous insulin of rosiglitazone seems to have been glibenclamide (35). Nateglinide has the infusion (pump) management is an ap- largely mitigated by the Rosiglitazone shortest duration of action and therefore care.diabetesjournals.org DIABETES CARE, VOLUME 33, NUMBER 8, AUGUST 2010 1899 Management of diabetes during Ramadan the lowest risk of severe fasting hypogly- dicious use of intermediate- or long- stopping insulin delivery from the pump. cemia among the secretagogues. acting insulin preparations plus a short- Such an advantage is not available to Incretin-based therapy. Therapies that acting insulin administered before meals. those treated with a conventional insulin affect the incretin system include gluca- Although hypoglycemia tends to be less injection in which insulin continues to be gon-like peptide-1 receptor agonists frequent, it is still a risk, especially in pa- released from the site of injection (GLP-1ras) and and tients who have required insulin therapy throughout its predetermined duration of dipeptidylpeptidase-4 inhibitors (DPP- for a number of years or in whom insulin action. Any excess insulin action can 4is) , , , deficiency predominates in the patho- only be counteracted by intake of and . These classes of agents physiology. Very elderly patients with carbohydrates. are not independently associated with hy- type 2 diabetes may be at especially high Fasting at Ramadan may be success- poglycemia, though they can increase the risk. fully accomplished in people with type 1 hypoglycemic effects of sulfonylureas, Using one injection of a long-acting diabetes if they are fully educated and fac- glinides, and insulin. Exenatide in partic- or intermediate-acting insulin can pro- ile with the use of the insulin pump and ular can be dosed before meals to mini- vide adequate coverage in some patients are otherwise metabolically stable and mize appetite and promote weight loss. as long as the dosage is appropriately in- free from any acute illnesses. Prior to Ra- With its short half-life of 2 h, it is not dividualized; however, most patients will madan, they should receive adequate associated with a substantial effect on fast- require rapid- or short-acting insulin ad- training and education, particularly with ing glucose. Liraglutide is dosed once a ministered in combination with the basal respect to self-management and insulin day, independent of meals, and is more insulin at meals, particularly at the dose adjustment. They should adjust effective in controlling fasting glycemia. evening meal, which typically contains a their infusion rates carefully according to Both require titration to effective doses larger caloric load. There is some evi- results of frequent home blood glucose over a period of 2–4 weeks and are asso- dence suggesting that use of a rapid- monitoring. Most will need to reduce ciated with mild to moderate nausea in acting instead of regular their basal infusion rate while increasing almost half of those exposed on at least human insulin before meals in patients the bolus doses to cover the predawn and one occasion, particularly as therapy is with type 2 diabetes who fast during Ra- sunset meals. initiated. DPP-4is are among the best tol- madan is associated with less hypoglyce- erated drugs for the treatment of diabetes. mia and smaller postprandial glucose Diabetic medication adjustment They are moderately less effective in A1C excursions (38,39). In a recent study, the during Ramadan lowering than GLP-1ras and, importantly use of premixed lispro with neutral pro- Illustrative examples and recommenda- vis-a`-vis treatment during Ramadan, do tamine lispro in a 50:50 ratio for the tions for adjusting therapy during Ra- not require titration. Many have touted evening meal and regular human insulin madan in patients with type 2 diabetes are their potential role as a substitute for sul- with NPH in a 30:70 ratio at the early shown in Table 3. fonylureas. However, there are no specific morning meal during Ramadan com- studies of these agents during periods of pared with regular human insulin at Pregnancy and fasting during fasting with respect to either tolerability 30:70 twice daily was associated with Ramadan or efficacy (36). moderate improvement in glycemic con- Pregnancy is a state of increased insulin ␣-Glucosidase inhibitors. Acarbose, trol and hypoglycemia (40). resistance and insulin secretion and of re- , and voglibose slow the absorp- Insulin pumps. An insulin pump pro- duced hepatic insulin extraction. Fasting tion of carbohydrates when taken with vides continuous insulin delivery over glucose concentrations are lower and the first bite of a meal. Because they are 24 h with basal infusion rates pro- postprandial glucose and insulin levels not associated with an independent risk grammed and individualized for each pa- are substantially higher in healthy preg- of hypoglycemia, particularly in the fast- tient. Patients self-administer boluses of nant women than in healthy women who ing state, they may be particularly useful insulin with meals or at times of hypergly- are not pregnant. Elevated blood glucose during Ramadan. However, they are only cemia, often with mathematical support and A1C levels in pregnancy are associ- modestly effective and exert little or no from the pump. The reliance on exclu- ated with increased risk for major congen- effect on fasting glucose, and therefore are sively rapid-acting or short-acting insulin ital malformations. Fasting during usually used in combination with other allows for flexibility over an extremely pregnancy would be expected to carry a agents to control fasting glucose. ␣-Glu- wide range of insulin doses with great high risk of morbidity and mortality to the cosidase inhibitors are associated with precision. However, frequent glucose fetus and mother, although controversy frequent mild to moderate gastro- monitoring is required because failure of exists (41). While pregnant Muslim intestinal effects, particularly flatulence. the pump or the infusion site can result in women are exempt from fasting during Using modest doses and slowly initiating severe deterioration in control over a few Ramadan, some with known diabetes therapy are reported to minimize the fre- hours. Theoretically, the combined risks (type 1, type 2, or gestational) insist on quency of these adverse effects (37). of hypoglycemia from prolonged daytime fasting. These women constitute a high- Patients treated with insulin. Prob- fasting and hyperglycemia from excessive risk group, and their management re- lems facing patients with type 2 diabetes nighttime eating can be better managed quires intensive care (42). who administer insulin are similar to by an insulin pump–based regimen than In general, women with pregesta- those with type 1 diabetes, except that the by multiple insulin dose–injection ther- tional or gestational diabetes are at very incidence of hypoglycemia is less. Again, apy. Hypoglycemia can be aborted, re- high risk and may be strongly advised not the aim is to maintain necessary levels of duced, prevented, and even more readily to fast during Ramadan. However, if they basal insulin to prevent fasting hypergly- treated in pump-treated patients by insist on fasting, special attention should cemia. An effective strategy would be ju- timely downward adjustments or even be given to their care. Pre-Ramadan eval-

1900 DIABETES CARE, VOLUME 33, NUMBER 8, AUGUST 2010 care.diabetesjournals.org Al-Arouj and Associates

Table 3—Recommended changes to treatment regimen in patients with type 2 diabetes who cose monitoring, and dosage and timing fast during Ramadan of medications. Newer pharmacological agents have Before Ramadan During Ramadan lesser hypoglycemic potential and may have specific advantages during Ra- Patients on diet and exercise Consider modifying the time and intensity of physical madan. Similarly, insulin pump therapy control activity; ensure adequate fluid intake may provide greater safety in the Ra- Patients on oral hypoglycemic Ensure adequate fluid intake madan setting. There are a few studies of agents these newer techniques in the Ramadan , metformin 500 mg, Metformin, 1,000 mg at the sunset meal, 500 mg at setting with encouraging results, but in three times daily the predawn meal general this challenging therapeutic situ- TZDs, AGIs, or incretin-based No change needed ation has not been adequately addressed therapies in clinical trials. Sulfonylureas once a day Dose should be given before the sunset meal; adjust the dose based on the glycemic control and the risk of hypoglycemia Acknowledgments— The Egyptian Diabetes Sulfonylureas twice a day Use half the usual morning dose at the predawn meal Center, with support from Les Laboratoires and the usual dose at sunset meal Servier, made this work possible. J.B.’s institution has received funds from Patients on insulin Ensure adequate fluid intake multiple pharmaceutical and device compa- Premixed or intermediate-acting Consider changing to long-acting or intermediate nies for his services as an investigator and con- insulin twice daily insulin in the evening and short or rapid-acting sultant, he currently has grants pending, and insulin with meals; take usual dose at sunset meal his travel accommodation expenses have been and half usual dose at predawn meal covered by the American Diabetes Associa- AGI, ␣-glucosidase inhibitor; TZD, thiazolidinedione. Note: The recommendations given in this table are for tion. A.T. is on the speaker’s bureau for Novo illustrative purposes and are largely based on expert clinical opinion and not on scientific data derived from Nordisk and Takeda and has received research clinical studies. The recommendations must be adjusted for each specific patient. Adapted from Akbani et al. grants from Roche, Eli Lilly and Company, (43). and Pharmaceuticals. No other poten- tial conflicts of interest relevant to this article were reported. uation of their medical condition is es- drates and saturated fats is increased dur- The authors are thankful for all of the com- sential. This includes preconception care ing Ramadan. Appropriate counseling ments received regarding their previous re- with emphasis on achieving near-normal should be given to avoid this practice, and port, each of which were considered during blood glucose and A1C values, counsel- agents that were previously prescribed for the preparation of this update. ing about maternal and fetal complica- the management of elevated cholesterol tions associated with poor glycemic and triglycerides should be continued. control, and education focused on self- References management skills. Ideally, patients CONCLUSIONS — Fasting during 1. Al-Arouj M, Bouguerra R, Buse J, Hafez S, should be managed in high-risk clinics Ramadan for patients with diabetes car- Hassanein M, Ibrahim MA, Ismail-Beigi F, El-Kebbi I, Khatib O, Kishawi S, Al-Madani staffed by an obstetrician, diabetologists, ries a risk of an assortment of complica- A, Mishal AA, Al-Maskari M, Nakhi AB, Al- a nutritionist, and diabetes nurse educa- tions. In general, patients with type 1 Rubean K. Recommendations for manage- tors. The management of pregnant pa- diabetes are at very high risk of life- ment of diabetes during Ramadan. Diabetes tients during Ramadan is based on an threatening complications. Patients with Care 2005;28:2305–2311 appropriate diet and intensive insulin type 1 diabetes who have a history of re- 2. Bravis V, Hui E, Salih S, Mehar S, Hassa- therapy. The issues discussed above con- current hypoglycemia or hypoglycemia nein M, Devendra D. Ramadan Education cerning the management of type 1 and unawareness or who are poorly con- and Awareness in Diabetes (READ) pro- type 2 diabetes also apply to this group, trolled are at very high risk for developing gramme for Muslims with type 2 diabetes with the exception that more frequent severe hypoglycemia. On the other hand, who fast during Ramadan. Diabetic Med- monitoring and insulin dose adjustment an excessive reduction in the insulin dos- icine 2010;27:327–331 3. Benbarka MM, Khalil AB, Beshyah SA, is necessary. age in these patients (to prevent hypogly- Marjei S, Awad SA. Insulin pump therapy cemia) may place them at risk for in Moslem patients with type 1 diabetes Management of hypertension and hyperglycemia and diabetic ketoacidosis. during Ramadan fasting: an observational dyslipidemia Hypo- and hyperglycemia may also occur report. Diabetes Technol Ther 2010;12: Dehydration, volume depletion, and a in patients with type 2 diabetes, but is 287–290 tendency toward hypotension may occur generally less frequent and has less severe 4. Miller T, Ed. Mapping the Global Muslim with fasting during Ramadan, especially if consequences than in patients with type 1 Population: A Report on the Size and Distri- the fast is prolonged and is associated diabetes. A patient’s decision to fast bution of the World’s Muslim Population, with excessive perspiration. Hence, the should be made after ample discussion [Internet] c2009. Washington, DC, Pew Re- dosage and/or the type of antihyperten- with a health care provider concerning search Center. Available from http:// pewforum.org/newassets/images/reports/ sive medications may need to be adjusted the risks involved. Patients who insist on Muslimpopulation/Muslimpopulation. to prevent hypotension. Diuretics may fasting should undergo pre-Ramadan as- pdf. Accessed 8 October 2009 not be appropriate during Ramadan for sessment and receive appropriate edu- 5. International Diabetes Federation. Diabe- some patients. It is a common practice cation and instructions related to tes Atlas 2009. 4th ed. Brussels, Interna- that the intake of foods rich in carbohy- physical activity, meal planning, glu- tional Diabetes Federation, 2009 care.diabetesjournals.org DIABETES CARE, VOLUME 33, NUMBER 8, AUGUST 2010 1901 Management of diabetes during Ramadan

6. Salti I, Be´nard E, Detournay B, Bianchi- 20. Laing SP, Swerdlow AJ, Slater SD, Botha once-daily gliclazide MR and glimepiride in Biscay M, Le Brigand C, Voinet C, Jabbar JL, Burden AC, Waugh NR, Smith AW, type 2 diabetic patients. Eur J Clin Invest A, EPIDIAR study group. A population- Hill RD, Bingley PJ, Patterson CC, Qiao 2004;34:535–542 based study of diabetes and its character- Z, Keen H. The British Diabetic Associ- 33. Rendell M. The role of sulphonylureas in istics during the fasting month of ation Cohort Study, II: cause-specific the management of type 2 diabetes melli- Ramadan in 13 countries: results of the mortality in patients with insulin- tus. Drugs 2004;64:1339–1358 epidemiology of diabetes and Ramadan treated diabetes mellitus. Diabet Med 34. ADVANCE Collaborative Group, Patel A, 1422/2001 (EPIDIAR) study. Diabetes 1999;16:466–471 MacMahon S, Chalmers J, Neal B, Billot L, Care 2004;27:2306–2311 21. Miller CD, Phillips LS, Ziemer DC, Woodward M, Marre M, Cooper M, Glas- 7. International Meeting on Diabetes and Gallina DL, Cook CB, El-Kebbi IM. Hypo- ziou P, Grobbee D, Hamet P, Harrap S, Ramadan Recommendations. Edition of glycemia in patients with type 2 diabetes. Heller S, Liu L, Mancia G, Mogensen CE, the Hassan II Foundation for Scientific and Arch Int Med 2001;161:1653–1659 Pan C, Poulter N, Rodgers A, Williams B, Medical Research on Ramadan. Casablanca, 22. Intensive blood-glucose control with sul- Bompoint S, de Galan BE, Joshi R, Travert Morocco, FRSMR, 1995 phonylureas or insulin compared with F. Intensive blood glucose control and 8. Cryer PE, Davis SN, Shamoon H. Hypo- conventional treatment and risk of com- vascular outcomes in patients with type 2 glycemia in diabetes (Review). Diabetes plications in patients with type 2 diabetes diabetes. N Engl J Med 2008;358:2560– Care 2003;26:1902–1912 (UKPDS 33). UK Prospective Diabetes 2572 9. Felig P. Starvation. In Endocrinology. De- Study (UKPDS) Group. Lancet 1998;352: Groot LJ, Ed. New York, Grune & Strat- 837–853 35. Mafauzy M. Repaglinide versus gliben- ton, 1979, p. 1927–1940 23. Beckman JA, Creager MA, Libby P. Diabe- clamide treatment of type 2 diabetes dur- 10. Cahill GF Jr. Starvation in man. N Engl tes and atherosclerosis: epidemiology, ing Ramadan fasting. Diabetes Res Clin J Med 1970;282:668–675 pathophysiology, and management. JAMA Pract 2002;58:45–53 11. Reiter J, Wexler ID, Shehadeh N, Tzur A, 2002;287:2570–2581 36. Drucker DJ, Sherman SI, Gorelick FS, Zangen D. Type 1 diabetes and prolonged 24. Akhan G, Kutluhan S, Koyuncuoglu HR. Bergenstal RM, Sherwin RS, Buse JB. In- fasting. Diabet Med, 2007;24:436–439 Is there any change in stroke incidence cretin-based therapies for the treatment of 12. Halberg N, Henriksen M, So¨derhamn N, during Ramadan? Acta Neurol Scandin type 2 diabetes: evaluation of the risks Stallknecht B, Ploug T, Schjerling P, Dela 2000;101:259–261 andbenefits.DiabetesCare2010;33:428– F. Effect of intermittent fasting and 25. Alghadyan AA. Retinal vein occlusion in 433 refeeding on insulin action in healthy Saudi Arabia: possible role of dehydra- 37. Van de Laar FA, Lucassen PL, Akkermans men. J Appl Physiol 2005;99:2128–2136 tion. Ann Ophthalmol 1993;25:394–398 RP, Van de Lisdonk EH, Rutten GE, Van 13. Varady KA, Hellerstein MK. Alternate-day 26. Temizhan A, Do¨nderici O, Ouz D, Demir- Weel C. Alpha-glucosidase inhibitors for fasting and chronic disease prevention: a bas B. Is there any effect of Ramadan fast- type 2 diabetes mellitus. Cochrane Data- review of human and animal trials. Am J ing on acute coronary heart disease base Syst Rev 2005:CD003639 Clin Nutr 2007;86:7–13 events? Int J Cardiol 1999;70:149–153 38. Mattoo V, Milicevic Z, Malone JK, 14. Uysal AR, Erdog˘an MF, Sahin G, Kamel 27. Hassanein M, Bravis V, Hui E, Devendra D. Schwarzenhofer M, Ekangaki A, Levitt N, Erdog˘an G. Clinical and metabolic ef- Ramadan-focused education and awareness LK, Liong LH, Rais N, Tounsi H, Ramadan fects of fastingin 41 type 2 diabetic pa- in type 2 diabetes. Diabetologia 2009;52: Study Group. A comparison of insulin lis- tients during Ramadan (Letter). Diabetes 367–368 pro Mix25 and human insulin 30/70 in Care 1998;21:2033–2034 28. Writing Team for the Diabetes Control the treatment of type 2 diabetes during 15. Laajam MA. Ramadan fasting and non- and Complications Trial/Epidemiology of Ramadan. Diabetes Res Clin Pract 2003; insulin-dependent diabetes: effect on Diabetes Interventions and Complica- 59:137–143 metabolic control. East Afr Med J 1990; tions Research Group. Sustained effect of 39. Akram J, De Verga V. 67:732–736 intensive treatment of type 1 diabetes (Lys(B28), Pro(B29) in the treatment of 16. Mafauzy M, Mohammed WB, Anum MY, mellitus on development and progression diabetes during the fasting month of Ra- Zulkifli A, Ruhani AH. A study of the fast- of diabetic nephropathy: the Epidemiol- madan. Ramadan Study Group. Diabet ing diabetic patient during the month of ogy of Diabetes Interventions and Com- Med 1999;16:861–866 Ramadan. Med J Malaysia 1990;45: plications (EDIC) study. JAMA 2003;290: 40. Hui E, Bravis V, Salih S, Hassanein M, 14–17 2159–2167 Devendra D. Comparison of Humalog 17. Belkhadir J, el Ghomari H, Klo¨cker N, 29. Mucha GT, Merkel S, Thomas W, Bantle Mix 50 with human insulin Mix 30 in type Mikou A, Nasciri M, Sabri M. Muslims JP. Fasting and in individ- with non-insulin dependent diabetes fast- uals with type 1 diabetes. Diabetes Care 2 diabetes patients during Ramadan. Int ing during Ramadan: treatment with glib- 2004;27:1209–1210 J Clin Pract. 10 March 2010 [Epub ahead enclamide. BMJ 1993;307:292–295 30. Kadiri A, Al-Nakhi A, El-Ghazali S, Jabbar of print] 18. Katibi IA, Akande AA, Bojuwoye BJ, A, Al Arouj M, Akram J, Wyatt J, Assem A, 41. Malhotra A, Scott PH, Scott J, Gee H, Okesina AB. Blood sugar control among Ristic S. Treatment of type 1 diabetes with Wharton BA. Metabolic changes in Asian fasting Muslims with type 2 diabetes mel- insulin lispro during Ramadan. Diabete Muslim pregnant mothers observing the litus in Ilorin. Niger J Med 2001;10:132– Metab 2001;27:482–486 Ramadan fast in Britain. Br J Nutr 1989; 134 31. Retnakaran R, Zinman B. Thiazolidinedio- 61:663–672 19. The effect of intensive treatment of diabetes nes and clinical outcomes in type 2 diabetes. 42. Azizi F. Research in Islamic fasting and on the development and progression of Lancet 2009;373:2088–2090 health. Ann Saudi Med 2002;22:186–191 long-term complications in insulin-depen- 32. Schernthaner G, Grimaldi A, Di Mario U, 43. Akbani MF, Saleem M, Gadit WU, Ahmed dent diabetes mellitus. The Diabetes Con- Drzewoski J, Kempler P, Kvapil M, Novi- M, Basit A, Malik RA. Fasting and feasting trol and Complications Trial Research als A, Rottiers R, Rutten GE, Shaw KM. safely during Ramadan in the diabetic pa- Group. N Engl J Med 1993;329:977–986 GUIDE study: double-blind comparison of tient. Pract Diab Int 2005;22:100–104

1902 DIABETES CARE, VOLUME 33, NUMBER 8, AUGUST 2010 care.diabetesjournals.org