- Email: [email protected]
Catamenial Diabetic Ketoacidosis and Catamenial Hyperglycemia: Case Report and Review of the Literature
Catamenial Diabetic Ketoacidosis and Catamenial Hyperglycemia: Case Report and Review of the Literature FERNANDO OVALLE, MD; TOM B. VAUGHAN III, MD; JENNIFER E. SOHN, MD; BARBARA GOWER, PHD
ABSTRACT: Background: Diabetic ketoacidosis (DKA), an acute and potentially life-threatening complication of diabetes mellitus, is frequently the result of a precipitating illness or nonadherence to treatment; however, despite a thorough history, physical, radiologic, and laboratory examination, a specific cause or precipitating event cannot be identified in a significant number of cases. Aims: To report 2 cases of recurrent DKA, and hyperglycemic crisis, associated to the menstrual cycle, and provide a review of the medical literature. Methods: The clinic and hospital medical records of 2 subjects with type 1 diabetes mellitus (T1DM) and unexplained DKA are reviewed. An electronic MEDLINE search of relevant medical literature published from 1965 to 2007 was performed; additionally, the reference lists of the identified articles and other sources, such as textbook chapters and meeting abstracts, were reviewed for related publications. Results: To date there have been 7 reported cases in the literature demonstrating an associ-
ation between menstruation and DKA. The 2 new cases presented here highlight the potentially significant changes in glucose metabolism that may occasionally be observed during the late luteal and decidual phases of the menstrual cycle; moreover, these cases underscore the importance of a thorough medical and gynecologic history when evaluating a patient with DKA. Conclusion: Through unclear mechanisms, some women with diabetes mellitus demonstrate significant changes in glucose control around the time of their menses, including DKA. Accordingly, we propose that the terms catamenial DKA and catamenial hyperglycemia be used to refer to these disorders and that catamenial DKA be included in the differential diagnosis list of causes or precipitating events that can lead to DKA. KEY INDEXING TERMS: Catamenial; Menstruation; Diabetes; Ketoacidosis; Hyperglycemia; Women. [Am J Med Sci 2008; 335(4):298–303.]
D
ating and managing patients with DKA. The most common and frequently highlighted causes of DKA are infections, new-onset diabetes, cerebrovascular accidents, alcohol abuse, pancreatitis, myocardial infarction, trauma, drugs, eating disorders, poor compliance and treatment errors; however, it is well recognized that in a significant number of cases, despite a careful search, no precipitating event is identified. It is also well known that some patients have a tendency to develop repeated episodes of DKA and, in frustration, physicians label them as “brittle diabetics.” Unfortunately, some of these patients may be erroneously accused of poor compliance or even malingering. The influences of the menstrual cycle and female sex hormones on insulin resistance, glucose metabolism, and glycemic control have generated significant controversy over the years. In our clinical experience, we have frequently observed a cyclic deterioration of glycemic control that closely follows the menstrual cycle in some, but not all, women with T1DM. Here, we describe 2 cases of cyclic DKA and cyclic hyperglycemic crisis, which developed during the
iabetic Ketoacidosis (DKA) is an acute and potentially life-threatening complication of diabetes mellitus with reported mortality rates between 2% and 5%.1– 4 The annual incidence of DKA has been reported as high as 1.5 episodes per 100 patients with diabetes per year, and the rates appear to be higher in blacks, lower socioeconomic groups, and, interestingly, in adolescent females.5,6 Appropriately, most textbook chapters and review articles on DKA emphasize the need to rapidly identify the cause or precipitating factor(s) when evalu-
From the Department of Medicine (FO, TBV, JES), Division of Endocrinology, Diabetes and Metabolism; and Department of Nutrition Sciences (BG), The University of Alabama at Birmingham School of Medicine, Birmingham, Alabama. Submitted February 18, 2007; accepted in revised form September 11, 2007. Correspondence: Fernando Ovalle, MD, FACE, Department of Medicine, Division of Endocrinology, Diabetes, and Metabolism, University of Alabama at Birmingham School of Medicine, 510 Twentieth Street South/FOT 702, Birmingham, AL 35294 (E-mail: [email protected]).
298
April 2008 Volume 335 Number 4
Ovalle et al
premenstrual period without any other clear precipitating event. Case Reports Case 1 A 32-year-old white woman with T1DM was referred to our diabetes clinic for evaluation of recurrent episodes of DKA and difficult glycemic control. Six years earlier at the age of 26, she had been diagnosed with T1DM when she reported to her family physician with symptoms from severe hyperglycemia and DKA. She had been managed with insulin since then and her glycemic control had been poor, with HgA1c’s usually around 10%; nonetheless, she had not yet developed any chronic microvascular complications from her diabetes. Just over 1 year prior, she had been started on continuous subcutaneous insulin infusion (CSII) therapy in an attempt to improve her glycemic control. She gave a history of several episodes of DKA since her diagnosis, in average once or twice a year, without any clear or previously identified precipitating events; however, over the last year, she had been admitted to her community hospital 7 times, almost on a monthly basis, with DKA. During each of these admissions to the intensive care unit, her treating physicians were unable to identify a precipitating illness or event despite seemingly thorough physical and laboratory workups, which included among other tests: blood counts and chemical analyses, thyroid function tests, blood gases, urinalysis, chest and abdominal x-ray examinations, electrocardiograms, blood and urine cultures, and even a lumbar spine puncture for cerebrospinal fluid analysis. Remarkably, the patient had noticed that all of these episodes of DKA had developed 2 to 3 days before her next menstrual period, and remembered that her menstrual period had started while she had been in the hospital every single time. Her periods, which remained regular (every 28 days), lasted 3 or 4 days and the flow was perhaps only slightly increased over the last 2 years. The patient was married, had 2 children aged 8 and 11 and worked as a caregiver in a nursing home. Her medical history was noteworthy only for a laparoscopic cholecystectomy 3 years previously. Her medications included only lispro insulin (Humalog), through her Medtronic MiniMed 508 Insulin Pump, with an average total daily insulin dose of 35 to 40 units per 24 hours. Her gynecologic history revealed a menarche at age 12. Since then, she had always had regular menstrual periods every 28 days; however, she had noticed a slight increase in the flow (3– 4 days) over the last 2 years, and she had also developed more cramps and associated loose stools or diarrhea and nausea during her premenstrual periods. She had been pregnant 3 times, and had a miscarriage and 2 cesarean deliveries; her last delivery was 8 years prior. On physical examination, she looked healthy and had a normal facies and body habitus. Her height was 5 feet 4 inches, weight was 125 lbs, blood pressure was 100/70, and her heart rate was 88 per minute. The remainder of the examination was normal and demonstrated no evidence of any chronic complications from her diabetes. Case 2 A 35-year-old black woman with diabetes mellitus was referred to our Diabetes Clinic to asses the possibility of using CSII therapy because of extreme difficulty controlling her blood glucose. The patient had been diagnosed with T1DM, 18 years earlier, at the age of 17 and had been under poor control since then. Over the last 6 months, she began experiencing greater difficulty with her glucose control and had been in and out of the hospital with recurrent episodes of nausea and vomiting; for this reason, she had been diagnosed with diabetic gastroparesis. She had also been recently diagnosed with proliferative diabetic retinopathy (PDR) and diabetic nephropathy, which manifested with proteinuria, but with normal creatinine clearance; additionally, her medical history was remarkable for hypertension, iron deficiency anemia, and 2 cesarean deliveries. Her only medication
THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
was insulin in a multiple daily injection regimen, providing her with an average total daily insulin dose of 60 units per day and metochlopramide before meals. She was divorced, had 2 children, and, although she used to manage a restaurant, was now on disability on the basis of her multiple diabetic complications. There was no immediate family history of diabetes mellitus. Her gynecologic history revealed regular, 30-day cycle menstrual periods since her menarche at age 11. She had been pregnant twice and delivered through cesarean delivery both times, the last time 13 years prior. On physical examination she was thin, with a height of 5 feet 2 inches, weight of 124 lbs, blood pressure of 130/90, and a regular heart rate of 100 per minute. She had obvious evidence of bilateral proliferative diabetic retinopathy on direct ophthalmoscopic examination, and evidence of distal symmetric sensory polyneuropathy in both lower extremities, on neurologic examination. Her skin also revealed mild acanthosis nigricans in her neck. Laboratory tests revealed the presence of anti-GAD antibodies and undetectable c-peptide levels in plasma, confirming the autoimmune origin of her disorder and the diagnosis of type 1 diabetes mellitus. Two weeks later, without a clear reason, her blood glucose control rapidly deteriorated, and she developed nausea and vomiting. The patient was admitted to the hospital with DKA. During her second day in the hospital, her menstrual period started at the expected date. The patient then recalled that her glucose had a tendency to run higher during the immediate premenstrual period, starting 2 or 3 days before the onset of her menses, and on through the first 3 or 4 days of her menstrual cycle. She, like the first patient, noted that her menses would often begin in the hospital while she was admitted for DKA. Outside hospital records were obtained and confirmed 6 hospital admissions for DKA over the previous 2 years. The physicians caring for her tested for common causes of DKA, but interestingly, no mention of her gynecologic history was made in any of her hospital records.
Discussion Although, the association between menstrual cycle and diabetes mellitus, glycemic control, glucose metabolism, insulin resistance, and even DKA has long been documented, it remains poorly recognized by many clinicians, perhaps due, at least in part, to the lack of a distinct diagnostic term to describe it. Hereby, we propose that the terms catamenial DKA and catamenial hyperglycemia be used to refer to these disorders. The adjective catamenial is derived from the Greek expression katamenios, which means monthly, and it is frequently used to describe medical disorders that are associated, or occur, around the time of menses. Exacerbation of many medical disorders in connection with the different phases of the menstrual cycle is a rather common and well-described phenomenon; however, the mechanism responsible for these associations is still poorly understood, although often attributed to reproductive hormones. Entities as diverse as epilepsy, pneumothoraces, migraines, asthma, flares of rheumatoid arthritis, and many others, have been attributed to catamenial causes.7 The earliest description of the association between menstrual cycles and diabetes mellitus dates back to 1906, when menstruation was associated with a tendency for increased glycosuria.8 In addition, to date, there have been 7 reports published in the medical 299
Catamenial DKA
Table 1. Summary of Reported Cases Describing Catamenial Hyperglycemic Crisis or DKA Authors
Year Reported
Patient Age (yr)
Race
Clinical Presentation
Country
Reference No.
Harrop et al Rosenbloom Peperkorn
1918 1921 1932
Black
9 10 11
1946 1954 1958 1969 2008
DKA DKA DKA DKA Hyperglycemic crisis Hyperglycemic crisis DKA DKA Hyperglycemic crisis DKA DKA DKA
USA USA Germany
Morton et al Hubble Green Sandström Ovalle et al
18 30 19 25 32 49 31 15 18 17 32 35
USA UK USA Sweden USA
12 13 14 15 —
White White White White Black
literature, including 10 patients, demonstrating an association between menstruation and diabetic ketoacidosis, or severe hyperglycemic crisis9 –15 (Table 1). The first report was by Harrop and Mosenthal from the Medical Clinic of the John Hopkins Hospital in 1918. They described the unfortunate case of an 18-year-old black female who developed T1DM 1 year earlier. During the last few months of her life, her condition seemed to worsen during her menstruation until she ultimately died of DKA during one of her menstrual periods.9 Subsequently, 6 other reports have appeared very sporadically in the literature: In 1921, Rosenbloom observed significant and recurrent glycosuria during menstruation in 2 women, one of whom later developed DKA during her menstrual period and died.10 In 1932, Peperkorn described 2 cases of recurrent DKA occurring during menstruation, as well as 2 with recurrent severe hyperglycemia.11 In 1946, Morton and McGavack reported a 31-year-old woman with diabetes who demonstrated premenstrual glycosuria, hyperglycemia and acidosis, and postmenstrual hypoglycemic episodes.12 Later, in 1954, Hubble described an extreme case of a 15-year-old girl’s phasic insulin resistance, occasionally complicated by ketoacidosis, requiring up to 19,250 units of insulin per day; the resistance phases appeared to be premenstrual with sensitive phases occurring postmenstrually.13 Similarly, in 1958, Green reported a case of phasic insulin resistance associated with the menstrual cycle, but no ketoacidosis, that responded to treatment with fluorohydrocortisone given only during the luteal phase and menstruation.14 Lastly, in 1969, Sandstrom described a case of a 17-year-old patient who developed DKA with every menstrual period during one year, and who became free from these attacks after sequential administration of ethinyloestradiol and dydrogesterone.15 In addition, Cramer performed a retrospective study to investigate the influence of menstruation on DKA among regularly menstruating female patients aged 14 to 45 admitted to their hospital with DKA during the period between 1922 and 1940. He 300
reported that 47.2% of the patients, in whom there was no other known or identifiable precipitating event, were menstruating at the time of their hospitalization; furthermore, in the same study, the difference in the incidence of DKA in men and women was also striking: there were practically 3 times as many cases of acidosis in women as in men in the age period in which the normal woman menstruates; this finding was particularly remarkable given the fact that in this age period the incidence of diabetes was known to be greater in men than in women.16 More recently, Walsh and Malins prospectively investigated the incidence of diabetic ketoacidosis, its relation to sex and age over a period of 5 years (1971–1975), as well as the association between DKA and menstruation in patients attending a diabetic clinic. Here again, despite the fact that 50% more men than women in that age group attended their clinic, the investigators reported that DKA occurred more often in women than in men, particularly in younger, reproductive age women, with a female-to-male ratio of 2.6. Moreover, they also found that menstruation was associated with ketoacidosis more often (58% of cases) than would be expected by chance.17 The mechanism through which the menstrual cycle may provoke DKA remains elusive. The body of literature investigating the possible mechanism is substantial yet inconclusive. It seems that there is some factor that leads to altered glucose metabolism at certain points of the menstrual cycle in some individuals, but studies have not reliably demonstrated what that factor might be. An excellent review of the literature was provided by Widom et al in 1992.18 Little definitive knowledge has been gained since, and the following is a brief review of the major avenues of investigation. It has long been known that estrogen is at its peak at the end of the luteal phase, just before the start of menstruation. The earliest investigations of carbohydrate metabolism (late 1960s to early 1970s) during the menstrual cycle involved simply performing April 2008 Volume 335 Number 4
Ovalle et al
oral glucose tolerance tests in normal women during various phases of their cycle. In general, studies performed in nondiabetic women demonstrated fairly consistent results and showed significant differences in insulin sensitivity between the follicular and luteal phases of the menstrual cycle, with a lower insulin sensitivity during the luteal phase of the cycle, which can explain the increase insulin requirements during this phase of the cycle reported by many diabetic women.19,20 Although the findings were not reproduced in an early assessment of the same in diabetics, some studies demonstrated impaired glucose tolerance in the luteal phase during the estrogen peak.21 More recently, studies using better methodology, including the frequently sampled intravenous glucose tolerance test with minimal modeling and glucose clamp studies, continue to demonstrate conflicting results.18,22–30 Diamond et al were able to demonstrate impaired glucose tolerance during the luteal phase using the hyperglycemic glucose clamp technique in healthy women, a state which might more closely represent the physiology of T1DM.25 A follow-up study was performed in 1992 in diabetics in which a paired hyperglycemic glucose clamp was performed in sixteen women with T1DM during their mid-follicular and mid-luteal phases. The investigators noted, that as a whole, the glucose levels were similar in the groups, but there was a subgroup of 7 women with a history of worsening glycemic control around their menses. These women indeed demonstrated a decrease in insulin sensitivity in the luteal phase that did not reach statistical significance; alternatively, the women without this history actually showed a slight improvement in insulin sensitivity, again not statistically significant. The women in the 2 groups were similar in terms of age, duration of diabetes, body mass index, insulin dosages, and A1Cs. The sex hormone levels were also similar except for a much more dramatic rise in the level of estrogens from the follicular to luteal phase in the subjects with a history of poor menstrual hyperglycemia.18 Although, in the study by Widom et al, the investigators were unable to assert with certainty a definite property of the subgroup of women that demonstrated menstrual hyperglycemia, post hoc analysis revealed differential responses on circulating estradiol levels; women who exhibited significantly higher estradiol levels in the luteal phase were found to have significantly lower insulin sensitivity during the same phase of the menstrual cycle. Thus, in this particular study, estradiol levels appeared to influence potential menstrual cycle changes in insulin sensitivity.18 Then again, 2 other small studies using similar methodology demonstrated no differences in insulin sensitivity during the different phases of the menstrual cycle of women with type 1 diabetes.29,30 THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
Another potential mechanism that has been proposed involves the affinity of insulin for its receptor and potential alterations based on the hormonal milieu. Here again, the evidence is contradictory; various investigators have assessed insulin binding in adipocytes, monocytes and erythrocytes and have failed to demonstrate a clear pattern of variation during the menstrual cycle.31–34 Similarly, assessments of hepatic glucose output have failed to yield an explanation.35 The integrity of the counter-regulatory response in the follicular versus luteal phase has also been investigated. In a small group of healthy, non diabetic women, Diamond and others were unable to demonstrate a significant difference in terms of counterregulatory hormones released or their metabolic end points. Each woman was tested both, during the luteal and during the follicular phases of her cycle.36 An obvious question is whether or not suppression of the ovulatory cycle will attenuate menstrual hyperglycemia. Interestingly, one of the only reports addressing the relationship between gonadotropin releasing hormone agonists and diabetes is in a patient with catamenial hypoglycemia. Letterie and Fredlund describe a patient with an increase in hypoglycemic events just before and during her menstrual cycle. After failing to improve with medroxyprogesterone acetate, the patient’s menses were stopped using depot leuprolide (a medical oophorectomy) and replacement of estrogen and progesterone. This procedure brought about a dramatic improvement in her hypoglycemic episodes. One other case report in the literature notes a deterioration of glycemic control in a diabetic patient on leuprolide.37,38 Recently, Goldner et al utilized continuous glucose monitoring as a tool to assess glycemia and the menstrual cycle in patients with type 1 diabetes. Although the population was small, the study was useful in that it studied women serially, over 3 cycles, to assess whether patterns persisted. Four women completed the study, and 2 of the 4 showed an increased frequency of hyperglycemia in the luteal phase in all cycles. Estrogen and progesterone levels were followed, and in only 1 of the 2 patients with cyclical hyperglycemia was there a statistically significant positive correlation between progesterone levels and hyperglycemia.39 Although sex hormones are favorite suspects in the search for the etiology and pathophysiology of these catamenial changes in glucose homeostasis, other lessrecognized factors may play a significant role in these glycemic excursions. These include the luteal phase disorders, premenstrual syndrome (PMS), and premenstrual dysphoric disorder. An excellent review of the relationship between insulin sensitivity and PMS, by Trout and Teff, was recently published.40 Evidence supporting a possible role of PMS as a causative factor in catamenial dysglycemia include 301
Catamenial DKA
the observation that women with diabetes mellitus have a higher prevalence of menstrual disorders in general.41,42 There is one epidemiologic study that found that women with diabetes and PMS adjust their insulin doses around the time of their menses to a greater extent than those with no PMS which suggests that PMS may influence insulin sensitivity; however, to date, there is no properly designed studies to answer this question.43 Changes in the quality and quantity of food intake have also been observed in association with the different phases of the menstrual cycle, including bingeing and increased carbohydrate intake during the luteal phase, in studies of nondiabetic women.44,45 Although this phenomenon has not been studied in women with diabetes, it is reasonable to hypothesize that similar dietary changes might take place and could play a role in the deterioration of glycemic control sometimes seen in women with diabetes around the time of their menses. Lastly, endometriosis must be included in the list of potential factors leading to catamenial dysglycemic disorders. Although, endometriosis has been a well-recognized etiologic factor in the pathogenesis of other catamenial disorders such as catamenial pneumothorax, hemothorax, and catamenial neuropathies, it has not been implicated in the etiology of diabetic catamenial dysglycemic disorders. Our first patient did have a syndrome that could be consistent with endometriosis; however, this diagnosis could not be confirmed because she was lost to follow-up before a more complete gynecologic evaluation, specifically aimed at ruling out endometriosis, could be performed. In view of the potential pathophysiologic mechanisms involved, and the various approaches that have been described when dealing with these and other catamenial disorders, we believe the following measures can be recommended in the management of catamenial dysglycemic disorders: (a) instruct patient to use a menstrual calendar with detailed blood glucose records; (b) instruct patient to anticipate exacerbations and to try to avoid triggering factors, including dietary changes; (c) instruct patient to make insulin adjustments in anticipation, as well as in response, to blood glucose changes. Conclusion Our observations underscore the significant influence that menstrual cycles can exert on the plasma glucose levels of women with T1DM. Catamenial medical disorders are well established and described, yet poorly understood. Entities as diverse as epilepsy, pneumothoraces, migraines, asthma, flares of rheumatoid arthritis, and many others, have been attributed to catamenial causes.7 Catamenial dysglycemic disorders, including catamenial DKA, catamenial hyperglycemia, and catamenial hypoglycemia, need to be recognized as clinical 302
phenomena. Additionally, catamenial DKA should be considered when no other explanation is found during the evaluation of reproductive age women presenting with diabetic ketoacidosis. Clearly, more studies are needed to further understand the pathophysiology of these disorders and the effects of the menstrual cycle, and related conditions, on glucose metabolism in general. References 1. Faich GA, Fishbein HA, Ellis SE. The epidemiology of diabetic ketoacidosis: a population based study. Am J Epidemiol 1983;117:551– 8. 2. Fishbein H. Diabetic ketoacidosis, hyperosmolar nonketotic coma, lactic acidosis and hypoglycemia. In: Harris MI, Hamman RF, editors. Diabetes in America. Washington, DC: National Diabetes Group, US Department of Health and Human Services; XII 1985 p. 1–16. 3. Umpierrez GE, Kelly JP, Navarrete JE, et al. Hyperglycemic crises in urban blacks. Arch Intern Med 1997;157:669 –75. 4. Kitabchi AE, Umpierrez GE, Murphy MB, et al. Management of hyperglycemic crises in patients with diabetes. Diabetes Care 2001;24:131–53. 5. Anonymus, Diabetes Surveillance 1980 –1987. Atlanta, GA: Division of Diabetes Translation, Centers for Disease Control, U.S. Department of Health and Human Services; 1990. p. 27. 6. Snorgaard O, Eskildesn PC, Vadstrup S, et al. Diabetic ketoacidosis in Denmark: epidemiology, incidence rates, precipitating factors and mortality rates. J Int Med 1989;226: 223. 7. Case AM, Reid RL. Effect of the menstural cycle on medical disorders. Arch Intern Med 1998;158:1405–12. 8. Naunyn B. Der diabetes mellitus. Vienna: Holder; 1906. p. 235. 9. Harrup GA, Mosenthal HO. The influence of menstruation on acidosis in diabetes mellitus. Report of a case. Johns Hopkins Hospital Bulletin, 1918. 10. Rosenbloom J. Influence of menstruation on the food tolerance in diabetes mellitus. JAMA 1921;76:1742. 11. Peperkorn R. Menstruation und diabetes mellitus. Munch Med Wochenschr 1932;79:1748 –50. 12. Morton JH, McGavack TH. The influence of ovarian activity and administered estrogens upon diabetes mellitus: case report. Ann Intern Med 1946;25:154 – 61. 13. Hubble D. Insulin resistance. Br Med J 1954;2:1022– 4. 14. Green R. Phasic insulin resistance associated with the menstrual cycle and controlled by fluorohydrocortisone: a case report. Metabolism 1958;7:90 –2. 15. Sandstro¨m B. Diabetes mellitus och menstruation. Nord Med 1969;81:727– 8. 16. Cramer HI. The influence of menstruation on carbohydrate tolerance in diabetes mellitus. Can Med Assoc J 1942;47: 51–5. 17. Walsh CH, Malins JM. Menstruation and control of diabetes. Br Med J 1977;2:177–9. 18. Widom B, Diamond MP, Simonson DC. Alterations in glucose metabolism during menstrual cycle in women with IDDM. Diabetes Care 1992;15:213–20. 19. Jarret RJ, Graver HJ. Changes in oral glucose tolerance during the menstrual cycle. Br Med J 1968;2:528 –9. 20. Roy SH, Ghosh BP, Bhattacharjee SK. Changes in oral glucose tolerance during normal menstrual cycle. J Indian Med Assoc 1971;57:201– 4. 21. Walsh CH, O’Sullivan DJ. Carbohydrate tolerance during the menstrual cycle in diabetics. Lancet 1973;2:413–5. 22. Pulido JME, Salazar MA. Changes in insulin sensitivity, secretion and glucose effectiveness during menstrual cycle. Arch Med Res 1999;30:19 –22.
April 2008 Volume 335 Number 4
Ovalle et al
23. Valdes CT, Elkind-Hirsch KE. Intravenous glucose tolerance test-derived insulin sensitivity changes during the menstrual cycle. J Clin Endocrinol Metab 1991;72:642– 6. 24. Ezenwaka EC, Akanji AO, Adejuwon CA, et al. Insulin responses following glucose administration in menstruating women. Int J Gynecol Obstet 1993;42:155–9. 25. Diamond MP, Simonson DC, DeFronzo RA. Menstrual cyclicity has a profound effect on glucose homeostasis. Fertil Steril 1989;52:204 – 8. 26. Diamond MP, Jacob R, Connally-Diamond M, et al. Glucose metabolism during the menstrual cycle: assessment with the euglycemic, hyperinsulinemic clamp. J Reprod Med 1993;38:417–21. 27. Yki-Jarvinen, Hannele. Insulin sensitivity during the menstrual cycle. J Clin Endocrinol Metab 1984;59:350 –3. 28. Spellacy WN, Ellingson AB, Keith G, et al. Plasma glucose and insulin levels during the menstrual cycles of normal women and premenstrual syndrome patients. J Reprod Med 1990;35:508 –11. 29. Moberg E, Kollind M, Lins PE, et al. Day-to-day variation of insulin sensitivity in patients with type 1 diabetes: role of gender and menstrual cycle. Diabet Med 1995;12:224 – 8. 30. Scott AR, Macdonald IA, Bowman CA, et al. Effect of phase of menstrual cycle on insulin sensitivity, peripheral blood flow and cardiovascular responses to hyperinsulinemia in young women with type 1 diabetes. Diabet Med 1990;7:57– 62. 31. Pederson O, Hjollund E, Lindkov H. Insulin binding and action on fat cells from young healthy females and males. Am J Phys 1982;243:E158 – 67. 32. Bertoli A, De Pirro R, Fusco A, et al. Differences in insulin receptors between men and menstruating women and influence of sex hormones on insulin binding during the menstrual cycle. J Clin Endocrinol Metab 1980;50:246 –50. 33. Moore P, Kolterman O, Weyant J, et al. Insulin binding in human pregnancy: comparisons to the postpartum, luteal and follicular states. J Clin Endocrinol Metab 1981;52:937– 41. 34. DePirro R, Fusco A, Bertoli A, et al. Insulin receptor
THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
35. 36.
37. 38.
39.
40. 41. 42. 43. 44. 45.
during the menstrual cycle in normal women. J Clin Endocrinol Metab 1978;47:1387–9. Toth EL, Suthijumroon A, Crockford PM, et al. Insulin action does not change during the menstrual cycle in normal women. J Clin Endocrinol Metab 1987;64:74 – 80. Diamond MP, Grainger DA, Rossi G, et al. Counterregulatory response to hypoglycemia in the follicular and luteal phases of the menstrual cycle. Fertil Steril 1993;60: 988 –93. Letterie GS, Fredlund PN. Catamenial insulin reactions treated with long-acting gonadotropin releasing hormone agonist. Arch Intern Med 1994;154:1868 –70. Coddington CC, Hassiakos DK, Harrision HC, et al. Effect of a gonadotropin releasing hormone analogue on the glucose metabolism in a diabetic patient. Gynecol Obstet Invest 1990;30:246 – 8. Goldner W, Krause VL, Sivitz W, et al. Cyclic changes in glycemia assessed by continuous glucose monitoring system during multiple complete menstrual cycles in women with type 1 diabetes. Diabetes Technol Ther 2004;6:473– 80. Trout KK, Teff KL. Insulin sensitivity and premenstrual syndrome. Curr Diab Rep 2004;4:273– 80. Strotmeyer ES, Steenkiste AR, Foley TP, et al. Menstrual cycle differences between women with type 1 diabetes and women without diabetes. Diabetes Care 2003;26:1016 –21. Solomon CG, Hu FB, Dunaif A, et al. Long or highly irregular menstrual cycles as a marker for risk of type 2 diabetes mellitus. JAMA 2001;286:2421– 6. Cawood EH, Bancroft J, Steel JM. Perimenstrual symptoms in women with diabetes mellitus and the relationship to diabetic control. Diabet Med 1993;10:444 – 8. Dalvit-McPhillips SP. The effect of human menstrual cycle on nutrient intake. Physiol Behav 1983;31:209 –12. Both-Orthman B, Rubinow D, Hoban C, et al. Menstrual cycle phase-related changes in appetite in patients with premenstrual syndrome and in control subjects. Am J Psychiatry 1988;145:628 –31.
303
ABSTRACT: Background: Diabetic ketoacidosis (DKA), an acute and potentially life-threatening complication of diabetes mellitus, is frequently the result of a precipitating illness or nonadherence to treatment; however, despite a thorough history, physical, radiologic, and laboratory examination, a specific cause or precipitating event cannot be identified in a significant number of cases. Aims: To report 2 cases of recurrent DKA, and hyperglycemic crisis, associated to the menstrual cycle, and provide a review of the medical literature. Methods: The clinic and hospital medical records of 2 subjects with type 1 diabetes mellitus (T1DM) and unexplained DKA are reviewed. An electronic MEDLINE search of relevant medical literature published from 1965 to 2007 was performed; additionally, the reference lists of the identified articles and other sources, such as textbook chapters and meeting abstracts, were reviewed for related publications. Results: To date there have been 7 reported cases in the literature demonstrating an associ-
ation between menstruation and DKA. The 2 new cases presented here highlight the potentially significant changes in glucose metabolism that may occasionally be observed during the late luteal and decidual phases of the menstrual cycle; moreover, these cases underscore the importance of a thorough medical and gynecologic history when evaluating a patient with DKA. Conclusion: Through unclear mechanisms, some women with diabetes mellitus demonstrate significant changes in glucose control around the time of their menses, including DKA. Accordingly, we propose that the terms catamenial DKA and catamenial hyperglycemia be used to refer to these disorders and that catamenial DKA be included in the differential diagnosis list of causes or precipitating events that can lead to DKA. KEY INDEXING TERMS: Catamenial; Menstruation; Diabetes; Ketoacidosis; Hyperglycemia; Women. [Am J Med Sci 2008; 335(4):298–303.]
D
ating and managing patients with DKA. The most common and frequently highlighted causes of DKA are infections, new-onset diabetes, cerebrovascular accidents, alcohol abuse, pancreatitis, myocardial infarction, trauma, drugs, eating disorders, poor compliance and treatment errors; however, it is well recognized that in a significant number of cases, despite a careful search, no precipitating event is identified. It is also well known that some patients have a tendency to develop repeated episodes of DKA and, in frustration, physicians label them as “brittle diabetics.” Unfortunately, some of these patients may be erroneously accused of poor compliance or even malingering. The influences of the menstrual cycle and female sex hormones on insulin resistance, glucose metabolism, and glycemic control have generated significant controversy over the years. In our clinical experience, we have frequently observed a cyclic deterioration of glycemic control that closely follows the menstrual cycle in some, but not all, women with T1DM. Here, we describe 2 cases of cyclic DKA and cyclic hyperglycemic crisis, which developed during the
iabetic Ketoacidosis (DKA) is an acute and potentially life-threatening complication of diabetes mellitus with reported mortality rates between 2% and 5%.1– 4 The annual incidence of DKA has been reported as high as 1.5 episodes per 100 patients with diabetes per year, and the rates appear to be higher in blacks, lower socioeconomic groups, and, interestingly, in adolescent females.5,6 Appropriately, most textbook chapters and review articles on DKA emphasize the need to rapidly identify the cause or precipitating factor(s) when evalu-
From the Department of Medicine (FO, TBV, JES), Division of Endocrinology, Diabetes and Metabolism; and Department of Nutrition Sciences (BG), The University of Alabama at Birmingham School of Medicine, Birmingham, Alabama. Submitted February 18, 2007; accepted in revised form September 11, 2007. Correspondence: Fernando Ovalle, MD, FACE, Department of Medicine, Division of Endocrinology, Diabetes, and Metabolism, University of Alabama at Birmingham School of Medicine, 510 Twentieth Street South/FOT 702, Birmingham, AL 35294 (E-mail: [email protected]).
298
April 2008 Volume 335 Number 4
Ovalle et al
premenstrual period without any other clear precipitating event. Case Reports Case 1 A 32-year-old white woman with T1DM was referred to our diabetes clinic for evaluation of recurrent episodes of DKA and difficult glycemic control. Six years earlier at the age of 26, she had been diagnosed with T1DM when she reported to her family physician with symptoms from severe hyperglycemia and DKA. She had been managed with insulin since then and her glycemic control had been poor, with HgA1c’s usually around 10%; nonetheless, she had not yet developed any chronic microvascular complications from her diabetes. Just over 1 year prior, she had been started on continuous subcutaneous insulin infusion (CSII) therapy in an attempt to improve her glycemic control. She gave a history of several episodes of DKA since her diagnosis, in average once or twice a year, without any clear or previously identified precipitating events; however, over the last year, she had been admitted to her community hospital 7 times, almost on a monthly basis, with DKA. During each of these admissions to the intensive care unit, her treating physicians were unable to identify a precipitating illness or event despite seemingly thorough physical and laboratory workups, which included among other tests: blood counts and chemical analyses, thyroid function tests, blood gases, urinalysis, chest and abdominal x-ray examinations, electrocardiograms, blood and urine cultures, and even a lumbar spine puncture for cerebrospinal fluid analysis. Remarkably, the patient had noticed that all of these episodes of DKA had developed 2 to 3 days before her next menstrual period, and remembered that her menstrual period had started while she had been in the hospital every single time. Her periods, which remained regular (every 28 days), lasted 3 or 4 days and the flow was perhaps only slightly increased over the last 2 years. The patient was married, had 2 children aged 8 and 11 and worked as a caregiver in a nursing home. Her medical history was noteworthy only for a laparoscopic cholecystectomy 3 years previously. Her medications included only lispro insulin (Humalog), through her Medtronic MiniMed 508 Insulin Pump, with an average total daily insulin dose of 35 to 40 units per 24 hours. Her gynecologic history revealed a menarche at age 12. Since then, she had always had regular menstrual periods every 28 days; however, she had noticed a slight increase in the flow (3– 4 days) over the last 2 years, and she had also developed more cramps and associated loose stools or diarrhea and nausea during her premenstrual periods. She had been pregnant 3 times, and had a miscarriage and 2 cesarean deliveries; her last delivery was 8 years prior. On physical examination, she looked healthy and had a normal facies and body habitus. Her height was 5 feet 4 inches, weight was 125 lbs, blood pressure was 100/70, and her heart rate was 88 per minute. The remainder of the examination was normal and demonstrated no evidence of any chronic complications from her diabetes. Case 2 A 35-year-old black woman with diabetes mellitus was referred to our Diabetes Clinic to asses the possibility of using CSII therapy because of extreme difficulty controlling her blood glucose. The patient had been diagnosed with T1DM, 18 years earlier, at the age of 17 and had been under poor control since then. Over the last 6 months, she began experiencing greater difficulty with her glucose control and had been in and out of the hospital with recurrent episodes of nausea and vomiting; for this reason, she had been diagnosed with diabetic gastroparesis. She had also been recently diagnosed with proliferative diabetic retinopathy (PDR) and diabetic nephropathy, which manifested with proteinuria, but with normal creatinine clearance; additionally, her medical history was remarkable for hypertension, iron deficiency anemia, and 2 cesarean deliveries. Her only medication
THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
was insulin in a multiple daily injection regimen, providing her with an average total daily insulin dose of 60 units per day and metochlopramide before meals. She was divorced, had 2 children, and, although she used to manage a restaurant, was now on disability on the basis of her multiple diabetic complications. There was no immediate family history of diabetes mellitus. Her gynecologic history revealed regular, 30-day cycle menstrual periods since her menarche at age 11. She had been pregnant twice and delivered through cesarean delivery both times, the last time 13 years prior. On physical examination she was thin, with a height of 5 feet 2 inches, weight of 124 lbs, blood pressure of 130/90, and a regular heart rate of 100 per minute. She had obvious evidence of bilateral proliferative diabetic retinopathy on direct ophthalmoscopic examination, and evidence of distal symmetric sensory polyneuropathy in both lower extremities, on neurologic examination. Her skin also revealed mild acanthosis nigricans in her neck. Laboratory tests revealed the presence of anti-GAD antibodies and undetectable c-peptide levels in plasma, confirming the autoimmune origin of her disorder and the diagnosis of type 1 diabetes mellitus. Two weeks later, without a clear reason, her blood glucose control rapidly deteriorated, and she developed nausea and vomiting. The patient was admitted to the hospital with DKA. During her second day in the hospital, her menstrual period started at the expected date. The patient then recalled that her glucose had a tendency to run higher during the immediate premenstrual period, starting 2 or 3 days before the onset of her menses, and on through the first 3 or 4 days of her menstrual cycle. She, like the first patient, noted that her menses would often begin in the hospital while she was admitted for DKA. Outside hospital records were obtained and confirmed 6 hospital admissions for DKA over the previous 2 years. The physicians caring for her tested for common causes of DKA, but interestingly, no mention of her gynecologic history was made in any of her hospital records.
Discussion Although, the association between menstrual cycle and diabetes mellitus, glycemic control, glucose metabolism, insulin resistance, and even DKA has long been documented, it remains poorly recognized by many clinicians, perhaps due, at least in part, to the lack of a distinct diagnostic term to describe it. Hereby, we propose that the terms catamenial DKA and catamenial hyperglycemia be used to refer to these disorders. The adjective catamenial is derived from the Greek expression katamenios, which means monthly, and it is frequently used to describe medical disorders that are associated, or occur, around the time of menses. Exacerbation of many medical disorders in connection with the different phases of the menstrual cycle is a rather common and well-described phenomenon; however, the mechanism responsible for these associations is still poorly understood, although often attributed to reproductive hormones. Entities as diverse as epilepsy, pneumothoraces, migraines, asthma, flares of rheumatoid arthritis, and many others, have been attributed to catamenial causes.7 The earliest description of the association between menstrual cycles and diabetes mellitus dates back to 1906, when menstruation was associated with a tendency for increased glycosuria.8 In addition, to date, there have been 7 reports published in the medical 299
Catamenial DKA
Table 1. Summary of Reported Cases Describing Catamenial Hyperglycemic Crisis or DKA Authors
Year Reported
Patient Age (yr)
Race
Clinical Presentation
Country
Reference No.
Harrop et al Rosenbloom Peperkorn
1918 1921 1932
Black
9 10 11
1946 1954 1958 1969 2008
DKA DKA DKA DKA Hyperglycemic crisis Hyperglycemic crisis DKA DKA Hyperglycemic crisis DKA DKA DKA
USA USA Germany
Morton et al Hubble Green Sandström Ovalle et al
18 30 19 25 32 49 31 15 18 17 32 35
USA UK USA Sweden USA
12 13 14 15 —
White White White White Black
literature, including 10 patients, demonstrating an association between menstruation and diabetic ketoacidosis, or severe hyperglycemic crisis9 –15 (Table 1). The first report was by Harrop and Mosenthal from the Medical Clinic of the John Hopkins Hospital in 1918. They described the unfortunate case of an 18-year-old black female who developed T1DM 1 year earlier. During the last few months of her life, her condition seemed to worsen during her menstruation until she ultimately died of DKA during one of her menstrual periods.9 Subsequently, 6 other reports have appeared very sporadically in the literature: In 1921, Rosenbloom observed significant and recurrent glycosuria during menstruation in 2 women, one of whom later developed DKA during her menstrual period and died.10 In 1932, Peperkorn described 2 cases of recurrent DKA occurring during menstruation, as well as 2 with recurrent severe hyperglycemia.11 In 1946, Morton and McGavack reported a 31-year-old woman with diabetes who demonstrated premenstrual glycosuria, hyperglycemia and acidosis, and postmenstrual hypoglycemic episodes.12 Later, in 1954, Hubble described an extreme case of a 15-year-old girl’s phasic insulin resistance, occasionally complicated by ketoacidosis, requiring up to 19,250 units of insulin per day; the resistance phases appeared to be premenstrual with sensitive phases occurring postmenstrually.13 Similarly, in 1958, Green reported a case of phasic insulin resistance associated with the menstrual cycle, but no ketoacidosis, that responded to treatment with fluorohydrocortisone given only during the luteal phase and menstruation.14 Lastly, in 1969, Sandstrom described a case of a 17-year-old patient who developed DKA with every menstrual period during one year, and who became free from these attacks after sequential administration of ethinyloestradiol and dydrogesterone.15 In addition, Cramer performed a retrospective study to investigate the influence of menstruation on DKA among regularly menstruating female patients aged 14 to 45 admitted to their hospital with DKA during the period between 1922 and 1940. He 300
reported that 47.2% of the patients, in whom there was no other known or identifiable precipitating event, were menstruating at the time of their hospitalization; furthermore, in the same study, the difference in the incidence of DKA in men and women was also striking: there were practically 3 times as many cases of acidosis in women as in men in the age period in which the normal woman menstruates; this finding was particularly remarkable given the fact that in this age period the incidence of diabetes was known to be greater in men than in women.16 More recently, Walsh and Malins prospectively investigated the incidence of diabetic ketoacidosis, its relation to sex and age over a period of 5 years (1971–1975), as well as the association between DKA and menstruation in patients attending a diabetic clinic. Here again, despite the fact that 50% more men than women in that age group attended their clinic, the investigators reported that DKA occurred more often in women than in men, particularly in younger, reproductive age women, with a female-to-male ratio of 2.6. Moreover, they also found that menstruation was associated with ketoacidosis more often (58% of cases) than would be expected by chance.17 The mechanism through which the menstrual cycle may provoke DKA remains elusive. The body of literature investigating the possible mechanism is substantial yet inconclusive. It seems that there is some factor that leads to altered glucose metabolism at certain points of the menstrual cycle in some individuals, but studies have not reliably demonstrated what that factor might be. An excellent review of the literature was provided by Widom et al in 1992.18 Little definitive knowledge has been gained since, and the following is a brief review of the major avenues of investigation. It has long been known that estrogen is at its peak at the end of the luteal phase, just before the start of menstruation. The earliest investigations of carbohydrate metabolism (late 1960s to early 1970s) during the menstrual cycle involved simply performing April 2008 Volume 335 Number 4
Ovalle et al
oral glucose tolerance tests in normal women during various phases of their cycle. In general, studies performed in nondiabetic women demonstrated fairly consistent results and showed significant differences in insulin sensitivity between the follicular and luteal phases of the menstrual cycle, with a lower insulin sensitivity during the luteal phase of the cycle, which can explain the increase insulin requirements during this phase of the cycle reported by many diabetic women.19,20 Although the findings were not reproduced in an early assessment of the same in diabetics, some studies demonstrated impaired glucose tolerance in the luteal phase during the estrogen peak.21 More recently, studies using better methodology, including the frequently sampled intravenous glucose tolerance test with minimal modeling and glucose clamp studies, continue to demonstrate conflicting results.18,22–30 Diamond et al were able to demonstrate impaired glucose tolerance during the luteal phase using the hyperglycemic glucose clamp technique in healthy women, a state which might more closely represent the physiology of T1DM.25 A follow-up study was performed in 1992 in diabetics in which a paired hyperglycemic glucose clamp was performed in sixteen women with T1DM during their mid-follicular and mid-luteal phases. The investigators noted, that as a whole, the glucose levels were similar in the groups, but there was a subgroup of 7 women with a history of worsening glycemic control around their menses. These women indeed demonstrated a decrease in insulin sensitivity in the luteal phase that did not reach statistical significance; alternatively, the women without this history actually showed a slight improvement in insulin sensitivity, again not statistically significant. The women in the 2 groups were similar in terms of age, duration of diabetes, body mass index, insulin dosages, and A1Cs. The sex hormone levels were also similar except for a much more dramatic rise in the level of estrogens from the follicular to luteal phase in the subjects with a history of poor menstrual hyperglycemia.18 Although, in the study by Widom et al, the investigators were unable to assert with certainty a definite property of the subgroup of women that demonstrated menstrual hyperglycemia, post hoc analysis revealed differential responses on circulating estradiol levels; women who exhibited significantly higher estradiol levels in the luteal phase were found to have significantly lower insulin sensitivity during the same phase of the menstrual cycle. Thus, in this particular study, estradiol levels appeared to influence potential menstrual cycle changes in insulin sensitivity.18 Then again, 2 other small studies using similar methodology demonstrated no differences in insulin sensitivity during the different phases of the menstrual cycle of women with type 1 diabetes.29,30 THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
Another potential mechanism that has been proposed involves the affinity of insulin for its receptor and potential alterations based on the hormonal milieu. Here again, the evidence is contradictory; various investigators have assessed insulin binding in adipocytes, monocytes and erythrocytes and have failed to demonstrate a clear pattern of variation during the menstrual cycle.31–34 Similarly, assessments of hepatic glucose output have failed to yield an explanation.35 The integrity of the counter-regulatory response in the follicular versus luteal phase has also been investigated. In a small group of healthy, non diabetic women, Diamond and others were unable to demonstrate a significant difference in terms of counterregulatory hormones released or their metabolic end points. Each woman was tested both, during the luteal and during the follicular phases of her cycle.36 An obvious question is whether or not suppression of the ovulatory cycle will attenuate menstrual hyperglycemia. Interestingly, one of the only reports addressing the relationship between gonadotropin releasing hormone agonists and diabetes is in a patient with catamenial hypoglycemia. Letterie and Fredlund describe a patient with an increase in hypoglycemic events just before and during her menstrual cycle. After failing to improve with medroxyprogesterone acetate, the patient’s menses were stopped using depot leuprolide (a medical oophorectomy) and replacement of estrogen and progesterone. This procedure brought about a dramatic improvement in her hypoglycemic episodes. One other case report in the literature notes a deterioration of glycemic control in a diabetic patient on leuprolide.37,38 Recently, Goldner et al utilized continuous glucose monitoring as a tool to assess glycemia and the menstrual cycle in patients with type 1 diabetes. Although the population was small, the study was useful in that it studied women serially, over 3 cycles, to assess whether patterns persisted. Four women completed the study, and 2 of the 4 showed an increased frequency of hyperglycemia in the luteal phase in all cycles. Estrogen and progesterone levels were followed, and in only 1 of the 2 patients with cyclical hyperglycemia was there a statistically significant positive correlation between progesterone levels and hyperglycemia.39 Although sex hormones are favorite suspects in the search for the etiology and pathophysiology of these catamenial changes in glucose homeostasis, other lessrecognized factors may play a significant role in these glycemic excursions. These include the luteal phase disorders, premenstrual syndrome (PMS), and premenstrual dysphoric disorder. An excellent review of the relationship between insulin sensitivity and PMS, by Trout and Teff, was recently published.40 Evidence supporting a possible role of PMS as a causative factor in catamenial dysglycemia include 301
Catamenial DKA
the observation that women with diabetes mellitus have a higher prevalence of menstrual disorders in general.41,42 There is one epidemiologic study that found that women with diabetes and PMS adjust their insulin doses around the time of their menses to a greater extent than those with no PMS which suggests that PMS may influence insulin sensitivity; however, to date, there is no properly designed studies to answer this question.43 Changes in the quality and quantity of food intake have also been observed in association with the different phases of the menstrual cycle, including bingeing and increased carbohydrate intake during the luteal phase, in studies of nondiabetic women.44,45 Although this phenomenon has not been studied in women with diabetes, it is reasonable to hypothesize that similar dietary changes might take place and could play a role in the deterioration of glycemic control sometimes seen in women with diabetes around the time of their menses. Lastly, endometriosis must be included in the list of potential factors leading to catamenial dysglycemic disorders. Although, endometriosis has been a well-recognized etiologic factor in the pathogenesis of other catamenial disorders such as catamenial pneumothorax, hemothorax, and catamenial neuropathies, it has not been implicated in the etiology of diabetic catamenial dysglycemic disorders. Our first patient did have a syndrome that could be consistent with endometriosis; however, this diagnosis could not be confirmed because she was lost to follow-up before a more complete gynecologic evaluation, specifically aimed at ruling out endometriosis, could be performed. In view of the potential pathophysiologic mechanisms involved, and the various approaches that have been described when dealing with these and other catamenial disorders, we believe the following measures can be recommended in the management of catamenial dysglycemic disorders: (a) instruct patient to use a menstrual calendar with detailed blood glucose records; (b) instruct patient to anticipate exacerbations and to try to avoid triggering factors, including dietary changes; (c) instruct patient to make insulin adjustments in anticipation, as well as in response, to blood glucose changes. Conclusion Our observations underscore the significant influence that menstrual cycles can exert on the plasma glucose levels of women with T1DM. Catamenial medical disorders are well established and described, yet poorly understood. Entities as diverse as epilepsy, pneumothoraces, migraines, asthma, flares of rheumatoid arthritis, and many others, have been attributed to catamenial causes.7 Catamenial dysglycemic disorders, including catamenial DKA, catamenial hyperglycemia, and catamenial hypoglycemia, need to be recognized as clinical 302
phenomena. Additionally, catamenial DKA should be considered when no other explanation is found during the evaluation of reproductive age women presenting with diabetic ketoacidosis. Clearly, more studies are needed to further understand the pathophysiology of these disorders and the effects of the menstrual cycle, and related conditions, on glucose metabolism in general. References 1. Faich GA, Fishbein HA, Ellis SE. The epidemiology of diabetic ketoacidosis: a population based study. Am J Epidemiol 1983;117:551– 8. 2. Fishbein H. Diabetic ketoacidosis, hyperosmolar nonketotic coma, lactic acidosis and hypoglycemia. In: Harris MI, Hamman RF, editors. Diabetes in America. Washington, DC: National Diabetes Group, US Department of Health and Human Services; XII 1985 p. 1–16. 3. Umpierrez GE, Kelly JP, Navarrete JE, et al. Hyperglycemic crises in urban blacks. Arch Intern Med 1997;157:669 –75. 4. Kitabchi AE, Umpierrez GE, Murphy MB, et al. Management of hyperglycemic crises in patients with diabetes. Diabetes Care 2001;24:131–53. 5. Anonymus, Diabetes Surveillance 1980 –1987. Atlanta, GA: Division of Diabetes Translation, Centers for Disease Control, U.S. Department of Health and Human Services; 1990. p. 27. 6. Snorgaard O, Eskildesn PC, Vadstrup S, et al. Diabetic ketoacidosis in Denmark: epidemiology, incidence rates, precipitating factors and mortality rates. J Int Med 1989;226: 223. 7. Case AM, Reid RL. Effect of the menstural cycle on medical disorders. Arch Intern Med 1998;158:1405–12. 8. Naunyn B. Der diabetes mellitus. Vienna: Holder; 1906. p. 235. 9. Harrup GA, Mosenthal HO. The influence of menstruation on acidosis in diabetes mellitus. Report of a case. Johns Hopkins Hospital Bulletin, 1918. 10. Rosenbloom J. Influence of menstruation on the food tolerance in diabetes mellitus. JAMA 1921;76:1742. 11. Peperkorn R. Menstruation und diabetes mellitus. Munch Med Wochenschr 1932;79:1748 –50. 12. Morton JH, McGavack TH. The influence of ovarian activity and administered estrogens upon diabetes mellitus: case report. Ann Intern Med 1946;25:154 – 61. 13. Hubble D. Insulin resistance. Br Med J 1954;2:1022– 4. 14. Green R. Phasic insulin resistance associated with the menstrual cycle and controlled by fluorohydrocortisone: a case report. Metabolism 1958;7:90 –2. 15. Sandstro¨m B. Diabetes mellitus och menstruation. Nord Med 1969;81:727– 8. 16. Cramer HI. The influence of menstruation on carbohydrate tolerance in diabetes mellitus. Can Med Assoc J 1942;47: 51–5. 17. Walsh CH, Malins JM. Menstruation and control of diabetes. Br Med J 1977;2:177–9. 18. Widom B, Diamond MP, Simonson DC. Alterations in glucose metabolism during menstrual cycle in women with IDDM. Diabetes Care 1992;15:213–20. 19. Jarret RJ, Graver HJ. Changes in oral glucose tolerance during the menstrual cycle. Br Med J 1968;2:528 –9. 20. Roy SH, Ghosh BP, Bhattacharjee SK. Changes in oral glucose tolerance during normal menstrual cycle. J Indian Med Assoc 1971;57:201– 4. 21. Walsh CH, O’Sullivan DJ. Carbohydrate tolerance during the menstrual cycle in diabetics. Lancet 1973;2:413–5. 22. Pulido JME, Salazar MA. Changes in insulin sensitivity, secretion and glucose effectiveness during menstrual cycle. Arch Med Res 1999;30:19 –22.
April 2008 Volume 335 Number 4
Ovalle et al
23. Valdes CT, Elkind-Hirsch KE. Intravenous glucose tolerance test-derived insulin sensitivity changes during the menstrual cycle. J Clin Endocrinol Metab 1991;72:642– 6. 24. Ezenwaka EC, Akanji AO, Adejuwon CA, et al. Insulin responses following glucose administration in menstruating women. Int J Gynecol Obstet 1993;42:155–9. 25. Diamond MP, Simonson DC, DeFronzo RA. Menstrual cyclicity has a profound effect on glucose homeostasis. Fertil Steril 1989;52:204 – 8. 26. Diamond MP, Jacob R, Connally-Diamond M, et al. Glucose metabolism during the menstrual cycle: assessment with the euglycemic, hyperinsulinemic clamp. J Reprod Med 1993;38:417–21. 27. Yki-Jarvinen, Hannele. Insulin sensitivity during the menstrual cycle. J Clin Endocrinol Metab 1984;59:350 –3. 28. Spellacy WN, Ellingson AB, Keith G, et al. Plasma glucose and insulin levels during the menstrual cycles of normal women and premenstrual syndrome patients. J Reprod Med 1990;35:508 –11. 29. Moberg E, Kollind M, Lins PE, et al. Day-to-day variation of insulin sensitivity in patients with type 1 diabetes: role of gender and menstrual cycle. Diabet Med 1995;12:224 – 8. 30. Scott AR, Macdonald IA, Bowman CA, et al. Effect of phase of menstrual cycle on insulin sensitivity, peripheral blood flow and cardiovascular responses to hyperinsulinemia in young women with type 1 diabetes. Diabet Med 1990;7:57– 62. 31. Pederson O, Hjollund E, Lindkov H. Insulin binding and action on fat cells from young healthy females and males. Am J Phys 1982;243:E158 – 67. 32. Bertoli A, De Pirro R, Fusco A, et al. Differences in insulin receptors between men and menstruating women and influence of sex hormones on insulin binding during the menstrual cycle. J Clin Endocrinol Metab 1980;50:246 –50. 33. Moore P, Kolterman O, Weyant J, et al. Insulin binding in human pregnancy: comparisons to the postpartum, luteal and follicular states. J Clin Endocrinol Metab 1981;52:937– 41. 34. DePirro R, Fusco A, Bertoli A, et al. Insulin receptor
THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
35. 36.
37. 38.
39.
40. 41. 42. 43. 44. 45.
during the menstrual cycle in normal women. J Clin Endocrinol Metab 1978;47:1387–9. Toth EL, Suthijumroon A, Crockford PM, et al. Insulin action does not change during the menstrual cycle in normal women. J Clin Endocrinol Metab 1987;64:74 – 80. Diamond MP, Grainger DA, Rossi G, et al. Counterregulatory response to hypoglycemia in the follicular and luteal phases of the menstrual cycle. Fertil Steril 1993;60: 988 –93. Letterie GS, Fredlund PN. Catamenial insulin reactions treated with long-acting gonadotropin releasing hormone agonist. Arch Intern Med 1994;154:1868 –70. Coddington CC, Hassiakos DK, Harrision HC, et al. Effect of a gonadotropin releasing hormone analogue on the glucose metabolism in a diabetic patient. Gynecol Obstet Invest 1990;30:246 – 8. Goldner W, Krause VL, Sivitz W, et al. Cyclic changes in glycemia assessed by continuous glucose monitoring system during multiple complete menstrual cycles in women with type 1 diabetes. Diabetes Technol Ther 2004;6:473– 80. Trout KK, Teff KL. Insulin sensitivity and premenstrual syndrome. Curr Diab Rep 2004;4:273– 80. Strotmeyer ES, Steenkiste AR, Foley TP, et al. Menstrual cycle differences between women with type 1 diabetes and women without diabetes. Diabetes Care 2003;26:1016 –21. Solomon CG, Hu FB, Dunaif A, et al. Long or highly irregular menstrual cycles as a marker for risk of type 2 diabetes mellitus. JAMA 2001;286:2421– 6. Cawood EH, Bancroft J, Steel JM. Perimenstrual symptoms in women with diabetes mellitus and the relationship to diabetic control. Diabet Med 1993;10:444 – 8. Dalvit-McPhillips SP. The effect of human menstrual cycle on nutrient intake. Physiol Behav 1983;31:209 –12. Both-Orthman B, Rubinow D, Hoban C, et al. Menstrual cycle phase-related changes in appetite in patients with premenstrual syndrome and in control subjects. Am J Psychiatry 1988;145:628 –31.
303