ORIGINAL ARTICLES Hyperosmolarity and Acidosis in Diabetes Mellitus: A Three-year Experience in Rhode Island TOM J. WACHTEL, MD, LINDA M. TETU-MOURADJIAN, RN, DONA L. GOLDMAN, RN, MPH, SUSAN E. ELLIS, MA, PATRICIA S. O'SULLIVAN, EdD Objectives: Diabetic acidosis (DA) a n d the diabetic hypero s m o i a r state (DHS) a r e generally c o n s i d e r e d to be t w o distinct clinical entities. However, clinical experience a n d the literature suggest that there m a y be some overlap. The p u r p o s e s o f this study were 1) to establish the p r o p o r t i o n o f overlap cases, 2 ) to identify a n y ocCurrence o f DHS i n diabetic p a t i e n t s u n d e r the age o f 3 0 y e a r s (likely type I ) a n d a n y occurrence o f DA i n diabetic p a t i e n t s o v e r the age o f 60 y e a r s (likely type II), 3 ) to describe clinical f a c t o r s associated with the development o f DA a n d DHS, a n d 4) to identify f a c t o r s that influence the p r o g n o s i s o f DHS. Patients and methods: 613 cases were identified by retrospective c h a r t r~riew, u s i n g discharge i n f o r m a t i o n f r o m all 15 R h o d e I s l a n d c o m m u n i t y hospitals d u r i n g 1986, 1987, a n d p a r t o f 1988. DA a l o n e [serum glucose (glu) > 300 mg/dL, bicarbonate (HCO~) < 15mEq/L, calculated total s e r u m osmoiarity ( o s m ) < 320 m O s m / L ] w a s the d i a g n o s i s f o r 134 subjects (22%), DHS a l o n e (glu > 600 mg/dL, HCO 3 >--15mEq/L, o s m > 320 m O s m / L ) , f o r 278 subjects (45%), a n d a m i x e d case ( g l u > 600 mg/dL, HCO~ < 15 mEq/L, osm > 320 m O s m / L ) , f o r 200 subjects (33%). I n f o r m a t i o n a b o u t s e r u m o r u r i n a r y ketones w a s available f o r 109 subjects w h o h a d DA a l o n e [103 h a d diabetic ketoacidosis (DKA) ] a n d 144 subjects w h o h a d m i x e d DA a n d DHS (131 h a d m i x e d DKA a n d DHS). All the d a t a w e r e also a n a l y z e d u s i n g the effective osmoiarity a n d a c u t o f f o f 310 m O s m / L f o r this alternative case definition. Results: Patients with DA a l o n e were y o u n g e r ( m e a n age 33 y e a r s ) a n d p a t i e n t s with DHS alone were older ( m e a n age 63years). However, 28 (10%) o f the 2 78 cases o f OHS a l o n e a n d 72 (36%) o f the 200 cases o f m i x e d DA a n d DHS ocCurred i n p a t i e n t s u n d e r the age o f 30. Eighteen cases (13%) o f DA alone a n d 62 cases (31%) o f m i x e d DA a n d DHS ocCurred in p a t i e n t s o v e r the age o f 60. The results were n o t substantially c h a n g e d w h e n effective osmolarity > 310 m O s m / L was used to define h y p e r o s m o l a r i t y a n d w h e n only cases with d o c u m e n t e d OKA were included. A n infection w a s the m o s t c o m m o n p r e c i p i t a t i n g f a c t o r o f DA (30%), DHS (27%), a n d m i x e d cases (32%). Other corn-

Received from the Rhode Island Hospital, Rhode Island Department of Health, and the Brown University Program in Medicine, Providence, Rhode Island. Presented at the 13th annual meeting of the Society of General Internal Medicine, Arlington, Virginia, May 3, 1990. Supported in part by Grant #U32/CCU 100351-09 from the Division of Diabetes Translation of the Centers for Chronic Disease Prevention and Health Promotion of the Centers for Disease Control to the Rhode Island Department of Health, Office of Disease Control, Chronic Disease Division, Diabetes Control Program. Address correspondence and reprint requests to Dr. Wachtel: Division of General Internal Medicine, Rhode Island Hospital, 593 Eddy Street, Providence, RI 02903.

mort associated f a c t o r s i n c l u d e d n o n c o m p l i a n c e with t r e a t m e n t ( 2 0 % f o r DA, 1 2 % f o r DHS, a n d 2 2 % f o r m i x e d cases) a n d p r e v i o u s l y u n d i a g n o s e d diabetes ( 2 4 % f o r DA, 18% f o r DHS, a n d 10% f o r m i x e d cases). N u r s i n g h o m e residents a c c o u n t e d f o r O. 7% o f DA cases, 18% o f DHS cases, a n d 4.5% o f m i x e d cases. Mortality w a s 4% f o r DA, 12% f o r DHS, a n d 9% f o r m i x e d cases. The mortality f o r DHS is the lowest r e p o r t e d in the literature, c o n t i n u i n g a d o w n w a r d t r e n d that began in the 1970s. N o n s u r v i v a l w a s associated with o l d e r age, h i g h e r osmoiarity, a n d nursing h o m e residence. Survival was associated with the p r e s e n c e o f a n infection. Conclusions: 1) m a n y p a t i e n t s experience m i x e d DA (DKA) a n d DHS r a t h e r t h a n e i t h e r c o n d i t i o n alone, 2 ) both DA (DKA) a n d DHS ocCur i n y o u n g a n d old diabetic p e r sons, 3 ) infection is the m o s t c o m m o n p r e d i s p o s i n g f a c t o r f o r either condition, a n d 4) h i g h e r osmoiarity, older age, a n d n u r s i n g h o m e residence a r e associated with n o n s u r viral in DHS. Key words: diabetes meilitus; acidosis; hyperosmoiarity; infection; compliance; n u r s i n g homes. J GEN INTERN IVIED

1991;6:495- 502. THE CLASSIFICATION o f d i a b e t e s m e l l i t u s d i s t i n g u i s h e s t w o m a j o r t y p e s o f d i a b e t e s : t y p e I, o r i n s u l i n - d e p e n d e n t , a n d t y p e II, o r n o n - i n s u l i n - d e p e n d e n t . ~ T y p e I d i a b e t e s u s u a l l y b e g i n s d u r i n g y o u t h b u t c a n o c c u r at a n y age; t y p e II d i a b e t e s g e n e r a l l y b e g i n s a f t e r t h e a g e o f 4 0 y e a r s b u t c a n o c c u r at a n y age. T y p e I d i a b e t e s r e q u i r e s e x o g e n o u s i n s u l i n to p r e v e n t d i a b e t i c k e t o a c i d o s i s (DKA). T y p e II d i a b e t e s is g e n e r a l l y n o t ketosis-prone. The use of the term "diabetic nonketotic hyperosm o l a r c o m a , " is s o m e w h a t r e s t r i c t i v e , b e c a u s e s e v e r a l authors have described mixed cases of diabetic acidosis (DA) a n d h y p e r o s m o l a r i t y ; for e x a m p l e , C a r r o l l a n d Matz 2 r e p o r t e d t h e p r o p o r t i o n o f m i x e d c a s e s to b e as h i g h as 18% in t h e i r s e r i e s o f 2 7 5 s u b j e c t s , Arieff a n d Carroll 3 established the presence of lactic acidosis or k e t o a c i d o s i s in 60% o f t h e i r 33 r e p o r t e d cases, a n d W a c h t e l et al. 4 r e p o r t e d an a n i o n g a p o f > 15 m E q / L in 67% o f 135 cases. T h e w o r d " c o m a " is r e s t r i c t i v e because the level of consciousness, while associated with the level of hyperosmolarity, 5 can vary substant i a l l y a m o n g i n d i v i d u a l s . F o r e x a m p l e , an o s m o l a r i t y > 4 0 0 m O s m / L has b e e n d e s c r i b e d f o r a p a t i e n t w i t h a n o r m a l state o f c o n s c i o u s n e s s . 5 T h e r e f o r e , t h i s

495

496

Wachte/et a/.. HYPEROSMOLARITYAND ACIDOSISIN DtABETF_..S

m e t a b o l i c d e r a n g e m e n t might be m o r e a p p r o p r i a t e l y called the diabetic h y p e r o s m o l a r state (DHS).4 The m e c h a n i s m of DKA is w e l l u n d e r s t o o d and described in several general internal m e d i c i n e textbooks. 6 The m e c h a n i s m w h e r e b y a diabetic person bec o m e s e x t r e m e l y h y p e r g l y c e m i c and h y p e r o s m o l a r is not well understood. H y p e r o s m o l a r i t y itself may result f r o m h y p e r g l y c e m i a - i n d u c e d osmotic diuresis. In addition to free-water loss and severe dehydration, the patient also e x p e r i e n c e s large deficits of s o d i u m and potassium. 7 R e d u c e d thirst after w a t e r deprivation in healthy elderly patients a m a y predispose the geriatric diabetic patient to enter a vicious circle of w o r s e n i n g h y p e r o s m o l a r i t y and osmotic diuresis. Progressive dehydration also affects renal function, limiting the kidney's capacity to excrete glucose, and further aggravates the h y p e r g l y c e m i c state. H y p e r o s m o l a r i t y and potassium d e p l e t i o n m a y also inhibit insulin secretion, 9 thus p e r p e t u a t i n g the vicious circle. Why s o m e patients w h o b e c o m e h y p e r o s m o l a r do not d e v e l o p k e t o n e m i a is not k n o w n at the present time, but several hypotheses have b e e n suggested. The simplest explanation is that the regulation of lipolysis is so sensitive to insulin that the levels of insulin in m a n y patients w i t h DHS may be sufficient to control this p a t h w a y but not to prevent catabolism in the carbohydrate and protein pathways. 1o This situation m a y exist in most type II diabetic persons and in type I diabetic persons w h o are receiving e x o g e n o u s insulin. The m o s t convincing a r g u m e n t against this hypothesis is the observation that circulating plasma insulin concentrations are similar in DHS and DKA. 11 Gerich et al. 9 have suggested that hyperosmolarity inhibits lipolysis, thus minimizing the t e n d e n c y to d e v e l o p DKA. O t h e r m e c h anisms may also play a y e t - u n d e t e r m i n e d role. Cortisol and growth h o r m o n e , b o t h lipolytic hormones, are l o w e r in DHS than in DKA. 11 Finally, s o m e patients w i t h DHS have plasma-free fatty acid levels that are as high as those in DKA, suggesting in these cases i m p a i r e d hepatic synthesis of ketones f r o m free fatty acids w i t h o u t any inhibition of lipolysis.12 Diabetic ketoacidosis and DHS have b e e n conside r e d traditionally two distinct clinical entities. However, classifying diabetic patients into true type I and true type II diabetes in clinical practice is often tentative. Furthermore, DHS has b e e n w e l l described in type I diabetic patients w h o are given an a m o u n t of insulin sufficient to prevent ketosis b u t insufficient to control hyperglycemia. 6 An e p i d e m i o l o g i c study of DA in Rhode Island revealed that 66% of all cases o c c u r in patients u n d e r 55 years of age, 30%, b e t w e e n the ages of 55 and 85, and 4%, over the age of 85.13 The diabetic h y p e r o s m o l a r state usually occurs in patients with type II diabetes and, thus, tends to o c c u r in patients w h o are old. 4 However, the condition has also b e e n described in children as y o u n g as 18 months old, 14 and maturityonset diabetes of the y o u n g is a well-established clini-

cal entity. The most c o m p e l l i n g a r g u m e n t suggesting overlap b e t w e e n DKA and DHS is the notion that hyperosmolarity is c o m m o n in DKA, s, 15 and acidosis has b e e n well d o c u m e n t e d in patients described in case series as having diabetic nonketotic h y p e r o s m o l a r coma. 2-4 The p u r p o s e s of this r e p o r t are 1) to establish the p r o p o r t i o n of cases with overlap features of DA (DKA) and DHS, 2) to identify any o c c u r r e n c e of DHS in diabetic patients u n d e r the age of 30 (likely type I ) and any o c c u r r e n c e of DA (DKA) in diabetic patients over the age of 60 (likely type I I ) , 3) to describe clinical factors associated w i t h the d e v e l o p m e n t of DA (DKA) and DHS, and 4) to describe factors associated w i t h survival from DHS. This research was c o n d u c t e d as part o f a larger study. The Rhode Island DHS project was designed by the Rhode Island D e p a r t m e n t of Health, primarily to describe the e p i d e m i o l o g y of DHS in the state, and to identify c o m m u n i t y - b a s e d interventions to r e d u c e this c o m p l i c a t i o n of diabetes. The p o p u l a t i o n in the State of Rhode Island is a p p r o x i m a t e l y 1 , 0 0 0 , 0 0 0 p e o p l e , and the p r e v a l e n c e of diabetes is 3.5% (personal c o m m u n i c a t i o n , based on a survey d o n e by the Rhode Island D e p a r t m e n t of Health).

METHODS Rhode Island provides a m o d e l for e p i d e m i o l o g i c studies. Rhode Island has 15 c o m m u n i t y hospitals~ all of w h i c h participated in this study. All hospital admissions to acute care beds w e r e surveyed during 1986, 1987, and part of 1988 for admissions related to diabetes. Hospitals made available to the researchers all records of patients w h o had a discharge diagnosis of diabetes. The records w e r e r e v i e w e d and cases w e r e identified. For a record to b e abstracted, a patient had to have a serum glucose > 300 m g / d L and a bicarbonate (HCO3) < 15 mEq/L, or a s e r u m g l u c o s e > 6 0 0 mg/dL. To be i n c l u d e d in the study as a case of DHS, the patient had to have an admission-calculated total s e r u m osmolarity > 320 m O s m / L and a s e r u m glucose > 6 0 0 mg/dL. 16 To be included as a case of DA, the patient had to have a bicarbonate level 15 mEq/L and a s e r u m glucose > 300 mg/dL. Arterial b l o o d gases and s e r u m or urine ketone m e a s u r e m e n t s w e r e frequently unavailable and w e r e not used to define DA. Therefore, s o m e subjects w e r e classified as having DA, although the precise cause of the acidosis was not known. Overlap cases w e r e patients w i t h a b i c a r b o n a t e level < 15 mEq/L and a total s e r u m osmolarity > 320 mOsm/L. Again, the cause of the acidosis was often u n k n o w n and may have b e e n lactic acidosis in a substantial p r o p o r t i o n of cases. To evaluate this potential source of bias, all the analyses w e r e repeated in the s u b g r o u p of subjects for w h o m information a b o u t s e r u m or urinary ketones was available. Patients w i t h HCO 3 < 15mEq/L and positive ket o n e m i a or ketonuria w e r e classified as having DKA. The standard formula used to calculate total osmo-

JOURNALOFGENERALINTERNALMEDICINE,Volume6 (November/December), 1991

larity was: mOsm/L = 2 (Na + K) + g / u / 1 8 + BUN/ 2.8, where Na = sodium concentration in mEodL, K = potassium in mEq/L, glu = serum glucose in mg/dL, and BUN = b l o o d urea nitrogen in mg/dL. The reason for our choice of 320 mOsm/L as a c u t o f f p o i n t for total serum osmolarity was based on p r e c e d e n t in the literature, 16 and because the highest level for total osmolarity is 318 mOsm/L w h e n one replaces each c o m p o n e n t of the formula with its value at the u p p e r limit of the normal range (glu = 180 mg/dL, BUN = 24 mg/dL, Na = 145 mEq/L, K = 5.0 mEq/L). Thus, 320 mOsm/L is clearly abnormal. Many clinicians prefer to use the effective osmolarity (Eosm), which excludes the BUN from its definition. While lower than the measured osmolarity ( w h i c h is approximated by the above formula), the Eosm is more important pathophysiologically in the hyperosmolar state because it is the driving force moving water across cell membranes. Using the definition of total serum hyperosmolarity w o u l d capture azotemic patients who truly do not have DHS because the BUN is the main contributor to their hypersmolar state. Therefore, all the analyses were repeated using the following alternative calculation o f effective osmolarity: Eosm = 2(Na + K) + glu/18. A cutoff of 310 mOsm/L was used to define DHS with this alternative definition. Our c h o i c e of 600 mg/dL as a cutoff point for serum glucose was based on p r e c e d e n t in the l i t e r a t u r e ) 6 The anion gap was calculated according to the following formula: gap = (Na + K) -- (CI + HCO3). The following data were collected for all patients: age; sex; place of residence, dichotomized into h o m e or nursing h o m e / l o n g - t e r m care institution; o u t c o m e of hospitalization (discharged dead or alive); and three predisposing factors for DA or DHS, including 1) new onset of diabetes, 2) presence of an infection, and 3) compliance with treatment (all three as described by the attending physician in the medical record). If a subject had more than one predisposing factor, only one was noted, with the following priority: new onset of diabetes, infection, and noncompliance. In addition, w e r e c o r d e d the following biochemical values obtained for patients u p o n admission to the hospital: serum glucose, BUN, sodium, potassium, chloride, bicarbonate, and the presence of ketones in the b l o o d or urine. We also collected information about level of consciousness on admission according to the physicians' observations as r e c o r d e d in the subjects' charts. Since the terms used by the physicians to describe their patients' levels of consciousness were often vague, w e constructed a three-level classification, including normal (level 1), lethargic or o b t u n d e d but responsive (level 2), and unresponsive (level 3). Sixty-three patients for w h o m information was unavailable about sodium, potassium, or BUN were e x c l u d e d from the analyses involving total serum osmolarity. Seventeen of those (for w h o m BUN was the only missing informa-

497

tion) w e r e included in the analyses involving effective osmolarity. The patients were stratified with three distinct objectives in mind. First, they were classified into three groupings by level o f bicarbonate and osmolarity to demonstrate the existence of a large p r o p o r t i o n of overlap cases. Second, they were divided into three n e w groups by levels of serum osmolarity and glucose to find clinical factors associated with severity and to evaluate the effect of osmolarity on outcome. Third, they were stratified by age, with particular focus on those younger than 30 and those older than 60, because the under-30 group should include a p r e d o m i n a n c e of patients with type I diabetes, while the over-60 group should include a p r e d o m i n a n c e of patients with type II diabetes, based on the epidemiologic characteristics of type I and type II diabetes. These stratifications were p e r f o r m e d for cases defined by calculated total serum osmolarity, repeated for cases defined by calculated effective osmolarity, and repeated again for cases with available information about ketogenesis. Descriptive statistics were used to establish the characteristics of the subjects included in the various groups formed by stratification. We did not perform significance tests to compare these various groups, because they were created arbitrarily. Bivariate significance tests were p e r f o r m e d to compare survivors with nonsurvivors. Significance of differences b e t w e e n proportions was calculated by chi-square tests. Significance o f differences between means was assessed using a two-tailed Student's t test. In addition, multivariate regression was used to assess the independent prognostic importance of the various potential factors; this was done in a stepwise manner.

RESULTS Overlap between DA and DHS Four hundred seventy-eight patients had a calculated total osmolarity > 320 mOsm/L; they comprised the subjects with DHS in this study. Two hundred seventy-eight of these patients (45%) had DHS alone. Three h u n d r e d thirty-four patients had HCO 3 < 15 mEq/L; they comprised all the cases of DA; among these, 200 patients (33%) met both definitions, they were mixed cases of DA and DHS; 134 patients (22%) had DA alone (Fig. 1). O f the 478 cases with total serum hyperosmolarity, 456 (95%) had an effective o s m o l a r i t y > 3 1 0 mOsm/L (Fig. 2). These 456 patients included 262 cases of DHS alone and 194 cases of mixed DA and DHS. Using Eosm --< 310 mOsm/L and HCO 3 < 15 mEq/L as case definition, 157 cases of DA alone were identified, including 23 additional cases, six of w h i c h no longer met the definition of mixed DA and DHS without the contribution of BUN, and 17 in w h i c h BUN was not

498

Wachtel et aL, HYPEROSVaXARrrYAND AaOOSlSIN DIABETES

FIGURE 1. Overlapbetween diabetic acidosis(DA) and the diabetic hyperosmolar state (DHS) (defined by calculatedtotal osmolarity) in 612 subjects in the study.

available and thus these had been e x c l u d e d from the group of 134 cases identified as having a total osmolarity 15 mEq/L).

DHS and DA (Table 1). Two h u n d r e d thirty-four of the 478 subjects with hyperosmolarity (49%) were over the age of 60 years. Eighty of the 334 subjects with acidosis (24%) were over the age of 60 years, including 18 (13%) of the 134 subjects with DA alone and 62 (31%) of the 200 subjects with mixed DAand DHS. One hundred forty-four o f the 334 subjects with DA alone or mixed DA and DHS (43%) were u n d e r the age of 30. The absence of acidosis was m u c h more c o m m o n in the older patients (68% of 252 subjects over the age of 60 vs. 16% of 172 subjects under the age of 30). Unresponsiveness was u n c o m m o n (5%) and mortality was very low (1%) in the less-than-30 age group, c o m p a r e d with 24% unresponsiveness and 20% mortality for those over the age of 60 (not shown in the table). The same comments apply to the subgroup of patients with available information about ketogenesis (Table 2). It is important to note that old patients in the mixed-DKAand-DHS group were more likely to lack this information than were young patients. The DHS-alone group is larger in Table 2 than in Table 1 because it contains 13 additional cases with nonketotic acidosis.

Occurrence of DA (DKA) and DH$ in Young ( < 30 Years) and 01d ( > 60 Years) Patients Patients with DHS alone were older (mean age of 63 years) than those with mixed DA and DHS (mean age 43 years) and those with DAalone (mean age 33 years). One hundred of the 478 subjects with total serum hyperosmolarity (21%) were u n d e r the age of 30 years, representing 28 (10%) of the 278 subjects with DHS alone and 72 (36%) of the 200 subjects with mixed

FIGURE 3. Overlap between casesof diabetic ketoacidosis (DKA) and the diabetic hyperosmolar state (DHS), where DKA is defined by HC03 < 15 mEq/L and by the presenceof ketones in the blood or urine and DHS is defined by a calculatedtotal serum osmolafity > 320 mOsm/L in a subsampleof 525 subjects who were evaluated for the presenceof ketogenesis when acidosis(HC03 < 15 mEq/L) was present. Sixty-four DHS of the subjectshad positive ketones, however, their HCO~was > 15 mEq/L; Therefore, they were classifiedas DHS alone.

Of all 334 subjects w h o had acidosis as defined by HCO 3 < 15 mEq/L, only two had an anion gap < 15 mEq/L, suggesting h y p e r c h o l o r e m i c acidosis. Information about ketogenesis was lacking for b o t h cases, therefore, they remained in Table 1 but w e r e e x c l u d e d from Table 2. Sociodemographic and Clinical

FIGUREZ. Overlapbetween diabetic acidosis(DA) and the diabetic hyperosmolar state (DHS) (defined by calculatedeffective osmolarity) in 613 subjectsin the study. These are not all the same subjectsas in Figure 1, becausecaseswith missing information about blood urea nitrogen are includedhere (appliesto DA cases)and caseswith hyperosmolarity secondary to renal failure are excludedhere (applies mostly to DHS cases).

Characteristics

of Patients with DA and DH$ The 200 subjects w h o had mixed DHS (defined by total serum osmolarity) and DA were y o u n g e r than the 278 patients with DHS alone; they also had the same gender distribution and were less likely to be nursing

JOURNALOFGENERALINTERNALMEDICINE,Volume6 (November/December), 1991

499

TABLE 1 Characteristics of Subjects with AcidosisAlone, the DiabeticHyperosmolarState (DHS) Alone, and Mixed Cases Acidosis Alone (Osmolarity - 320 mOsm/L and HCO3< 15 mEq/L) n (%)

Mixed Acidosis and DHS (Osmolarity > 320 mOsm/L and HC0s < 15 mEcl/L)

134 (220/0)* (1%) (63%) (4%) (20%) (30%) (24%)

DHS Alone (Osmolarity > 320 mOsm/L and HCO3- 15 mEq/L)

200 (33%)t

278 (45o/0)t

9 (4.5%) 117 (58.5%) 18 (9%) 44 (22%) 64 (32%) 20 (109/0)

50 154 33 34 74 50

Nursing home residence Sex--female Mortality Noncompliance Infection New onset of diabetes

I 84 5 27 40 32

Level of consciousness I (normal) 2 (lethargic) 3 (unresponsive)

74 (55%) 54 (40%) 6 (5%)

82 (41%) 88 (44%) 30 (15%)

117 (43%) 110 (40%) 47 (17%)

Age--mean + SD < 30 years > 60 years

33 + 20 years 72 (54%) 18 (13%)

43 + 23 years 72 (36%) 62 (31%)

63 + 21 years 28 (10%) 172 (62%)

Osmolarity Glucose Sodium Potassium Blood urea nitrogen HC03

309 _+8 mOsm/L 515 + 138 mg/dL 131 _+ 5 rnEq/L 4.8 + 1.1 mEq/L 21 _+ 11 mg/dL 9 + 3 mEq/L

342 + 23 mOsm/L 808 + 295 mg/dL 135 --+B mEq/L 5.7 _+ 1.3 mEq/L 45 _+31 mg/dL 9 + 3 mEq/dL

(18%) (55%) (12%) (12%) (27%) (18%)

347 + 26 mOsm/L 861 + 250 mg/dL 136 + 10 mEq/L

5.0 _+0.9 mEq/L 52 + 30 mg/dL 24 + 5 mEq/dL

*All percentagesexcept those in the first row are column percentages. tOne of the 479 hyperosmolarcasesis missing from the analysisbecausethe HC03 level was unavailable.

h o m e residents, more likely to present with an underlying infection, less compliant with diet or drug treatment, and more likely to survive the episode (Table 1). There was no relationship b e t w e e n level o f consciousness and presence of acidosis. The 134 patients with DA alone were younger than the 478 patients with hyperosmolarity; they w e r e more likely to be female, less likely to reside in nursing homes, less compliant with treatment, and more likely to survive, and had a higher level of consciousness. The same results apply ( + 2%) both to the groups based on effective hyperosmolarity and to the subgroups of patients with available information about ketogenesis. Subgroup with Severe DHS. Subjects w h o had severe DHS (glu > 800 mg/dL and osm > 350 m O s m /

L) appeared not to differ from the overall subjects in regard to age and sex (Table 3). Furthermore, t h e y w e r e no more likely to have acidosis than were those with mild DHS, and among the subjects with mixed DA and DHS, the proportions of ketoacidosis were about the same in cases of severe hyperosmolarity (64% of 36) and cases of milder hyperosmolarity (68% of 120). The subjects with severe DHS w e r e more likely to reside in nursing homes, be compliant with diet and drug treatments, and have no prior history of diabetes. These patients were also more likely to have a decreased level of consciousness and to die before hospital discharge. Subgroup with Mixed DA a n d DHS a n d Not SeA subgroup o f 44 patients comprised those patients with DA w h o w e r e also hyverely Hyperglycemic.

TABLE Z Ages of the Subjects with DiabeticKetoacidosis(DKA) Alone, the DiabeticHyperosmolarState (DHS) Alone, and Mixed Casesin the Subgroupof Subjects Who Were Evaluatedfor the Presenceof Ketosis DKA (Osmolarity - 320 mOsm/L and HCO3< 15 mEq/L with PositiveKetones) n (%) Age--mean + SD Age < 30 years Age > 60 years

103 (20%)

Mixed DKA and DHS (Osmolarity > 320 mOsm/L and HCO3 < 15 mEq/L with Positive Ketones) 131 (25%)

DHS Alone (Osmolarity > 320 mOsm/L and HC03 >- 15 mEq/L or Negative Ketones) 291 (55%)

31 + 21 years

36 + 22 years

54 (52%)

68 (52%)

63 + 21 years

30 (I0%)

11 (I I%)

27 (21%)

179 (62%)

Wachtel et al., HYPEROSMOLARITY AND ACIDOSISIN DIABETES

500

TABLE 3

Characteristics of the Subjects with Hyperosmolarity (Osmolarity > 320 mOsm/L)

Glucose< 600 mg/dl; Osmolarity > 320 mOsm/L

Glucose>- 600 mg/dL and Osmolarity > 320 mOsm/L, Excluding Everyone with Both Glucose> 800 and Osmolaribj - 350 mOsm/L

Glucose> 800 mg/dL; Osmolarity -> 350 mOsm/L

Total

n (%)

44 (9%)*

338 (71%)

97 (20%)

479 (100%)

HC03 < 15 Nursing home residence Sex--female Mortality Noncompliance Infection New onset of diabetes

44 (100%) 2 (5%) 24 (55%) 4 (9%) 8 (18%) 18 (41%) 3 (7%)

120 (36%) 42 (12%)

36 (37%) 15 (15%)

189 (56Olo)

58 (600•0)

(9%) (18%) (27%) (12%)

17 (18%) 8 (8%) 28 (29%) 25 (26%)

200 (42%) 59 (12%) 271 (57%) 52(I I%) 78 (16%) 138 (29%)

Level of consciousness 1 (normal) 2 (lethargic) 3 (unresponsive)

23 (52%) 18 (41°/0) 3 (7%)

160 (48%) 136 (40%) 40 (12%)

16 (17%) 44 (46%) 35 (370/0)

Age--mean + SD

40 + 24 years

Osmolarity Glucose Sodium Potassium Blood urea nitrogen HC03

31 62 92 42

55 _+24 years

61 + 20 years

337 + 17 mOsm/L 798 + 192 mg/dL 134 ___9 mEq/L 5.2 ___1.1 mEq/L 43 _+ 25 mg/dL 19 -----9 mEq/L

328 + 11 mOsm/L 488 + 91 137 + 5 mEq/L 5.0 _+0.8 mEq/L 46 +_44 9 _+4 mEq/L

379 + 25 mOsm/L 1147 +-- 277 mg/dL 140 --+ 12 mEq/L 5.5 + 1.4 mEq/L 72 +-- 33 mg/dL 17 +- 9 mEq/L

70

(150/0)

199 (42%) 198 (42%) 78 (16%) 55 _+ 24 years 345 _+25 mOsm/L 840 + 272 rag/alL 135 ___10 mEq/L 5.3 __+1.2 mEq/L 49 ___31 mg/dL 17 ___9 mEq/L

*All percentagesexcept those in the first row are column percentages.

perosmolar (osm > 320 mOsm/L) but not severely hyperglycemic (glu --%600 mg/dL) (Table 3). When defined by effective hyperosmolarity, 44 patients were identified as having Eosm > 310 mOsm/L and glu -600 mg/dL. These patients probably represent patients

with DA who had become so dehydrated and hemoconcentrated that they met our definition of hyperosmolarity. They tended to have preserved mental function as did those with DA alone, but their mortality (9%) was closer to that of the overall hyperosmolar group (11%)

TABLE 4 Mortality of the SubJectswith Hyperosmolarity Lived

Died

n (%)

427 (89%)*

52 (11%)

Nurs!ng home residence Sex--female Noncompliance Infection New onset of diabetes

42 238 73 127 61

(10%) (56%) (17%) (30%) (14%)

17 (33%) 32 (62%) 4 (8%) 11 (21%) 9 (18%)

< 0.0001 0.53 0.12 0.25 0.71

Level of consciousness I (normal) 2 (lethargic) 3 (unresponsive)

189 (45%) 178 (42%) 55 (I 3%)

9 (17%) 2O (380/0) 23 (44%)

< 0.0001

75 _+ 14 years

< 0.001

Age--mean + SD Osmolarity Glucose Sodium Potassium Blood urea nitrogen HC03

53 + 24 years 344 +_ 23 mOsm/L 838 _ 272 mg/dL 134 _ 9 mEq/L 5.3 __+1.2 mEq/L 48 + 30 mg/dL 17 _+9 mEq/L

*All percentagesexceptthose in the first row are column percentages.

362 + 37 mOsm/L 845 _+274 mg/dL 142 + 14 mEq/L 5.0 _+0.9 mEq/L 60 ___30 mg/dL 19 ___9 mEq/L

p

< 0.001 0.69 < 0.001 0.03 0.007 0.26

SO 1

JOURNALOFGENERALINTERNALMEDICINE,Volume 6 (November/December), 1991

than that of the DA alone group (4%). (It is possible that DHS alone may o c c u r in patients with glu 320 mOsm/L. We could not ascertain the cause o f the acidosis for all 613 subjects, in the absence of information about serum or urine ketone levels and serum lactate levels. Nevertheless, the presence of docu m e n t e d ketoacidosis in many cases of DHS and the presence of hyperosmolarity in many cases of DKA lend strong evidence to the c o n c e p t of a continuous spectrum of metabolic changes, with " p u r e DKA" and " p u r e DHS" at the extremes. Therefore, a diagnosis of DKA does not exclude DHS, and vice versa. For the clinician, it is important to calculate the serum osmolarity of all diabetic persons with severe hyperglycemia or with DKA, and it is equally important to evaluate the a c i d - b a s e status and ketogenesis in all hyperglycemic hyperosmolar patients. The p r o p o r t i o n of overlap cases appears to be i n d e p e n d e n t of the definition of hyperosmolarity; indeed, w h e t h e r one defines DHS liberally, as glu -> 600 mg/dL and total serum osmolarity > 320 mOsm/L, or conservatively, as glu > 800 mg/dL and total serum osmolarity > - 350 mOsm/L, the proportions of subjects with serum bicarbonate < 15 mEq/L are approximately the same (36% and 37%, respectively). Second, DHS is not seen exclusively in the elderly or in persons with likely type II diabetes. While cases of DHS in young persons have appeared in the medical literature, w e did not e x p e c t to find that 1 O0 (21%) of the 478 hyperosmolar subjects were u n d e r the age of

TABLE S

Multivariate Predictors of Mortality--Final Equation from Stepwise Results

b Age 0.003 Osmolarity 0.002 Nursing home residence 0.116 Infection --0.069

~

p

Rz Step

0.21 O. 13 0.12 --0.10

< 0.0001 0.006 0.011 0.022

0.08 O.10 0.11 O. 12

30. Thirty-six percent of this younger subgroup had mixed DA and DHS, perhaps representing cases of DA complicated by fluid loss substantial e n o u g h to cause hyperosmolarity > 320 mOsm/L. Nevertheless, 10% of them had no acidosis, suggesting that DHS w i t h o u t acidosis in diabetic persons u n d e r the age of 30 with likely type I diabetes is by no means rare. Alternatively, it is possible that some or many of these cases represented patients with maturity-onset diabetes of the young. In addition, 80 (24%) of 334 cases of acidosis (alone or mixed) o c c u r r e d in patients over the age of 60 with likely type II diabetes (or patients with type I diabetes w h o survived to age 60). If one considers only those cases with available information about ketogenesis, 38 (16%) of the 234 cases of DKA (alone or mixed) occurred in patients over the age of 60. Third, several clinical factors are associated with DA and DHS. Fifteen percent of all cases of DHS o c c u r in patients with no known history of diabetes. This finding has been described previously for both DA and DHS. This observation may justify screening for diabetes in selected populations. A sizeable p r o p o r t i o n of patients with DHS reside in nursing homes. One can assume that this subgroup of patients is more debilitated than the general population in terms of physical and mental ability to function. Thus, these patients are less likely to q u e n c h their thirst. Furthermore, the sense of thirst is sometimes impaired in elderly persons and this will also affect their ability to correct a fluid deficit before the cycle leading to DHS is established. Hypernatremic dehydration has b e e n shown to be an indicator of neglect in nursing h o m e patients, t7 Perhaps an educational program targeted at nursing h o m e residents and staffs to provide information about symptoms and signs of hyperglycemia might prevent some cases of DHS. An acute infection appears to be the most c o m m o n predisposing factor for DA and DHS. This has b e e n described in every published report of a case series. Therefore, physicians should be prepared not only to treat infections p r o m p t l y and vigorously in their diabetic patients, but also to monitor closely glycemia, v o l u m e status, and ketogenesis until the infection has resolved. Finding that n o n c o m p l i a n c e is d o c u m e n t e d in 16% of all hyperosmolar cases and 20% of cases of DA alone is probably an underestimation of n o n c o m p l i a n c e in the subjects of this study. A recent study of n o n c o m p l i a n c e in the general population with regard to medication

50~.

Wachte/eta/..

HYPEROSMOLARrrYAND ACIDOSISIN DL~ETES

shows it to be about 40%; the same study indicates that, more often than not, the physician is not aware of his or her patients' noncompliance. 18 Since our data reflect the information available to the patients' physicians, it is not surprising to find a relatively low level of noncompliance. Noncompliance was highest (29%) in those u n d e r the age of 30. Regardless o f our subjects' levels of compliance, w e believe that patient encouragement to c o m p l y with management plans can only help to prevent DHS and DA. Level of consciousness is quite variable and, while it correlates with osmolarity, as previously reported,t9 17% of those patients in the most severe category of hyperosmolarity had normal mental function. Fourth, mortality in DHS was low at 11% and compared favorably with all published reports of previous series. Even for patients with "severe DHS," mortality was lower (18%) than previously reported. The reason for this finding is not clear, but it most likely represents the dissemination o f knowledge about this n e w syndrome throughout the medical community. The diabetic hyperosmolar state was first described in the South African Medical Journal in 1957,2° but the first series in the American medical literature appeared in the early 1970s. Case series published in the 1980s reported lower case fatality r a t e s - - C a r r o l l and Matz 2 in 1983, 14%, and Wachtel et al. 21 in 1987, 1 7 % than those rates observed in the 1970s: Arieff and Carroll 3 in 1972, 58% and McCurdy 7 in 1970, 40%. The standard management of DHS is n o w w e l l established in many textbooks, ~6 with little controversy, perhaps accounting for the downward trend in reported mortality figures. In the multivariate analysis, nursing h o m e residence, age, and level of hyperosmolarity w e r e the only i n d e p e n d e n t factors associated with nonsurvival. Acute infection was associated with survival. The multivariate analysis explains only 12% of the variance for mortality. This is not surprising, because mechanism of death is often difficult to determine, even with the help of postmortem examination, and because the effect o f treatment was not taken into account in our study. Age 2~ and hyperosmolarity~9. 2t have previously been found to be i n d e p e n d e n t predictors of nonsurvivorship. The presence of an infection as a predictor of a favorable prognosis may imply that patients do better w h e n a treatable, underlying process accompanies DHS. Older patients are more likely to have c o m o r b i d illnesses, w h i c h may affect prognosis unfavorably. Similarly, the nursing h o m e population is c o m p o s e d largely of frail individuals. A high osmolarity indicates a more advanced stage of DHS at the time of diagnosis; delays in diagnosis w o u l d be e x p e c t e d to have an adverse impact on prognosis. It is unfortunate that different authors have described case series of patients with DHS using different case definitions. As a result, direct comparisons among various studies are impossible. Nevertheless, our data

indicate that substantial overlap exists b e t w e e n DKA and DHS, regardless of the case definition o f DHS. We c o n c l u d e that 1) many patients e x p e r i e n c e mixed DA (DKA) and DHS rather than either condition alone, 2) b o t h DA (DKA) and DHS o c c u r in young and old diabetic persons, 3) infection is the most c o m m o n predisposing factor for either condition, and 4) higher osmolarity, older age, and nursing h o m e residence are associated with nonsurvival in DHS.

REFERENCES 1. National Diabetes Group. Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance. Diabetes. 1979;28:1039. 2. Carroll P, Matz R. Uncontrolled diabetes mellitus in adults: experience in treating diabetic ketoacidosis and hyperosmolar nonketotic coma with low-dose insulin and a uniform treatment regimen. Diabetes Care. 1983;6:579-85. 3. ArieffAI, Carroll HJ. Nonketotic hyperosmolar comawith hyperglycemia: clinical features, pathophysiology, renal function, acid-base balance, plasma-cerebrospinal fluid equilibria and the effects of therapy in 37 cases. Medicine. 1972;51:73-94. 4. Wachtel TJ, Silliman RA, Lamberton P. Predisposing factors for the diabetic hyperosmolar state. Arch Intern Med. 1987; 147:499-501. 5. Fulop M, Tannenbaum H, Dreyer N. Ketotic hyperosmolar coma. Lancet. 1973. 1973;2:635-9. 6. Foster D. Diabetes mellitus. In: Braunwald E, Isselbacher KJ, Petersdorf R, Wilson JD, MartinJB, Fauci A, eds. Harrison's principles of internal medicine, 1 lth ed. New York: McGraw-Hill, 1987; 1778. 7. McCurdy D. Hyperosmolar hyperglycemic nonketotic diabetic coma. Med Clin North Am. 1970;54:683-99. 8. Phillips PA, Rolls BJ, Ledingham JGG, et al. Reduced thirst after water deprivation in healthy elderly men. N Engl J ivied. 1984;311:753-9. 9. GerichJE, PenhosJC, Gutman RA, Recant L. Effect ofdehydration and hyperosmolarity on glucose, free fatty acid and ketone body metabolism in the rat. Diabetes. 1973;22:264-7I. 10. Davidson M. Hyperosmolar nonketotic coma. In: Diabetes mellitus: diagnosis and treatment. 2nd ed. New York: John Wiley & Sons, 1986;265-73. 11. Gerich JE, Martin MM, Recant L. Clinical and metabolic characteristics of hyperosmolar nonketotic coma. Diabetes. 1971; 20:228-38. 12. Podolsky S. Hyperosmolar nonketotic coma. In: Rifldn H, Raskin P, eds. Diabetes mellitus, volume V. Bowie, MD: Robert J. Brady/ Prentice-Hall, 1981;195-201. 13. Faich GA, Fishbein HA, Ellis SE. The epidemiology of diabetic acidosis: a population-based study. Am J Epidemiol. 1983; 117:551-8. 14. Gordon EE, Kabadi EM. The hyperglycemic hyperosmolar syndrome. AmJ Med Sci 1976;271:252-68. 15. Keller U, Berger W, Ritz R, Truog P. Course and prognosis of 86 episodes of diabetic coma. Diabetologia. 1975; 11:93-100. 16. Orland MJ, Saltman RJ. Type II and other diabetes mellitus. In: Manual of medical therapeutics. 25th ed. Boston/Toronto: Little, Brown and Company, 1986;327. 17. Himmelstein DU, Jones AA, Woolhandler S. Hypernatremic dehydration in nursing home patients: an indicator of neglect. J Am Geriatr Soc. 1983;31:466-71. 18. Spagnoli A, Ostino G, Borga AD, et al. Drug compliance and unreported drugs in the elderly. J Am Geriatr Soc. 1989; 7:619-24. 19. Fulop M, RosenblattA, Kreitzer S, Gerstenhaber B. Hyperosmolar nature of diabetic coma. Diabetes. 1975;24:594-9. 20. Sament S, Schwartz MB. Severe diabetic stupor without ketosis. SAfr MedJ. 1957;31:893-4. 21. Wachtel TJ, Silliman RA, Lamberton P. Prognostic factors in the diabetic hyperosmolar state. J Am Geriatr Soc. 1987;35:737-41.

Hyperosmolarity and acidosis in diabetes mellitus: a three-year experience in Rhode Island.

Diabetic acidosis (DA) and the diabetic hyperosmolar state (DHS) are generally considered to be two distinct clinical entities. However, clinical expe...
839KB Sizes 0 Downloads 0 Views