Women with insulin-dependent diabetes mellitus (IDDM) complicated by eating disorders are at risk for exacerbated alterations in lipid metabolism

European Journal of Clinical Nutrition (1997) 51, 462±466 ß 1997 Stockton Press. All rights reserved 0954±3007/97 $12.00 Women with insulin-dependent diabetes mellitus (IDDM) complicated by eating disorders are at risk for exacerbated alterations in lipid metabolism SG Affenito1,4, CJ Lammi-Keefe1, S Vogel1,4, JR Backstrand1,4, GW Welch2 and CH Adams3 1 Department of Nutritional Sciences, University of Connecticut, Storrs, Connecticut 06269; 2Mental Health Unit, Joslin Diabetes Centre and Department of Psychiatry, Harvard Medical School, Boston, MA 02215; and 3School of Allied Health, University of Connecticut, Storrs, Connecticut 06269, USA Objective: To examine lipid parameters that are affected in women with insulin-dependent diabetes mellitus (IDDM) who engaged in disordered eating behaviours. Design: Randomized, unmatched. Setting: Tertiary care. Subjects: Ninety women (18±46 y) with IDDM. Interventions: Classi®cation of subjects based on severity of eating disorder: clinical (n ˆ 14), subclinical (n ˆ 13) and control (n ˆ 63). Blood was analysed for glycosylated haemoglobin (HbA1c) and serum for triglycerides and cholesterol. Carotenoid and tocopherol concentrations were analysed by high performance liquid chromatography (HPLC). Dietary intake was assessed by the National Cancer Institute food frequency questionnaire. Results: HbA1c was signi®cantly increased im women demonstrating clinical and subclinical symptoms compared to control (10.4  2.6, 10.0  1.5 and 8.3  1.6%, respectively, P < 0.05). Triglyceride concentrations were signi®cantly increased in women with subclinical eating disorders compared to controls. In women who intentionally omitted or reduced insulin, triglyceride cholesterol and HbA1c were signi®cantly increased compared to controls. Women with IDDM and eating disorders who exhibited bulimic behaviours consumed signi®cantly more energy, total fat and cholesterol compared to controls and women with eating disorders who were restrained eaters. Conclusion: While IDDM is known to perturb lipid metabolism, these data demonstrate that eating disorders, in combination with IDDM, results in additional alterations in lipid metabolism. Sponsorship: Supported in part by the Storrs Agricultural Experiment Station, University of Connecticut Research Foundation and an NIH General Clinical Research Grant from the University of Connecticut Health Centre (M01RR06192). Descriptors: lipid metabolism, a-tocopherol; carotenoid; glycosylated haemoglobin; insulin-dependent diabetes mellitus; eating disorders; bulimia nervosa; anorexia nervosa Introduction Adolescent girls and young adult females with insulindependent diabetes mellitus (IDDM) are at signi®cant risk for the development of the classic eating disorders, anorexia nervosa and bulimia nervosa (Rodin et al, 1985; Steel et al, 1989; Hudson et al, 1985). Moreover, eating disorders are frequently accompanied by omission or reduction of insulin to induce glycosuria, thereby promoting weight loss (Birk & Spencer, 1989; Polonsky et al, 1994). Eating disordered behaviours and insulin misuse in women with IDDM may result in increased chronic complications of diabetes secondary to poor glycemic control (Polonsky et al, 1994; Colas et al, 1991; Steel et al, 1987). Poor glycemic control in individuals with IDDM (Glasgow et al, 1981; Lopez-Virella et al, 1982; Howard, 1987) Correspondence: 4Present Address: Dr SG Affenito, School of Allied Health, University of Connecticut, Storrs, Connecticut 0629 USA. Received 18 November 1996; revised 15 March 1997; accepted 28 March 1977 accelerates existing lipid perturbations (Mooradian & Moorley, 1987). Elevated plasma concentrations of triglyceride, cholesterol, very low density lipoproteins (VLDL) and low density lipoproteins (LDL) have been reported in individuals with IDDM in poor glycemic control (Glasgow et al, 1981). Additionally, plasma lipid peroxides have been found to be higher in individuals whose diabetes is not well controlled vs good control (Jain et al, 1989). This metabolic alteration may ultimately lead to increased long-term complications (Wolff, 1987). Lipid metabolism is also disturbed in people with eating disorders. Signi®cantly higher plasma total cholesterol concentrations were reported in individuals with anorexia nervosa and bulimia nervosa as compared to age and weight matched controls (Mira et al, 1989; Jaffe et al, 1987; Vize & Coke, 1994). Hypercholesterolemia was also more common in anorexia nervosa (Mira et al, 1987), which may be attributed to amenorrhoea (Nestel, 1988) or impaired clearance of LDL cholesterol (Mordasini et al, 1978). Lipid metabolism in women with IDDM and eating disorders SG Affenito et al Disturbances in the status of fat soluble antioxidant vitamins has been reported in people with eating disorders (Mira et al, 1989; Philipp et al, 1988; Langan & Farrell, 1985; VanBinsbergen et al, 1988). Hypercarotenemia in these individuals has been attributed to reduced clearance and degradation of LDL (Rock & Curran-Celetano, 1994). Alternatively, such alterations may be due to an acquired error in metabolism, enhanced absorption, and reduced tissue storage capacity (Rock & Curran-Celetano, 1994; Rock & Swendseid, 1993). The lipid status of women with IDDM complicated by co-existing eating disorders has heretofore not been investigated. The purpose of the current research was to examine lipid parameters that are affected in IDDM, as well as in eating disorders. Serum concentrations of triglyceride, cholesterol and total lipids, as well as plasma concentrations of vitamin E, retinol and carotenoids were assessed. Secondly, women from the subclinical and clinical groups were collapsed into one group for the purpose of analysis (n ˆ 27). Thirdly, women who misused insulin as a method of purging calories were identi®ed (n ˆ 12). Lastly, women with eating disorders, whether subclinical or clinical, were grouped as having bulimic (n ˆ 16) or not having bulimic behaviours (n ˆ 11) for dietary assessment purposes. Sample collections Blood was collected from a large antecubital vein into tubes containing ethylenediaminetetraacetic acid for whole blood analyses and antioxidant analyses and into a serum separation tube for serum analyses. Plasma and serum were separated by centrifugation (15006g for 15 min at 25 C). Aliquots were stored at 780 C. For analyses, frozen plasma samples were thawed at room temperature. All analyses were performed within one year of collection. Methods Subjects Subjects were women who had a history of IDDM for at least one year, who were aged 18±46 y, not pregnant or lactating and who were otherwise healthy. Recruitment occurred at diabetes clinics throughout Connecticut and Massachusetts. Ninety recruits were eligible to participate in the study. Ethical approval for this study was granted by the University of Connecticut Human Subjects Approval Committee and clinical facility institutional review boards. W (...truncated)