50

Artículo Original

Diabetes y depresión

José Francisco Caro, Juan Rosas Guzmán

Abstract algum grau de depressão, o que favorece um mau controle metabólico. Por outro lado, o diabetes compli- The relationship between diabetes mellitus and depres- cado com manifestações tardias agrava a depressão. sion is not always recognized. About one third of the A depressão pode aumentar até 37% o risco para patients with diabetes have some degree of depres- desenvolver diabetes mellitus tipo 2 quando se com- sion, which favours a poor metabolic control. On the para com quem não tem um estado depressivo. Embora other hand, late complications of diabetes further quando não há um fator etiopatológico identificável, aggravate depression. câmbios em diferentes sistemas hormonais (cortisol, Depression can increase up to 37% the risk for devel- insulina, hormônio de crescimento, etc.) neurotrans- oping mellitus when compared to missores (serotonina) e o sistema nervoso autônomo non-depressed subjects. Even when there is not an ativado jogam um papel relevante. identifiable etiopathologic factor, changes in different O tratamento deve-se realizar tal como corresponde a hormonal systems (cortisol, insulin, growth hormone, cada uma das entidades para conseguir o bem-estar etc.), neurotransmitters (serotonin), and the activated pleno da pessoa. autonomous nervous system play a relevant role. Treatment must be adjusted according to guidelines Introducción established for each entity present to achieve complete well-being of the person. In the past few years, it has been recognized an inter- action between diabetes mellitus and depression. Resumen Even being two separate entities, they have an intimate relationship over their natural history. Prevalence studies La relación entre diabetes mellitus y depresión no siem- have shown that depression favours the development of pre es reconocida. Se sabe que aproximadamente una diabetes mellitus, and that a high percentage of people tercera parte de los pacientes con diabetes tienen with diabetes usually suffer from depression as well. algún grado de depresión, lo que favorece un mal con- That is why we have decided to make an analysis of trol metabólico. Por otro lado, la diabetes complicada the existing evidence that relate diabetes mellitus and con manifestaciones tardías agrava la depresión. depression, answering the following questions: La depresión puede incrementar hasta un 37% el ries- go para desarrollar diabetes mellitus tipo 2 cuando se 1) Is there an association between diabetes compara con quien no tiene un estado depresivo. Aún and depression? cuando no hay un factor etiopatológico identificable, cambios en diferentes sistemas hormonales (cortisol, In the United States, about 7% of the population, or 20 insulina, hormona de crecimiento, etc), neurotrans- million people, have diabetes. Depression affects 10% misores (serotonina) y el sistema nervioso autónomo of the adult population, and women have a twofold activado juegan un papel relevante. higher prevalence of depression than men. One-third of El tratamiento deberá realizarse tal como corresponde people having diabetes are undiagnosed and therefore a cada una de las entidades para lograr el bienestar untreated. This is also a problem in depression. pleno de la persona. In a meta-analysis of 42 published studies including 21,351 adults, the prevalence of major depression in Resumo people having diabetes was 11 % and the prevalence of clinically relevant depression was 31%.1 A relação entre diabetes mellitus e depressão não Overall, studies have shown that individuals with sempre é reconhecida. Sabe-se que aproximadamente depression are more than twice as likely to have dia- uma terceira parte dos pacientes com diabetes têm betes as individuals without depression. Furthermore,

Trabajo recibido el 7/1/2008 y aceptado el 15/1/08 Dirección Postal: Dr. Juan Rosas Guzman. Jefe Unidad de Investigación Centro de Especialidades Médicas de Celay – Mexico. E-mail: [email protected]

VOL. XVI - Nº 2 - Año 2008 José Francisco Caro, Juan Rosas Guzmán 51

there are indications that depressive states are more magnitude to the combined risk of smoking and lack common in diabetes than in other diseases with com- of exercise. parable physical and psychological suffering.2 The and are both risk Although this is the first and simplest question of this factors for the development of diabetes. In a longitu- paper, its answer is not conclusive. The definitions dinal study women with a history of major depressive and methods of diagnosis for both depression and episodes were twice as likely to have the metabolic diabetes are different among studies. There is a pau- syndrome compared with those having no history of city of well-conducted and well-powered prospective depression.9,10 The relationship between depression and rigorously designed studies as was recognized and the metabolic syndrome remained after control- in November of 2002 by The Depression and Bipolar ling for age, race, education, smoking, physical inac- Support Alliance.3 Five years later, the continued lack tivity, carbohydrate consumption and alcohol use. of data and clear awareness of the problem of diabetes Interestingly, men with a history of depression were and depression in the medical and scientific communi- not significantly more likely to develop the metabolic ties is one of the reasons for this work. syndrome.9 The association between obesity and depression is well 2) How serious is the co-morbidity of diabetes recognized.10 However, most studies do not show con- and depression? clusively whether obesity leads to depression or vice versa. It is almost certain that the association works in The burden of co-morbid depression on patients with both directions. However, data from three longitudinal diabetes is profound. There is strong evidence that the studies including more than 10000 subjects indicate coexistence of diabetes and depression is associated that depression in childhood and adolescence may with poor diabetes outcomes. The impact of diabetes lead to adolescence or adult obesity.11 Furthermore, on depression outcomes has received little attention. the National Longitudinal Alcohol Epidemiology Survey In a large meta-analysis that included individual who had found that, among subjects age 18 or older, a BMI of 30 type 1 or type 2 diabetes, depression was associated or greater was positively associated with past depres- significantly with poor control.4 Patients having diabe- sion only in women.12 tes with co morbid depression suffer increased mor- Thus, it appears that depression is a predictor of obes- bidity, mortality, and health care costs, and decreased ity, the metabolic syndrome, and diabetes. However, work productivity and quality of life.5 Unsurprisingly, recognizing the heterogeneity of these conditions, it is much of this morbidity and mortality is related to car- possible that those processes can be bidirectional. diovascular disease. Adoption of appropriate self-care behaviors (regular exercise, dietary intake, home blood 4) Is there a mechanistic link between diabe- glucose monitoring, etc) is an integral part of good dia- tes and depression? betes care, yet very difficult for the depressed person with diabetes to do.5 Depression, obesity, the metabolic syndrome, diabe- We are not aware of any consensus conference or tes, and macrovascular disease are called “complex position statement by the ADA alone or in cooperation diseases”. They are polygenic diseases in which with any of the psychiatric associations focusing on the susceptibility genes are in constant interaction with problem of diabetes and depression. the environment, resulting in expression of a given phenotype. Thus, it is very unlikely that any mecha- 3) What is the temporal relationship between nistic links among these diseases would be simple diabetes and depression? and obvious. An established neuro-endocrine abnormality in Depression is often regarded as a co-morbid condition depression is the hyperactivity of the hypothalamic- that results from the daily burden of having diabetes pituitary-adrenal (HPA) axis. Increased cortisol pro- and its complication. However, it appears that the real- duction in depression may result in insulin resistance, ity is different. decreased insulin secretion, and central obesity, which A meta-analysis of nine longitudinal studies suggest- are premonitory of the metabolic syndrome and dia- ed that adults with depression have a 37% increase betes. Diminished sex steroids and growth hormone in risk of developing type 2 diabetes compared with in depression have metabolic consequences similar those who are not depressed.6 In some of these those of hypercortisolism. Elevated activity of the sym- studies, the risk persisted even after adjusting for pathetic nervous system in depression would result body weight, BMI, physical activity, family history of in elevated blood pressure and also abnormalities in diabetes, socioeconomic status, and education7 and insulin action and secretion. The decreased tone of the is stronger in females.8 Thus, depression appears to central serotonin system in depression may increase by be an independent risk factor for diabetes similar in a high carbohydrate diet and could facilitate the devel- 52 Diabetes y depresión

opment of obesity and the metabolic syndrome. Finally, ing. Accumulation of is the critical the pro-inflammatory state that leads to atherosclero- signal for the initiation of puberty. Considering that life sis is common to depression, obesity, the metabolic expectancy in the ancestral environment was probably syndrome and diabetes.2,13 measured in the teens, the initiation of menarche a The separate pathophysiologies of depression and few months earlier than those that did not acquire the diabetes are reasonably well understood. The changes “behavior shutdown” phenotype would have given them that the two co- morbidities together produce are only an extraordinary evolutionary advantage: increased inferred and not determined experimentally. There are fertility by a significant percentage of their brief life no longitudinal studies, observational or interventional, span. Today, given the environment of plenitude, those wherein any of the biomarkers of the two diseases have individual may have obesity, the metabolic syndrome, been studied together. or diabetes.

5) Why is depression so prevalent among peo- 6) Can diabetes and depression be treated? ple with diabetes? There are at least three hypotheses to answer this To the evolutionary biologist, human beings at every question: stage of their development represent “compromises” a) “Treatment of depression in people with diabetes in their continual adaptation to their changing environ- is possible and, without any additional interven- ment.14 Recent developments in evolutionary theory tion in diabetes management, will improve dia- suggest that negative emotions and depression may be betes control”. This hypothesis has been tested evolved strategies that facilitated behavior solutions to by two randomized controlled clinical trials. 19,20 problems in the ancestral environment.15 They demonstrated that depression can be Is depression a disease, or a “behavior shutdown treated in people with diabetes with remission mechanism”?16 Major Depression can be thought of rates similar to those without diabetes. However, in term of mismatch between modern and ancestral improvement of depression was not followed by environments. Although many of the characteristics any changes in diabetes outcomes.19,20 of the ancestral hunter-gatherer environment remain open to speculation, one can be quite certain that our The IMPACT (Improving Mood Promoting Access to hominid ancestors used to be much more vulnerable Collaborative Treatment) was a randomized clinical and much less in control of their environment than we trial which included 1801 patients 60 years of age or currently are. Famine, warfare, drought, disease and older with depression of whom 417 had coexistent infection, extreme weather fluctuations, high predation, diabetes. Eighteen primary care clinics from 8 health authoritarian leaders etc. were likely frequent problems care organizations in 5 states participated. Two hun- that the individual had little or no control over. As such, dred and twelve patients with both co morbidities one can be quite certain that behavioral activity was were randomized to the usual care and 205 formed much more dangerous then than now. Because of the intervention group. In the intervention group a care lesser control and greater danger, the propensity to manager offered education, problem-solving treat- shut down in unpropitious situations may very well have ment, or support for antidepressant management by been adaptive in ancestral environment. This evolution- the patient’s primary care physician; diabetes care was ary tendency toward depression may well be quite a not specifically enhanced. Outcome measurements disadvantage in modern cultures”.17 were obtained at baseline, 3, 6, and 12 months for the In 1962, the thrifty gene hypothesis was proposed to degree of depression, functional impairment, diabetes explain the tendency of certain ethnic groups toward self-management behaviors and Hemoglobin A1c. At obesity.18 It postulates that certain genes in humans 12 months, patients with diabetes who were assigned have evolved to maximize metabolic efficiency to sur- to the intervention group had less depression but dia- vive famine and that in times of abundance these genes betes outcomes remained unchanged.19 predispose their carrier to diseases caused by excess The PATHWAYS study was a randomized trial of col- nutritional intake, such as obesity. laborative care in patients with diabetes and depres- Is it possible to unify the behavior shut-down hypoth- sion. It included 329 patients, 12 months follow up and esis16 with the thrifty gene hypothesis?18 Studies have the setting, intervention, main outcome measures and shown that depression is more common in women. results ere similar to the IMPACT trial.20 Depression appears to precede the development of obesity, the metabolic syndrome and diabetes, a rela- b) The second hypothesis - “Active management tionship that is clearer in women. Thus, it is possible of both depression and diabetes is neces- that the phenotype of one of the putative thrifty genes sary to improve diabetes outcome in people is the shut-down behavior which is energy conserv- with co-morbid diabetes and depression”. At

VOL. XVI - Nº 2 - Año 2008 José Francisco Caro, Juan Rosas Guzmán 53

least five studies have tested this hypothesis. likely that work in this field will be highly productive and Unfortunately, most of these studies were not useful in improving the outcome of these patients. sufficiently powered. Thus, it is appropriate to conclude that this hypothesis has not been 6) Which should be the adequate treatment? adequately tested. Since there are no studies that show effective thera- Three randomized, placebo-controlled, double-blind peutic guidelines for diabetic patients with depression, trials in people with diabetes and depression using we recommend a treatment focused to achieve a good Nortriptyline (14 placebo, 14 active, 28 weeks);21 control of the diabetes as well as accomplishing an Sertraline (73 placebo, 78 active, 52 weeks);22 and improvement of the person’s state of feeling. Fluoxetine23 (27 placebo, 27 active, 8 weeks) demon- Treatment of diabetes mellitus should take the patient strated statistically significant improvement in depres- to control goals in glycated haemoglobin (A1c), fasting sion but not in glucose control. glucose and postprandial glycaemia as recommended An open label observational study without a controlled by the diagnosis and treatment guidelines by the Latin- arm showed improvement of both depression and glu- American Diabetes Association (ALAD) 28. cose control by Bupropion.24 Given the nature of the Even more, we should detect the patients on the study design, the conclusion of this clinical observation impaired glucose tolerance phase seeking delay or is not clear. prevent the development of type 2 diabetes mellitus. A randomized, controlled trial using cognitive behavior People with risk factors for diabetes, especially obesity therapy (CBT) (25 CBT, 26 control, 10 weeks) dem- and a sedentary lifestyle can be benefited with an ade- onstrated post treatment improvement of depression quate treatment.29 Correction of obesity and regular only, but statistically significant improvement of both physical activity can also prevent a depressive state. depression and glucose control 6 months thereafter.25 The patient that has had diabetes for over 5 years who However, as indicated by the investigators, the CBT has not had a good metabolic control will be exposed group had patient education almost a full year longer to develop micro and macrovascular complications. than controls. The difference in education was not Retinopathy, nephropathy, amputations, etc. will be statistically significant, but the extra educational expe- factors that complement in a meaningful way the devel- rience may have contributed to improved outcome in opment or aggravation of the depressive state. The the CBT group.25 secondary and tertiary prevention of diabetes mellitus is of capital importance.30 c) The final hypothesis- “Optimal management Treatment of depression using from behavioural tech- of diabetes in people with depression is pos- niques and/or drugs is justified. sible and, without any additional intervention in The option that has proven to be most effective is the depression management, will improve both co- one that includes psychotherapy and antidepressant morbidities”. This counter-intuitive hypothesis drugs, both in achieving a fast response as in main- has not been tested. taining the improvement 19. The first visits must be frequent, giving psychotherapy support, while the first The previous studies do not provide clear guideline for clinic results are observed, which should be within 4 to the treatment of diabetes and depression to the prac- 6 weeks. ticing physician. The failure to demonstrate efficacy The selection of the drug must be individualized and in diabetes outcomes with any given approach could consider for its prescription the medication’s profile, be due to several reasons: The optimal experimental possible side effects, drug interactions, price, dose design has not been used; The pathogenic relation- per day and safety. There is no such thing as the ideal ship between diabetes and depression speculated antidepressant, and two thirds of the patients respond previously is incorrect; The complexity of the two co- well to first and second generation medications. morbidities may require the treatment of a putative The premises will be: Solving depression, we will have third component which is elusive at this time; A new a person with less risk for diabetes. A person with dia- hypothesis need to be formulated and novel treatment betes and not depressed, will be willing to undergo an approaches may need to be tested. 26,27 optimal treatment of his/her disease and more easily Given the serious and often unmet medical needs of achieve the control goals which will avoid complica- people with co-morbid depression and diabetes, it is tions and will not affect his/her life expectancy. 54 Diabetes y depresión

Referencias 1. Anderson RJ, Freedland K E, Cluse R E, et al. The 102. prevalence of comorbid depression in adults with diabetes: 16. Henriques G. Depression: disease or behavioral shut- a meta- analysis. Diabetes Care 2001;24(6): 1069-78. down mechanism. Journal of Science and Health Policy. 2. Björntorp P. Epidemiology of the relationship between 17. Keller MC, Nesse RM. Is low mood an adaptation? depression and physical illness. Physical Consequences Evidence for subtypes with symptoms that match precipi- of Depression. 2001:67-85. tants. Journal of Affective Disorders, 2005;86:27-35. 3. Evans DL, et al. Mood disorders in the medically III: 18. Neel JV. Diabetes mellitus: a “thrifty” genotype ren- scientific review and recommendations. Biol Psychiatry, dered detrimental by “progress”? Am J Hum Genet, 2005;58:175-189. 1962;14:353-62. PMID 13937884. 4. Lustman PJ, Anderson RJ, Freedland KE, et al. Depression 19. Williams JW Jr., Katon W, Lin EHB, et al. The effective- and poor glycemic control: a meta-analytic review of the lit- ness of depression care management on diabetes-related erature. Diabetes Care 2000;23(7):934-42. outcomes in older patients. Annals of Internal Medicine, 5. Egede LE. Disease-focused or integrated treatment: dia- 2004;140(12):1015-101024. betes and depression. Medical Clinics of North America, 20. Lin EHB, Katon W, Rutter C, et al. Effects of enhanced 2006;90 (4):627-646. depression treatment on diabetes self-care. Annals of 6. Knol MJ, Twisk JWR, Beekman ATF, Heine RJ, Snoek Family Medicine, January/February 2006;4(1):46-53. FJ, Pouwer F. Depression as a risk factor for the onset of 21. Lustman PJ, Griffith LS, Clouse RE, et al. Effects of type 2 diabetes mellitus. A meta-analysis. Diabetologia, nortripltyline on depression and glycemic control in dia- 2006;49:837-845. betes: results of a double-blind, placebo-controlled trial. 7. Golden SH, Williams JE, Ford DF, Yeh HC, Sanford CP, Psychosomatic Medicine, 1997;59:241-250. Nieto FJ, Brancati FL, Depressive symptoms and the 22. Lustman PJ, Clouse RE, Nix BD, et al. Sertraline for risk of type 2 diabetes: The atherosclerosis risk in com- prevention of depression recurrence in diabetes mellitus. munities study. Diabetes Care, 2004;27:429-435. Arch Gen Psychiatry, 2006;63:521-529. 8. Zhao W, Chen Y, Lin M, Sigal RJ. Association between 23. Lustman PJ, Freedland KE, Griffith LS, Clouse RE. diabetes and depression: sex and age differences. Public Fluoxetine for depression in diabetes. Diabetes Care, Health, 2006;120:696-704. May 2000;23(5):618-623. 9. Kinder LS, Carnethon MR, Palaniappan LP, King AC, 24. Lustman PJ, Williams MM, Sayuk GS, Nix BD, Clouse Fortmann SP. Depression and the metabolic syn- RE. Factors influencing glycemic control in type 2 dia- drome in young adults: findings from the third national betes during acute- and maintenance-phase treatment health and nutrition examination survey. Psychosomatic of major depressive disorder with bupropion. Diabetes Medicine, 2004;66L316-322. Care, March 2007;30(3):459-465. 10. Onyike CU, Crum RM, Lee HB, Lyketsos CG, Eaton 25. Lustman PJ, Griffith LS, Freedland KE, Kissel SS, Clouse WW. Is obesity associated with major depression? RE. Cognitive behavior therapy for drepression in type 2 Results from the third national health and nutrition exami- diabetes mellitus. Ann Intern Med., 1998;129L613-621. nation survey. Am J Epidemiol, 2003;158L1139-1147. 26. Jacobson AM, Weinger K. Treating depression in diabetic 11. The child advocate: childhood depression and adult patients: is there an alternative to medications? Annals of obesity. Depression and Obesity, www.childadvocate. Internal Medicine, October 15, 1998;129(8):656-657. net/depression_and_obesity.htm. 27. Anderson D, Horton C, O’Toole ML, et al. Integrating 12. Carnethon, Mercedes R.; Kinder, Leslie S.; Fair, depression care with diabetes care in real-world settings: Joan M.; Stafford, Randall S.; Fortmann, Stephen P. lessons from the Robert wood Johnson foundation diabe- Symptoms of Depression as a Risk Factor for Incident tes initiative. Diabetes Spectrum, 2007;20(1):10-16. Diabetes: Findings from the National Health and Nutrition 28. Ascher P, Rosas GJ, López G et al. Guías ALAD de Examination Epidemiologic Follow-up Study, 1971-1992. diagnóstico, control y tratamiento de la Diabetes Mellitus American Journal of Epidemiology. 158(5):416-423, tipo 2. Rev. De la Asoc. Latinoamericana de Diabetes September 2003 2006; Vol XIV No. 3: 100- 134. 13. Gans ROB. The metabolic syndrome, depression, and 29. Lara A, Rosas GJ, et al. Recomendaciones para la pro- cardiovascular disease: interrelated conditions that share moción de la salud, prevención, detección, diagnóstico, pathophysiologic mechanisms. Medical Clinics of North tratamiento y control de la . Diabetes Hoy America, 2006;90 (4):572-591. 2006,. Vol VII No. 6:1708-1748 14. Nesse RM. Is depression an adaptation? Arch Gen 30. Rosas GJ, Gonzalez A., et al. Consenso para la preven- Psychiatry, 2000;57:14-20. ción de complicaciones crónica de la diabetes mellitus 15. Nettle D. Evolutionary origins of depression: a review and tipo 2. Annual Review del Colegio de Medicina Interna de reformulation. Journal of Affective Disorders, 2004;81:91- México. 2005.: 225-284. Intersistemas eds.

VOL. XVI - Nº 2 - Año 2008