Hypothesis 5. Treatment of depression may reduce the risks of diabetes and its complications.

Lustman PJ, Griffith LS, Freedland KE, Kissel SS, Clouse RE: Cognitive behavior therapy for depression in type 2 diabetes: a randomized, controlled trial. Ann Intern Med 129: 613–621, 1998

The preceding commentaries have reported evidence for the negative effects of depression on those with diabetes, e.g., poorer glycemic control and increased complications, as well as an elevated risk for development of diabetes in depressed subjects. Therefore, it is of great interest to know whether the risks conferred by depression can be modified if depression is successfully treated. Given the association between the above risks, lifestyle choices, and behaviors, it is likely that depression treatment that targets these factors may be particularly helpful, especially over the long term.

Cognitive behavior therapy (CBT) specifically addresses mistaken,maladaptive, or ruminative cognitions about one's self, one's relationships,and one's present and future circumstances. More accurate and helpful cognitive responses are formulated, and behavior is addressed in terms of adaptive problem-solving, improving relationships, and increasing enjoyable as well as productive activities.^46,47 CBT has been shown to improve depression in those with and without medical illness in numerous studies.^48–50

The first evidence that depression in patients with type 2 diabetes can be successfully treated by CBT was provided by this randomized, controlled trial reported by Lustman et al. For this study, 51 patients with type 2 diabetes and comorbid major depression were randomized to receive biweekly diabetes education either alone or in combination with 10 weekly CBT sessions for depression. Beck Depression Inventory score and glycated hemoglobin level were measured at the end of treatment and again 6 months after treatment. The dropout rate was low, and a completer analysis was used.

The percentage of patients achieving remission of depression was significantly higher in the CBT group than in the control group at both posttreatment evaluation points (85.0 vs. 27.3% at post-treatment, P< 0.001; 70.0 vs. 33.3% at follow-up, P = 0.03). Effects on glycemic control were particularly interesting. At posttreatment, glycated hemoglobin levels did not differ between groups. However, at the 6-month follow-up, mean glycated hemoglobin levels were significantly better in the CBT group than in the control group (9.5 vs. 10.9%, respectively, P =0.03). From post-treatment to follow-up, mean glycated hemoglobin levels fell in the CBT group and rose in the control group.

The Lustman group studied a small cohort of very well-characterized patients with type 2 diabetes. The diagnosis of both disorders was rigorous,using well-established standards of the American Diabetes Association, with the addition of a physician report in the case of diabetes and both interview and self-report questionnaires in the case of depression. Unfortunately, the cohort was very small, perhaps partially because of the strict exclusion criteria.

The authors did not explain their decision to exclude those with history of suicide attempt (no matter how remote), history of panic disorder, or those on any psychoactive drugs—medications that have become increasingly commonplace in medical practice whether they prove to be helpful or not. Also,it is known that drugs in many classes have psychoactive effects in addition to their intended treatment effects, so the notion that one can recruit a“clean” cohort of depressed individuals among a medical population is suspect.

In order to control for the nonspecific effects of supportive attention and information about improving glucose control, subjects in both groups were given the same diabetes education program. Although this strategy solves one potential problem, it creates another. If amount of contact with a knowledgeable and supportive professional is related to improvement in mood,then the increased contact time (10 hours) for the treatment group may explain some or all of the effects of treatment, rather than the type of treatment utilized.

It is helpful in this regard that glycated hemoglobin levels and depressive symptoms were measured at repeated intervals, because the differences in the timing of effects in the different areas provide support for the conclusion that the therapy itself is the best explanation for the results. If increased contact fully explained the improvement in mood, then the differences between the groups would not be expected to remain so wide 6 months later.

Additionally, the somewhat worse performance by the CBT group in terms of short-term glycated hemoglobin and glucose monitoring compliance suggest that the additional office visits and the work of therapy were temporarily detrimental to improving compliance behaviors. Thus, when one considers treatment for depression to lower both risk of development and incidence of complications, it is important to keep potential unintended treatment effects in mind in devising a treatment strategy.

The Lustman et al. study provided the first support for the efficacy of CBT for the treatment of depression in patients with type 2 diabetes. The findings suggest that a therapy that seeks to decrease maladaptive responses and improve skills may have far-reaching effects. In the research setting, when testing the hypothesis that treatment of depression leads to medical benefit,education and skill development that directly address the illness under question, e.g., making daily glucose monitoring a behavioral homework assignment, is avoided to limit confounding the effects of depression remission with “medical” intervention. Such limitations make it remarkable that medical endpoints improved more for treatment patients. Outside of research, such behaviors would be an important part of treatment for medical patients seen by a health psychologist practicing CBT. Thus, it is likely that patients in a community setting seen by an experienced and knowledgeable health psychologist would receive even greater benefit from treatment.

In 2001, the Psychosocial Therapies Working Group published its conclusions that psychosocial therapies have demonstrated efficacy in improving adherence,glycemic control, functioning, and quality of life for people with diabetes.⁵¹ More recently, the Atherosclerosis Risk in Communities study group reported that depressive symptoms predicted incident cases of type 2 diabetes.¹¹ The relationship was partially explained by lifestyle factors, such as increased body mass index, higher rates of smoking, physical inactivity, and caloric intake.

Symptoms of depression, such as loss of motivation, decreased energy,hopelessness, sleep and appetite disturbances, feelings of worthlessness, and thoughts of death and/or suicide all adversely affect the behaviors required to maintain a healthy lifestyle and decrease one's chances of developing diabetes. Such symptoms may have an even greater effect on those with the disease, because of the requirements for active self-monitoring, adjustments to the timing and dose of medications, and changing of eating and exercise habits that may be long-standing. Conversely, many of the changes those with and at risk for diabetes are asked to make may be quite disruptive to their lives socially and financially in terms of their work, relationships, and recreation. Such disruptions put vulnerable individuals at increased risk to develop depression as a consequence of the illness.

Large-scale psychosocial interventions to decrease the incidence of depression and encourage healthy choices should favorably affect the incidence rate of type 2 diabetes. Cognitive and social skills training could be taught throughout the educational process to work toward the achievement of both goals. Pilot programs already exist with the goal of decreasing violence among children and adolescents. Depression resistance or “hardiness training” would be roughly similar, and it could easily include the added benefit of fostering a health-oriented lifestyle.

For those who have diabetes, group CBT has been shown to be helpful and sufficient for many depressed individuals and cost-efficient as well. A team approach is already commonly utilized in the treatment of diabetes, and the addition of a clinical therapist with expertise in medical populations is already the norm in the treatment of cancer and many neurological diseases and conditions. For those individuals whose depression severity or specific circumstances require individual therapy, it is likely that treatment focused on improving one's mood, motivation, energy level, and ability to maintain focus and think clearly would greatly benefit control over and adjustment to illness, in addition to improving quality of life in general. Improved glucose control has been shown to lead to fewer complications, making an intervention that leads to such an end cost-efficient.

Even if depression management did not have direct effects on disease outcome, it might reduce resource utilization that stresses the medical system. Moreover, treatment of depression to help people with diabetes make a healthy adjustment and lead more enjoyable, fulfilling, and productive lives is a worthy goal in and of itself.