Too much of a good thing? Overexertion of self-control and dietary adherence in individuals with Type 2 diabetes.

Objectives: The resource model of self-control posits that self-control is a finite resource that can be depleted. Individuals with diabetes must continually restrict their diet, requiring self-control. As a result, dietary adherence is difficult, and lapses are common. People with diabetes who overexert self-control following a lapse may be especially likely to experience a subsequent relapse, as suggested by the resource model. This investigation used the resource model of self-control to test whether overexertion of dietary self-control following a lapse would be predictive of a subsequent relapse in dietary control.

Design: We tested this prediction in a daily diary study of 128 individuals with diabetes (Mage  = 66.12).

Methods: Participants' reports of their daily dietary adherence were used to define lapses in adherence, post-lapse adherence, and relapses.

Results: Individuals who overexerted self-control after a lapse were more likely to experience a subsequent relapse (OR = 3.276, p = .016) and to do so sooner (HR = 2.12, p = .023).

Conclusions: People with diabetes may seek to compensate for a lapse in adherence by overexerting self-control, but doing so may deplete their self-control and increase the risk of a future relapse. Statement of contribution What is already known on this subject? The resource model of self-control posits that self-control is a limited resource that can be temporarily depleted. Numerous experimental studies have demonstrated support for this model showing that when participants are instructed to engage in a self-control task, they produce less self-control on a subsequent task. The majority of the existing studies are not conducted in naturalistic settings and do not use patient populations. What does this study add? This study is an ecologically valid test of the resource model of self-control. This study applies the resource model of self-control to a patient population.