196 Diabetes Mellitus
Both human (Gonder-Frederick et al., 1990) and animal
(Lee, Konarska, & McCarty, 1989) studies have demonstrated
that stress has idiosyncratic effects on BG, which are manifest
in two ways: Different stressors may have different effects on
BG, and different individuals may respond to the same stres-
sor in different ways. Further, these individual response dif-
ferences appear to be stable over time (Gonder-Frederick
et al., 1990). This line of research has prompted an explo-
ration into the role of individual differences. Stabler et al.
(1987) found that the glucose response to experimental stress
was related to a Type A behavior pattern, but this “nding has
not been replicated in other studies (Aikens, Wallander, Bell,
& McNorton, 1994).
Life events have also been implicated in glycemic control
and symptomatology (Lloyd et al., 1999), although the associ-
ation tends to be weak (Cox et al., 1984). In contrast with major
life events, the role of daily stress variability has been shown
to provide more convincing data on a link between stress and
somatic health (Aikens, Wallander, Bell, & Cole, 1992).
Because relaxation techniques have been shown to de-
crease adrenocortical activity (DeGood & Redgate, 1982)
and circulating levels of catecholamines (Mathew, Ho,
Kralik, Taylor, & Claghorn, 1980), this intervention has been
proposed as a means of moderating the negative effects of the
stress response on glycemic control in diabetes. Relaxation
interventions with type 1 patients have produced mixed re-
sults (e.g., Feinglos, Hastedt, & Surwit, 1987; McGrady,
Bailey, & Good, 1991). This may be caused by heteroge-
neous glucose responses to stress in type 1 diabetes and/or
more labile glycemic control resulting from diet, insulin,
exercise, and illness (Feinglos et al., 1987).
Alternatively or concurrently, stress may also relate to di-
abetes management through the indirect effects on treatment
adherence (Peyrot & McMurray, 1985). This is particularly
relevant to individuals with type 1 diabetes or those requiring
insulin, since self-management in these patients is more com-
plex. Stress can disrupt self-care by promoting inappropriate
behaviors (e.g., drinking alcohol, binge eating) or by upset-
ting normal routine behaviors (Cox & Gonder-Frederick,
1991).
Finally, BG ”uctuations can indirectly af fect stress levels
through neuroendocrine changes that are subjectively per-
ceived as stress or mood states (Grandinetti, Kaholokula, &
Chang, 2000). At extreme BG levels, mental confusion,
disorientation, and coma can result. Diabetes is the leading
cause of adult blindness, lower extremity amputations, kid-
ney disease, and impotence (Glasgow, Fisher, et al., 1999).
Thus, glucose may also be responsible for indirectly inducing
stress secondary to the requirement for aversive therapeutic
interventions (Bernbaum, Albert, & Duckro, 1988).
A modest literature has developed over the past 20 years
on the effects of stress on control of type 2 diabetes. Studies
have demonstrated a relationship between life events and
diabetic symptomatology, although the association is some-
times weak (Grant, Kyle, Teichman, & Mendels, 1974) or ab-
sent (Inui et al., 1998). To explain the con”icting results,
Bradley (1979) suggested that type 2 patients may have some
degree of endogenous homeostatic control of their glucose
levels, making them less likely to experience disruption in re-
sponse to stress.
Physical stressors, such as elective surgery and anesthesia
(McClesky, Lewis, & Woodruff, 1978), as well as laboratory
stressors (Goetsch, Wiebe, Veltum, & Van Dorsten, 1990), af-
fect BG. Although the mechanisms for the metabolic re-
sponse to stress in type 2 diabetes are unknown, there is some
evidence for an altered adrenergic sensitivity and responsiv-
ity in type 2 diabetic humans and animal models, as sup-
ported by studies examining the role of alpha-adrenergic
blockades in altering glucose-stimulated insulin secretion
(e.g., Kashiwagi et al., 1986).
Some researchers propose that environmental stress,
which activates the sympathetic nervous system, may be
particularly deleterious to patients with type 2; therefore,
methods to reduce the effects of stress are believed to have
some clinical utility in this disease (Surwit et al., 1992).
With some exceptions (Lane, McCaskill, Ross, Feinglos, &
Surwit, 1993), well-controlled group studies have demon-
strated that relaxation training can have a signi“cant positive
impact on BG level or range with type 2 patients (Lammers,
Naliboff, & Straatmeyer, 1984; Surwit & Feinglos, 1983).
There is also evidence that anxiolytic pharmacotherapy ef-
fectively attenuates the effects of stress on hyperglycemia
in animals (Surwit & Williams, 1996) and humans (Surwit,
McCasKill, Ross, & Feinglos, 1991).SummarySpeculation regarding the role of stress in the development
and course of diabetes has continued for more than 300 years.
Only limited evidence supports the notion that stress is in-
volved in the onset of type 1 diabetes. The literature on the
effects of stress on the course of type 1 diabetes in experimen-
tal and clinical settings is complicated by a variety of method-
ological limitations and issues. Importantly, less than half of
individuals with type 1 diabetes may manifest a relationship
between stress and BG control (Kramer et al., 2000), and in-
dividuals who are •stress reactorsŽmay respond idiosyncrati-
cally (Goetsch et al., 1990; Riazi, Pickup, & Bradley, 1996).
Evidence that stress reduction strategies are effective in type 1
diabetes is limited and inconclusive. The literature on the