Handbook of Psychology

Psychosocial Factors in Diabetes Management 193

Several measurement considerations limit the study of
adherence and its relationship to health outcomes. McNabb
(1997) and Johnson (1992) provide excellent reviews of
these methodological dif“culties in adherence research. The
“rst dif “culty is in de“ning the set of behaviors involved in
the treatment regimen because of the wide variability in types
and intensities of treatment regimens, the lack of explicit
recommendations in medical charts, and/or the inability of
patients to recall recommendations. In addition, adherence to
one aspect of the regimen is relatively independent of adher-
ence to other aspects of the regimen (Glasgow, McCaul, &
Schafer, 1987), with adherence to medications the highest
while adherence to behaviors necessitating greater lifestyle
change (e.g., diet, exercise) is lower (Johnson, 1992). Thus,
global rating systems and judgments of patients as adherent
versusnonadherentare inappropriate.
As reviewed by Johnson (1992), methods used to evaluate
diabetes patients• adherence levels include physiological
outcomes (e.g., GHb), physician ratings, collateral reports,
measurement of permanent products (e.g., number of pills
consumed, data stored in memory BG meters), and patient
self-reports. There is no widely accepted, reliable measure of
adherence or approach to quantifying the level of adherence
at present (McNabb, 1997). Each method of assessment has
its advantages as well as its limitations. Despite reliability
and validity concerns, self-report measures are the most com-
monly used measures of adherence. A variety of psychomet-
rically sound questionnaires (e.g., the Summary of Diabetes
Self-Care Activities; Toobert, Hampson, & Glasgow, 2000);
self-monitoring diaries (e.g., Glasgow et al., 1987); and inter-
views (e.g., 24 Hour Recall Interview; Johnson, Silverstein,
Rosenbloom, Carter, & Cunningham, 1986) have been devel-
oped. Given the dif“culties in each of the measurement meth-
ods, Johnson and McNabb recommend selecting instruments
carefully, using a multicomponent measurement strategy,
and measuring adherence across time and within a time pe-
riod consistent with other measures of constructs to which the
researcher is seeking to relate adherence. Once measured,
however, decisions about how to evaluate the obtained ad-
herence levels must be made. Without a known standard of
adherence, researchers and practitioners are left without clear
guidelines for qualifying levels of behavior that fall below
this elusive standard (McNabb, 1997).
Adherence as a construct is important because of its pre-
sumed link with glycemic control and thus indirectly its link
to diabetes complications. Despite the clear logic of this rela-
tionship, research has been inconsistent in its ability to “nd a
direct link between patient adherence and metabolic control
in diabetes. This may be because of the multidetermined na-
ture of glycemic control, the limitations of GHb as a measure

of glycemia, methodological problems in adherence mea-

surement and analysis, and the potential for an idiosyncratic

effect of adherence on glycemic control between individuals

(Johnson, 1992; McNabb, 1997).

PSYCHOSOCIAL FACTORS

IN DIABETES MANAGEMENT

Despite the dif“culties in its conceptualization, accurate mea-

surement, interpretation, and relationship to glycemic con-

trol, adherence continues to be the focus of research efforts

and clinical interventions. Research, reviewed next, has

sought to (a) identify the factors associated with either the

promotion or suppression of adherence levels and (b) de-

velop effective interventions to enhance adherence levels and

subsequent health outcomes. This chapter focuses on six such

variables: patient knowledge, stress, depression, social sup-

port, patient practitioner relationships, and perceived barriers

and coping styles. The selection of these six factors was

based on the amount of research conducted with the variable

as the focus, the availability of empirically tested interven-

tions focusing on the factor, and clinical relevancy.

Knowledge

The increasing complexity of the diabetes regimen and

emphasis on self-management (M. Williams, Baker, Parker,

& Nurss, 1998) has placed higher educational demands on

patients (ADA, 1996). Education may facilitate patients• ac-

ceptance of their diagnosis, engagement in the behavioral

changes necessary for their active participation (M. Williams

et al., 1998), and ability to lead normal, productive lives

(Garrard et al., 1987).

Diabetes Education Programs

In the late 1970s, diabetes education programs were initiated

to ensure that patients had suf“cient knowledge and under-

standing of their disease (Beeney, Dunn, & Welch, 1994).

The need to evaluate these programs led to the development

of tests of diabetes knowledge (e.g., Garrard et al., 1987;

Hess & Davis, 1983; Miller, Goldstein, & Nicolaisen, 1978).

Diabetes education has historically had as its objective the

didactic transmission of facts about diabetes, based on the as-

sumption that increasing knowledge of the •factsŽ of diabetes

would improve BG control and, ultimately, reduce the inci-

dence and severity of complications (Beeney et al., 1994).

The traditional patient education has relied primarily on writ-

ten material about the disease process, medical management,