Diabetes Symptoms and Distress in ACCORD Trial Participants: Relationship to Baseline Clinical Variables

 
CONTINUE READING
F e a t u r e          A r t i c l e

    Diabetes Symptoms and Distress in ACCORD Trial
  Participants: Relationship to Baseline Clinical Variables
           Mark D. Sullivan, MD, PhD, Gregory Evans, MA, Roger Anderson, PhD, Patrick O’Connor, MD, MPH,
               Dennis W. Raisch, PhD, Debra L. Simmons, MD, MS, and K.M. Venkat Narayan, MD, MBA

T
        ype 2 diabetes may lead to               association of diabetes symptoms          or at least two additional CVD risk
        a variety of symptoms such               and distress with patients’ overall       factors (dyslipidemia, hyperten-
        as excessive thirst, frequent            health state as measured by a feeling     sion, current status as a smoker,
urination, fatigue, and burning feet.            thermometer. The feeling thermom-         or obesity). Key exclusion criteria
These symptoms diminish quality of               eter allows patients to rate their        included frequent or recent serious
life, impair functional status, and              current overall health between 100        hypoglycemia events, unwillingness
contribute to the psychological                  (perfect health) and 0 (death).           to perform self-monitoring of blood
distress experienced by patients with                                                      glucose or inject insulin, a BMI
diabetes.1–4 However, there is no                Study Methods                             > 45 kg/m2, a serum creatinine level
established metric for the severity of           Study population                          > 1.5 mg/dl, or other serious illness.
diabetes symptoms and associated                 The rationale and design of the           A total of 10,251 participants were
distress.                                        Action to Control Cardiovascular Risk     recruited and randomly assigned to
    Many diabetes symptoms                       in Diabetes (ACCORD) trial and its        either intensive glycemia-lowering
are linked through established                   health-related quality of life substudy   with a target A1C < 6.0% or standard
pathophysiological mechanisms                    have been reported previously.8,9         glycemia management with a target
to inadequate short- or long-term                Briefly, the ACCORD trial, spon-          A1C of 7.0–7.9%.
glucose control or acute hypoglyce-              sored by the National Heart, Lung,            The study protocol was approved
mia. But studies to date suggest that            and Blood Institute (NHLBI), was          by the institutional review board or
the relationship between severity of             conducted in 77 clinical centers across   ethics committee at each center, as
diabetes symptoms and measures                   the United States and Canada.             well as by an ethics review panel at
of glucose control such as A1C is                   We recruited participants with         the NHLBI. All patients provided
weak.5–7                                         type 2 diabetes and an A1C ≥ 7.5%         written informed consent.
    A better understanding of factors            and who either were 1) between the            Of the 10,251 patients enrolled
that amplify or dampen diabetes-                 ages of 40 and 79 years and had           in the trial, a randomly selected
related symptoms could lead to                   cardiovascular disease (CVD) or 2)        subsample of 2,053 participants
improved approaches to maxi-                     between the ages of 55 and 79 years       from each of the clinical centers
mize the quality of life of diabetes             and had anatomical evidence of            was enrolled in a substudy concern-
patients. The purpose of this study              significant atherosclerosis, albumin-     ing health-related quality of life
was to describe the relationship of              uria, left-ventricular hypertrophy,       (HRQL). The ACCORD HRQL
scores on the Diabetes Symptoms                                                            substudy was designed to assess
Distress Questionnaire to demo-                               In Brief                     three distinct outcomes: general
graphic and clinical variables for                                                         health, treatment satisfaction,
patients with type 2 diabetes. This               Our study demonstrates strong            and diabetes-related symptoms.
included an evaluation of the cross-              associations of diabetes symptoms        This report focuses on associa-
sectional association of diabetes                 and distress with female sex, higher     tions between baseline symptoms
                                                  BMI, history of neuropathy, and
symptoms and distress with demo-                                                           and symptom distress, a feeling
                                                  current depressive symptoms.
graphic and clinical variables such                                                        thermometer used to rate patients’
                                                  Many diabetes-specific symptoms
as A1C, LDL cholesterol, blood                                                             general health state, and multiple
                                                  may be significantly shaped by
pressure, diabetes duration and                                                            demographic and clinical variables.
                                                  factors such as depression and
complications, and depression                                                              It includes all ACCORD study
                                                  obesity.
status. We also examined the                                                               subjects who were enrolled in the

Clinical Diabetes • Volume 30, Number 3, 2012                                                                                 101
F e a t u r e          A r t i c l e

ACCORD HRQL substudy and who              physical exam. Systolic and diastolic    used to assess associations between
completed baseline data collection.       blood pressure levels were deter-        the four dependent variables and the
The ACCORD HRQL methods have              mined using the average of three         feeling thermometer. Results for the
been described previously.10              readings using an Omron device           two methods were quite similar (all
                                          (Omron Inc., Kyoto, Japan).              coefficients agreed ± 0.05), and only
Key measures                                  Depression was assessed using        Pearson correlation coefficients are
We used the Diabetes Symptom              the nine-item depression mea-            presented here.
Distress Questionnaire developed          sure from the Patient Health
by Anderson and Testa11 to assess 60      Questionnaire (PHQ-9). The PHQ-9         Ethical approval
individual symptoms of diabetes and       is the self-report version of the        The University of Washington
its treatment. This measure has been      Primary Care Evaluation of Mental        Institutional Review Board for
previously validated against physi-       Disorders questionnaire, a well-         the Protection of Human Subjects
cian report of patient symptoms.11 It     validated psychiatric diagnostic         approved this research, as did the
discriminates between patients with       interview for use in primary care        Institutional Review Boards of the
diabetes and those with hyperten-         settings.14 In this analysis, we used    other ACCORD clinical networks
sion12 and distinguishes between          the PHQ-9 mean score and whether         and the clinics where data were
patients with type 2 diabetes random-     the score exceeded the threshold         collected.
ized to glipizide or placebo.1 With       suggesting major depression (a score
this questionnaire, subjects reported                                              Study Results
                                          ≥ 10 on the PHQ-9). We also assessed
whether they had experienced a given      patients’ history of depression and      Baseline sample characteristics
symptom or feeling. If they had expe-     use of antidepressant medications at     Table 1 displays the characteristics
rienced it, then they were asked “How     baseline.                                of the 1,950 study subjects who were
distressing was it?” according to this        Finally, a feeling thermometer       included in the ACCORD HRQL sub-
scale: 0 = not at all; 1 = somewhat;      instrument15 was used to assess each     study and provided complete baseline
2 = moderately; 3 = very much; and        participant’s overall health percep-     data. The study sample had a mean
4 = extremely. For each participant,      tions.16 This instrument consists of a   age of 62 years and was 60% male;
we calculated the total symptom           single-item visual analog scale with     60% had post-secondary education,
count and the mean symptom distress       which participants are asked to rate     80% lived with other adults, and 66%
(assigning participants not experienc-    their current (today) health state       were Caucasian. The mean duration
ing the symptom a distress score of       from 0 (worst imaginable) to 100         of diabetes was just over 11 years. The
0). We also examined the relationship     (best imaginable).                       mean baseline A1C was 8.3 ± 1%, with
between these and an overall rating                                                56% of the sample having an A1C of
of the patient’s health state using a     Statistical analysis                     ≥ 8%.
feeling thermometer.                      We used both simple and multiple
    Participant age, sex, ethnicity,      linear regression analyses to assess     Patient responses to Diabetes
educational level, social support         relationships between demographic,       Symptoms Distress Questionnaire
(living alone vs. living with oth-        diabetes, and depression status and      The 10 most commonly endorsed
ers), diabetes duration, history of       the specified dependent variables.       symptoms on the Diabetes Symptoms
eye disease, history of neuropathy,       Separate models were fit for each        Distress Questionnaire were: drowsy
and medication use were based on          of the four dependent variables.         or sleepy (59%), getting up often at
self-report. A central laboratory         Multiple linear regression models        night to urinate (57%), feeling over-
(with National Glycohemoglobin            generally took this form: Diabetes       weight (57%), tired or being weary
Standardization Program level I           symptoms = demographics (step 1)         (57%), being thirsty (50%), numbness
certification) analyzed blood for A1C     + diabetes duration and A1C and          of hands or feet (50%), having to
and lipid levels. Total cholesterol,      complications (step 2) + depression      urinate frequently (50%), drinking
HDL cholesterol, and fasting triglyc-     (step 3) + glucose, blood pressure,      a lot of fluids (50%), general weak-
eride concentrations were measured        and lipid treatments (step 4). Given     ness or fatigue (48%), and lethargy
enzymatically, and LDL cholesterol        the large number of tests performed,     or no energy to do things (44%).
was calculated using the Friedewald       only those associations significant at   These were generally also rated as
formula.13 BMI and Michigan               the P < 0.001 level were considered      the most distressing symptoms. The
Neuropathy Screening Instrument           statistically significant. Pearson and   most distressing symptom (range
(MNSI) scores were determined by          Spearman rank correlations were          0–4) was feeling overweight (mean

102                                                                                   Volume 30 Number 3, 2012• Clinical Diabetes
F e a t u r e            A r t i c l e

                                                                                          majority reporting multiple symp-
 Table 1. Demographic and Clinical Characteristics of Study Subjects (n = 1,950)
                                                                                          toms. The mean symptom count was
 Characteristic                                          Percentage or Mean (SD)          17.1. The median number of symp-
 Mean age (years)                                        62.3 (6.7)                       toms was 15.0. Figure 2 displays the
                                                                                          distribution of symptom distress
 Female (%)                                              39.8                             in the study population. Symptom
 Having post-secondary education (%)                     59.8                             distress was generally low, with 73%
                                                                                          reporting mean distress between 0
 Living with other adults (%)                            80.0                             and 1 on the four-point scale. The
 Caucasian (%)                                           65.9                             overall mean for distress was 0.7, and
                                                                                          the median distress score was 0.6.
 Mean duration of diabetes (years)                       11.1 (7.9)                           Total symptom count was highly
 Mean baseline A1C (%)                                   8.3 (1)                          correlated with mean symptom dis-
                                                                                          tress (Pearson r = 0.88, P < 0.0001).
 Having baseline A1C ≥ 8% (%)                            55.6                             Symptom count (r = –0.35,
 Reporting history of neuropathy (%)                     27.5                             P < 0.0001) and symptom dis-
                                                                                          tress (r = –0.36, P < 0.0001) were
 Having foot amputation or MNSI score > 2 (%)            43.1
                                                                                          significantly and similarly correlated
 Reporting history of eye disease (%)                    29.7                             with the overall health state rating
                                                                                          on the feeling thermometer.
 Mean BMI (kg/m )      2
                                                         32.4 (5.5)
 Mean systolic blood pressure (mmHg)                     136.3 (17.1)                     Univariate relationships between
                                                                                          symptoms, distress, and clinical
 Mean diastolic blood pressure (mmHg)                    74.4 (10.9)                      variables
 Mean total cholesterol (mg/dl)                          182.8 (41.1)                     Table 2 displays the univariate rela-
                                                                                          tionships between symptoms, distress,
 Mean LDL cholesterol (mg/dl)                            104.1 (33.8)
                                                                                          and clinical variables. Total symptom
 Mean HDL cholesterol (mg/dl)                            42.1 (11.6)                      count was significantly and negatively
                                                                                          associated with age. It was signifi-
 Mean triglyceride level (mg/dl)                         189.2 (140.2)
                                                                                          cantly and positively associated with
 Reporting history of depression (%)                     24.6                             female sex, history of neuropathy,
 Using antidepressant medications (%)                    14.0                             BMI, serum triglygeride level, history
                                                                                          of depression, use of antidepressants,
 Mean PHQ-9 score                                        5.4 (5.1)                        mean PHQ-9 score, having a PHQ-9
 Having PHQ-9 score ≥ 10 (%)                             19.6                             score ≥ 10, and use of insulin. Mean
                                                                                          symptom distress was significantly
 Using oral hypoglycemic medications only (%)            56.9                             and negatively associated with age
 Using insulin (%)                                       35.9                             and use of only oral hypoglycemic
                                                                                          medications. Mean symptom distress
 Using thiazides (%)                                     26.8
                                                                                          was significantly and positively asso-
 Using β-blockers (%)                                    30.4                             ciated with female sex, baseline mean
                                                                                          A1C, having a baseline A1C ≥ 8%,
 Using ACE inhibitors (%)                                52.0
                                                                                          history of neuropathy, BMI, total
 Using statins (%)                                       63.5                             cholesterol, triglyceride level, history
                                                                                          of depression, use of antidepres-
distress score of 2.2). Eight symptoms           reaction, pain in legs or calves when    sants, mean PHQ-9 score, having a
tied for the second most distressing             walking, gaining weight, and inability   PHQ-9 ≥ 10, and use of insulin.
(mean distress score 1.8), including             to sleep or insomnia.
numbness or tingling of hands or feet,               Figure 1 displays the distribu-      Multivariable models for symptom
having to urinate frequently, lethargy           tion of symptom counts in the study      count, distress, and factors
or no energy to do things, foot cramps           population. Less than 2% of subjects     Multivariable models were derived in
or foot pain, high blood glucose                 reported no symptoms, with the vast      a progressive manner, entering demo-

Clinical Diabetes • Volume 30, Number 3, 2012                                                                                  103
F e a t u r e           A r t i c l e

                                                                                      associated with multiple factors when
                                                                                      these are considered individually,
                                                                                      including demographic and psycho-
                                                                                      logical variables and measures of
                                                                                      diabetes control and complications.
                                                                                      However, in multivariable models,
                                                                                      diabetes symptoms and distress are
                                                                                      significantly associated only with
                                                                                      sex, BMI, history of neuropathy,
                                                                                      and current depressive symptoms
                                                                                      (PHQ-9 score ≥ 10). These findings
                                                                                      suggest that efforts to reduce diabetes
                                                                                      symptoms should focus on strategies
                                                                                      to reduce neuropathy and depression.
                                                                                      We did not correct for any ongoing
                                                                                      depression treatment (medications or
                                                                                      psychotherapy) in our analyses.
Figure 1. Number of symptoms per participant.                                             Previous research on diabetes
                                                                                      symptoms has shown a stronger
                                                                                      relationship between diabetes
                                                                                      symptom severity and the mental
                                                                                      component score of the Short Form-
                                                                                      36 Health Survey than the physical
                                                                                      component score.2 Other research
                                                                                      has suggested an important role for
                                                                                      depression in diabetes symptoms.
                                                                                      In a primary care sample of 4,168
                                                                                      patients with diabetes, Ludman et
                                                                                      al.6 found that patients with major
                                                                                      depression had more diabetes symp-
                                                                                      toms after adjusting for demographic
                                                                                      characteristics, objective measures
                                                                                      of diabetes severity, and medical
                                                                                      comorbidity. The overall number of
                                                                                      diabetes symptoms was related to the
                                                                                      number of depressive symptoms, and
Figure 2. Distribution of mean distress scores.                                       depression was significantly related
                                                                                      to all of the 10 diabetes symptoms
graphic variables first, then diabetes      primary outcomes:                         assessed. Previously, Ciechanowski
and cardiovascular risk variables,          • Total symptom count was sig-            et al.5 reported strong associa-
then depression variables, and then           nificantly associated with female       tions between diabetes symptoms
treatments for diabetes, blood pres-          sex (β = 2.24), history of neuropathy   and depression in a sample of 273
sure, and lipid control. These models         (β = 3.71), and having a PHQ-9          diabetic patients recruited from a
are displayed in Table 3.                     score ≥ 10 (β = 11.13).                 specialty care setting.
   Most predictor variables that            • Symptom distress was significantly          Depression in patients with
entered the multivariable models              associated with history of neuropa-     diabetes is likely both a cause and a
were significant at the P ≤ 0.001             thy (β = 0.10), BMI (β = 0.01), and     consequence of diabetes symptoms,
level, and most remained signifi-             PHQ-9 score ≥ 10 (β = 0.48).            complications, and related health
cant in the more complete models.                                                     behaviors.17 Diabetes symptoms and
We therefore report only the final          Discussion                                complications, smoking, and obesity
models including variables signifi-         Diabetes symptom count and                (BMI) have all been associated with
cant at P < 0.0001 for each of the two      symptom-related distress appear to be     an increased risk of depression in

104                                                                                      Volume 30 Number 3, 2012• Clinical Diabetes
F e a t u r e           A r t i c l e

 Table 2. Univariate Relationships With Symptom Outcomes
 Characteristic                                                    Total Symptom Count                     Mean Distress
                                                          β             SE         P            β            SE            P
 Age (years)                                              –0.169        0.038      < 0.0001     –0.007       0.001         < 0.0001
 Sex (female vs. male)                                    3.817         0.512      < 0.0001     0.137        0.018         < 0.0001
 Post-secondary education (yes vs. no)                    –0.118        0.519      0.821        –0.018       0.018         0.317
 Living alone (yes vs. no)                                –1.00         0.636      0.114        –0.015       0.022         0.506
 Caucasian (yes vs. no)                                   0.246         0.537      0.646        0.001        0.019         0.962
 Duration of diabetes (years)                             0.054         0.033      0.098        0.002        0.001         0.150
 Baseline A1C (%)                                         0.557         0.243      0.022        0.037        0.009         < 0.0001
 Baseline A1C ≥ 8% (yes vs. no)                           1.590         0.511      0.002        0.079        0.018         < 0.0001
 History of neuropathy (yes vs. no)                       5.827         0.559      < 0.0001     0.195        0.020         < 0.0001
 Foot amputation or MNSI score > 2 (yes vs. no)           1.596         0.513      0.002        0.051        0.018         0.005
 History of eye disease (yes vs. no)                      0.797         0.558      0.154        0.030        0.020         0.125
 BMI (kg/m2)                                              0.387         0.045      < 0.0001     0.015        0.002         < 0.0001
 Systolic blood pressure (mmHg)                           0.000         0.015      0.989        0.000        0.001         0.892
 Diastolic blood pressure (mmHg)                          0.048         0.023      0.040        0.002        0.001         0.012
 Total cholesterol (mg/dl)                                0.021         0.006      0.001        0.001        0.000         < 0.0001
 LDL cholesterol (mg/dl)                                  0.007         0.008      0.3669       0.000        0.000         0.154
 HDL cholesterol (mg/dl)                                  –0.004        0.022      0.8242       0.000        0.001         0.799
 Triglyceride level (mg/dl)                               0.008         0.002      < 0.0001     0.000        0.000         < 0.0001
 History of depression (yes vs. no)                       7.280         0.570      < 0.0001     0.268        0.020         < 0.0001
 Using antidepressant medications (yes vs. no)            6.283         0.719      < 0.0001     0.258        0.025         < 0.0001
 PHQ-9 score                                              1.311         0.040      < 0.0001     0.050        0.001         < 0.0001
 PHQ-9 score ≥ 10 (yes vs. no)                            13.357        0.566      < 0.0001     0.559        0.019         < 0.0001
 Using oral hypoglycemics only (yes vs. no)               –1.879        0.512      0.000        –0.086       0.018         < 0.0001
 Using insulin (yes vs. no)                               2.447         0.527      < 0.0001     0.111        0.018         < 0.0001
 Using thiazides (yes vs. no)                             –0.238        0.574      0.678        –0.010       0.020         0.620
 Using β-blockers (yes vs. no)                            1.614         0.552      0.005        0.038        0.019         0.051
 Using ACE inhibitors (yes vs. no)                        –0.743        0.509      0.145        –0.030       0.018         0.091
 Using statins (yes vs. no)                               –0.229        0.529      0.666        –0.008       0.019         0.679

previous studies.18–21 This suggests             ing and eating) exist in a mutually              This pattern of reciprocal
that aversive symptoms (such as pain             reinforcing pattern that exerts a            interactions between symptom
from neuropathy), depression, and                significant effect on patients’ overall      severity, depression, and qual-
health behaviors associated with                 health state, as indicated by the feel-      ity of life has been found in other
mood regulation (such as smok-                   ing thermometer findings.                    chronic diseases such as asthma22

Clinical Diabetes • Volume 30, Number 3, 2012                                                                                       105
F e a t u r e           A r t i c l e

 Table 3. Multivariable Relationships With Symptom Burden and Distress
 Characteristic                                             Total Symptom Count                     Mean Distress
                                                   β             SE         P            β             SE             P
 Age (years)                                       –0.0223       0.0388     0.5658        –0.0003      0.0013         0.7909
 Sex (female vs. male)                             2.2421        0.5181     < 0.0001      0.0612       0.0169         0.0003
 Post-secondary education (yes vs. no)             0.5286        0.4683     0.2591        0.0077       0.0153         0.6149
 Living alone (yes vs. no)                         –0.0112       0.5607     0.9841        0.0211       0.0183         0.2479
 Caucasian (yes vs. no)                            –0.5190       0.5092     0.3082        –0.0251      0.0166         0.1311
 Duration of diabetes (years)                      0.0168        0.0322     0.6021        –0.0003      0.0011         0.7396
 Baseline A1C (%)                                  0.0060        0.2199     0.9784        0.0148       0.0072         0.0393
 History of neuropathy (yes vs. no)                3.7069        0.5373     < 0.0001      0.1042       0.0175         < 0.0001
 Foot amputation or MNSI score > 2 (yes vs. no)    0.8031        0.4757     0.0915        0.0219       0.0155         0.1572
 History of eye disease (yes vs. no)               –0.2710       0.5299     0.6091        –0.0050      0.0173         0.7726
 BMI (kg/m2)                                       0.1395        0.0443     0.0017        0.0072       0.0014         < 0.0001
 Systolic blood pressure (mmHg)                    0.0110        0.0167     0.5093        0.0004       0.0005         0.478
 Diastolic blood pressure (mmHg)                   0.0067        0.0277     0.8091        0.0001       0.0009         0.892
 Total cholesterol (mg/dl)                         0.0031        0.0072     0.6701        0.0001       0.0002         0.7111
 LDL cholesterol (mg/dl)                           0.0077        0.0232     0.7415        0.0011       0.0008         0.163
 Triglyceride level (mg/dl)                        0.0038        0.0018     0.0328        0.0002       0.0001         0.0006
 History of depression (yes vs. no)                2.1388        0.6303     0.0007        0.0517       0.0205         0.0118
 Using antidepressant medications (yes vs. no)     0.6774        0.7454     0.3636        0.0488       0.0243         0.0446
 PHQ-9 score ≥ 10 (yes vs. no)                     11.1252       0.6105     < 0.0001      0.4804       0.0199         < 0.0001
 Using oral hypoglycemic only (yes vs. no)         0.9480        0.9195     0.3026        0.0416       0.0300         0.1656
 Using insulin (yes vs. no)                        1.5094        0.9896     0.1274        0.0774       0.0322         0.0165
 Using thiazides (yes vs. no)                      –0.3656       0.5111     0.4744        –0.0085      0.0167         0.6084
 Using β-blockers (yes vs. no)                     1.1790        0.5056     0.0198        0.0275       0.0165         0.0949
 Using ACE inhibitors (yes vs. no)                 –0.9374       0.4524     0.0384        –0.0403      0.0147         0.0064
 Using statins (yes vs. no)                        –0.1261       0.5001     0.801         0.0010       0.0163         0.951

and chronic kidney disease requiring         Our analysis has some important           waves of data will be necessary to
hemodialysis.23 Our study adds to         limitations. First, the cross-sectional      fully understand these complex
these other studies in supporting the     study design precludes causal                relationships.
idea that diabetes-specific symptoms      inference. Longitudinal data                    Second, our sample is not rep-
are more significantly shaped by fac-     are necessary to prove causality.            resentative of the entire population
tors such as depression and obesity       Because we posit reciprocal relation-        of patients with type 2 diabetes.
than by severity and duration of          ships between diabetes symptoms              ACCORD participants had diabetes
diabetes per se.                          and their determinants, multiple             for a mean of 10 years, were at high

106                                                                                       Volume 30 Number 3, 2012• Clinical Diabetes
F e a t u r e               A r t i c l e

risk for cardiovascular events, and                           References                         diabetes. Postgrad Med J 64 (Suppl. 3):50–58,
                                                                                                 discussion 90–52, 1988
were willing to undergo intensive                    1
                                                      Testa MA, Simonson DC: Health
                                                                                                     13
                                                 economic benefits and quality of life                 Friedewald WT, Levy RI, Fredrickson
treatment to control glucose, includ-            during improved glycemic control in             DS: Estimation of the concentration of
ing frequent clinic visits and the use           patients with type 2 diabetes mellitus: a       low-density lipoprotein cholesterol in
                                                 randomized, controlled, double-blind            plasma, without use of the preparative
of insulin.                                      trial. JAMA 280:1490–1496, 1998                 ultracentrifuge. Clin Chem 18:499–502, 1972
    Third, although we assessed                       2
                                                        Gulliford MC, Mahabir D:                     14
                                                                                                        Spitzer RL, Kroenke K, Williams JB:
a broad range of factors for their               Relationship of health-related quality of       Validation and utility of a self-report version
                                                 life to symptom severity in diabetes            of PRIME-MD: the PHQ primary care study.
associations with diabetes symp-                 mellitus: a study in Trinidad and Tobago.       JAMA 282:1737–1744, 1999
toms, there are many other factors               J Clin Epidemiol 52:773–780, 1999                    15
                                                                                                        EuroQol Group: EuroQol: a new facility
that could have been examined
                                                     3
                                                      Goddijn PP, Bilo HJ, Feskens EJ,           for the measurement of health-related quality
                                                 Groeniert KH, van der Zee KI, Meyboom-          of life. Health Policy 16:199–208, 1990
in metabolic (C-reactive protein),               de Jong B: Longitudinal study on glycae-            16
                                                                                                        Schunemann HJ, Griffith L, Jaeschke
physiological (nerve conduction                  mic control and quality of life in patients
                                                 with type 2 diabetes mellitus referred for      R, Goldstein R, Stubbing D, Guyatt GH:
velocities), behavioral (exercise), and          intensified control. Diabet Med 16:23–30,       Evaluation of the minimal important
                                                 1999                                            difference for the feeling thermometer and
psychosocial (anxiety) domains that                                                              the St. George’s Respiratory Questionnaire
                                                     4
                                                       Currie CJ, Poole CD, Woehl A, Morgan      in patients with chronic airflow obstruction.
may have affected the final multi-               CL, Cawley S, Rousculp MD, Covington            J Clin Epidemiol 56:1170–1176, 2003
variable models.                                 MT, Peters JR: The health-related utility           17
                                                                                                        Golden SH, Lazo M, Carnethon M,
                                                 and health-related quality of life of
    In summary, our study demon-                 hospital-treated subjects with type 1 or        Bertoni AG, Schreiner PJ, Diez Roux AV, Lee
                                                                                                 HB, Lyketsos C: Examining a bidirectional
strates strong associations between              type 2 diabetes with particular reference to
                                                                                                 association between depressive symptoms
                                                 differing severity of peripheral neuropathy.
diabetes symptoms and distress and               Diabetologia 49:2272–2280, 2006                 and diabetes. JAMA 299:2751–2759, 2008
female sex, BMI, neuropathy, and                    5
                                                      Ciechanowski PS, Katon WJ, Russo JE,
                                                                                                     18
                                                                                                        Katon W, von Korff M, Ciechanowski P,
                                                 Hirsch IB: The relationship of depressive       Russo J, Lin E, Simon G, Ludman E, Walker
depression. These factors likely rein-                                                           E, Bush T, Young B: Behavioral and clinical
                                                 symptoms to symptom reporting, self-care
force each other, although precise               and glucose control in diabetes. Gen Hosp       factors associated with depression among
                                                 Psychiatry 25:246–252, 2003                     individuals with diabetes. Diabetes Care
specification of these relationships                                                             27:914–920, 2004
                                                     6
awaits longitudinal data.                              Ludman EJ, Katon W, Russo J, Von              19
                                                 Korff M, Simon G, Ciechanowski P, Lin                 Mier N, Bocanegra-Alonso A, Zhan
                                                 E, Bush T, Walker E, Young B: Depression        D, Wang S, Stoltz SM, Acosta-Gonzalez RI,
       Acknowledgments                           and diabetes symptom burden. Gen Hosp           Zuniga MA: Clinical depressive symptoms
                                                 Psychiatry 26:430–436, 2004                     and diabetes in a binational border
This work was supported by grants                                                                population. J Am Board Fam Med 21:223–
                                                    7
                                                     McKellar JD, Humphreys K, Piette JD:        233, 2008
from the NHLBI (N01-HC-95178,                    Depression increases diabetes symptoms by           20
N01-HC-95179, N01-HC-95180,                      complicating patients’ self-care adherence.            Sacco WP, Wells KJ, Friedman
                                                 Diabetes Educ 30:485–492, 2004                  A, Matthew R, Perez S, Vaughan CA:
N01-HC-95181, N01-HC-95182,                          8
                                                                                                 Adherence, body mass index, and
                                                       Buse JB, Bigger JT, Byington RP,          depression in adults with type 2 diabetes: the
N01-HC-95183, N01-HC-95184, IAA-                 Cooper LS, Cushman WC, Friedewald WT,           mediational role of diabetes symptoms and
Y1-HC-9035, and IAA-Y1-HC-1010);                 Genuth S, Gerstein HC, Ginsberg HN,             self-efficacy. Health Psychol 26:693–700, 2007
                                                 Goff DC Jr, Grimm RH Jr, Margolis KL,
by other components of the National              Probstfield JL, Simons-Morton DG, Sullivan
                                                                                                    21
                                                                                                      Sacco WP, Yanover T: Diabetes and
                                                 MD: Action to Control Cardiovascular Risk       depression: the role of social support and
Institutes of Health, including the                                                              medical symptoms. J Behav Med 29:523–531,
                                                 in Diabetes (ACCORD) trial: design and
National Institute of Diabetes and               methods. Am J Cardiol 99:21i–33i, 2007          2006
Digestive and Kidney Diseases,                       9
                                                       Gerstein HC, Riddle MC, Kendall
                                                                                                     22
                                                                                                        Richardson LP, Lozano P, Russo J,
                                                 DM, Cohen RM, Goland R, Feinglos                McCauley E, Bush T, Katon W: Asthma
the National Institute on Aging,                 MN, Kirk JK, Hamilton BP, Ismail-Beigi F,       symptom burden: relationship to asthma
and the National Eye Institute; by               Feeney P; ACCORD Study Group: Glycemia          severity and anxiety and depression
                                                 treatment strategies in the Action to Control   symptoms. Pediatrics 118:1042–1051, 2006
the Centers for Disease Control                  Cardiovascular Risk in Diabetes (ACCORD)            23
                                                                                                       Weisbord SD, Fried LF, Arnold RM,
and Prevention; and by General                   trial. Am J Cardiol 99:34i–43i, 2007            Fine MJ, Levenson DJ, Peterson RA, Switzer
Clinical Research Centers. The                       10
                                                        Sullivan MD, Anderson RT, Aron D,        GE: Prevalence, severity, and importance of
                                                 Atkinson HH, Bastien A, Chen GJ, Feeney         physical and emotional symptoms in chronic
following companies provided study               P, Gafni A, Hwang W, Katz LA, Narayan KM,       hemodialysis patients. J Am Soc Nephrol
medications, equipment, or sup-                  Nwachuku C, O’Connor PJ, Zhang P; ACCORD        16:2487–2494, 2005
                                                 Study Group: Health-related quality of life
plies: Abbott Laboratories, Amylin               and cost-effectiveness components of the
Pharmaceuticals, AstraZeneca,                    Action to Control Cardiovascular Risk in
                                                 Diabetes (ACCORD) trial: rationale and          Mark D. Sullivan, MD, PhD, is a
Bayer HealthCare, Closer                         design. Am J Cardiol 99:90i–102i, 2007          professor in the Department of
Healthcare, GlaxoSmithKline, King                   11
                                                       Anderson R, Testa M: Symptom distress     Psychiatry and Behavioral Sciences
Pharmaceuticals, Merck, Novartis,                checklists as a component of quality of life
                                                 measurement. Drug Inf J 28:89–114, 1994         at the University of Washington in
Novo Nordisk, Omron Healthcare,                     12
                                                      Testa MA, Simonson DC: Measuring
                                                                                                 Seattle. Gregory Evans, MA, is an
Sanofi-Aventis, and Schering-Plough.             quality of life in hypertensive patients with   assistant professor in the Departments

Clinical Diabetes • Volume 30, Number 3, 2012                                                                                               107
F e a t u r e          A r t i c l e

of Biostatistical Sciences and           O’Connor, MD, MPH, is assistant           Albuquerque. Debra L. Simmons,
Neurology at Wake Forest University      medical director and a senior clinical    MD, MS, is a professor at the
School of Medicine in Winston-Salem,     research investigator at HealthPartners   University of Arkansas for Medical
N.C. Roger Anderson, PhD, is a           Research Foundation in Minneapolis,       Sciences in Little Rock. K.M. Venkat
professor in the Department of Health    Minn. Dennis W. Raisch, PhD, is a         Narayan, MD, MBA, is a professor at
Sciences at Penn State College of        professor at the College of Pharmacy      the Rollins School of Public Health at
Medicine in Hershey, Pa. Patrick         of the University of New Mexico in        Emory University in Atlanta, Ga.

108                                                                                   Volume 30 Number 3, 2012• Clinical Diabetes
You can also read