Association between depression and diabetes

This study examines the association between major depressive disorder and type 2 diabetes in the Chinese population. The findings reveal that major depressive episode is associated with an increased risk of type 2 diabetes.

Running head: HEALTH CARE
Association between depression and diabetes
Name of the student:
Name of the University:
Author’s note

1HEALTH CARE
PART A:
Critical appraisal of cohort study design:
Did the study address a clearly focused issue? The study by Meng et al. (2018) was based on a
focussed issue which included examination of the
association between major depressive disorder and
type 2 diabetes in Chinese population.
Was the cohort recruited in an acceptable way? The population for the study was taken from the
Chin Kadoorie BioBank study which includes
512891 individuals aged between 30-79 years from
10 diverse areas across China. The cohort size is
acceptable. The sample is representative of the
target population as the study considered diversity
elements like socioeconomic status, diversity of
exposures, disease patterns and local
infrastructures.
Was the exposure accurately measured to minimise
bias?
There was no exposure related assessment.
However, the study measured physical
measurements of participants like height, weight
and blood pressure. In addition, the depression
status of individual participants was measured.
Medical records were used to identify diabetes
cases.
Was the outcome accurately measured to minimise
bias?
The main outcome of interest for this study was
depression and risk of diabetes. The prevalence of
major depressive episode in Chinese cohort were

2HEALTH CARE
measured using the modified Chinese version of
Composite International Diagnostic Interview
Short-Form (CIDI-SF). Use of this tool is likely to
minimize bias as the study confirmed the
sensitivity and specificity of the tool in Chinese
population.
Have the authors identified all important
confounding factors?
To enhance the quality and validity of research
findings, it is necessary that any confounding factor
is accurately identified and addressed in research
(Kyriacou & Lewis, 2016). Taking data from past
medical record is a major confounding factor that
can contribute to biases.
Have they taken account of the confounding factors
in the design and/or analysis?
In case of using CIDI-SF tool for depression
assessment, the sensitivity and specificity element
of the tool has been considered. These
considerations are likely to eliminate risk of
confounding factors in result. The statistical
analysis method was also comprehensive reducing
risk of confounding factors.
Was the follow up of subjects complete enough? In cohort study, adequate follow up overtime is
necessary to ascertain the occurrence of an
outcome (Howe et al., 2016). The follow up of
Chinese population was done for 8 years and loss
to follow-up was considered during statistical
analysis. The significance of this research was that

3HEALTH CARE
there was high-follow up rate.
Was the follow up of subjects long enough? The follow-up of the cohort group was done for 8
years which is adequate time to interpret risk of
development of depression and diabetes in an
individual.
What are the results of this study? The statistical analysis of Chinese population data
from China Kadoorie Biobank study revealed that
major depressive episode is associated with
increased risk of type 2 diabetes. High rate of
incidence of type 2 diabetes was found for people
with depression compared to those without
depression. Hence, the research question was
addressed.
How precise are the results? . In case of participants with a depressive episode,
the risk of type 2 diabetes was 32% with CI 1.05–
1.68. Hence, the result was statistically significant
indicating the preciseness of the study findings.
Do you believe the results? Sensitivity aspect was considered for use of tools.
In addition, in case of review of diabetes cases
from medical record, the lab results were
reconfirmed. In addition, no loss-to-follow up cases
and long term-period of analysis enhanced the
validity of the study findings.
Can the results be applied to the local population? The findings cannot be applied to local population
because it was a prospective analysis and there is a

4HEALTH CARE
need to validate the finding in Chinese population
currently as health care systems and ways to seek
treatment has changed.
Do the results of this study fit with other available
evidence?
Yes, the study finding is consistent with other
available evidences as the study indicated other
research papers were similar type of association
between depression and type 2 diabetes.
Critical appraisal of case control study design:
Did the study address a clearly focused issue? The study by Islam et al. (2015) focused on
analysing the association between depression and
diabetes among adults in Bangladesh. The study
has clear focussed question as it defines the
research outcome of interest and target population
group
Did the authors use an appropriate method to
answer their question?
Islam et al. (2015) used matched case-controlled
study design to answer the research question by
taking 591 patients with diabetes in the Bangladesh
Institute of Health Sciences (BIHS) hospital as
cases and control group matched with age, sex and
area of residence. This research design is chosen
when the purpose of research is to determine the
exposure to the risk factor of interest in two groups
of individuals particularly the control and the cases

5HEALTH CARE
group (Lewis, 2015).
Were the cases recruited in an acceptable way? The study also defined the manner in which the
cases were diagnosed which included diagnosis of
diabetes by BIHS physician using the WHO
criteria. Hence, established criteria have been used
to recruit cases thus reducing the risk of bias. The
number of cases selected is also appropriate.
However, no power calculation was done.
Were the controls selected in an acceptable way? For each index case, controls were taken which
matched for age, sex and area of residence.
However, socioeconomic factor was not
considered. This should be done to reduce risk of
confounding factors completely.
Was the exposure accurately measured to minimise
bias?
The exposure for the study included presence of
type 2 diabetes in study subjects. The exposure
assessment was done by taking those cases where
physicians diagnosed diabetes using the WHO
criteria. Blood tests on HbA1c were measured at
the BIHS research laboratory. Other associated risk
factor of diabetes such as blood pressure, BMI and
history of disease was considered using
questionnaire. Hence, both subjective and objective
data was used to evaluate exposure thus reducing
risk of biases in study result.
Aside from the experimental intervention, were the
groups treated equally?
As this study aimed to evaluate association
between diabetes and risk of diabetes, there was no

6HEALTH CARE
experimental intervention to be studied. Hence, no
equal treatment was required. However, matching
of controls with index cases was done by matching
for age, sex and area of residence.
Have the authors taken account of the potential
confounding factors in the design and/or in their
analysis?
The confounding factor in study design has been
considered by ensuring that validated criteria and
tool is used to evaluate exposure and outcome. For
example, WHO criteria for diabetes diagnosis were
selected. The author also considered influence of
language barrier on study result and translated all
tools into Bengali thus reducing biases in study
design.
How large was the treatment effect? The study finding revealed that prevalence of
depressive illness was higher among persons with
diabetes. The treatment effect is understood from
the fact that incidence of depression was found to
be 28.3% for patient with diabetes compared to
16.9% for patient without diabetes. Hence, not very
large but significant treatment effect has been
found.
How precise was the estimate of the treatment
effect?
The preciseness of treatment effect can understood
from confidence interval data of exposure and
outcome. The study revealed that mild as well as
moderate depression was linked with diabetes with
CI of 5.4 to 18.0 thus suggesting statistically

7HEALTH CARE
significant effect between exposure and outcome.
Do you believe the results? The study finding is reliable as appropriate sample
size and data collection techniques have been used
to study findings. All associated risk factors of
diabetes and depression has been considered. As
the author considered biases in various areas of
study design, the study is likely to be reliable.
Can the results be applied to the local population? One limitation that affects the application of the
study in local population is the assessment of
depression at a single time instead of long term
follow-up. In addition, several data related to eye
problem and heart disease were based on self-
reported data.
. Do the results of this study fit with other available
evidence?
The study finding is consistent with perveious
studies as the study gave evidence regarding
association between depression and poor diabetes
in meta analysis as well as quantitative research
design
3. Critical appraisal of cross-sectional study design
Did the study address a clearly focused issue? The study by Akpalu et al. (2018) has clear
focussed question as the outcome of interest,
research variables and population group was well-
defines. The key research aim included
investigating association between glycemic control

8HEALTH CARE
and depression in type 2 diabetes patients attending
a tertiary health care facility in Ghana.
Was the cohort recruited in an acceptable way? The manner of sample recruitment is appropriate as
both sample size calculation and power analysis
was done. Systematic random sampling method
was used to recruit 400 type 2 diabetes patients in
the study. The exclusion criteria and effect of
medication shows that sample is recruited in
acceptable way.
Was the exposure accurately measured to minimise
bias?
The exposure included patients with type 2 diabetes
and this was measured by blood test to record
HbA1c value. The criteria for good glycamic
control were defined and ethical requirement was
maintained. Hence, use of objective measurements
ensures that there is low risk of bias during
exposure measurement.
Was the outcome accurately measured to minimise
bias?
The outcome related to risk of depression was
measured by screening patients using the Patient
Health Questionnaire-9. It is a validated tool tha
has been found to be used for detection of general
depressive symptoms (Woldetensay et al. 2018)
Have the authors identified all important
confounding factors?
Yes, all important confounding factors has been
identified.
Have they taken account of the confounding factors
in the design and/or analysis?
Confounding factors include those elements thus
has impact on study results and that contributes to

9HEALTH CARE
misinterpretation of study findings. The author
identified effect of sample size and hence used
appropriate sample size by calculating sample size
and power calculation. Hence, risk of biases at
critical point of research was considered.
Was the follow up of subjects complete enough? No details regarding follow up time provided
Was the follow up of subjects long enough? The loss to follow up case and long term period of
analysis was not considered.
What are the results of this study? The study revealed that depression is common
among Ghana people with type 2 diabetes mellitus.
How precise are the results? The preciseness of the result is understood from the
fact that 31.3% of sample group were found to have
depression.
Do you believe the results? The results are reliable as valid instruments and
both objective and subjective data has been used to
confirm the findings. It eliminates risk of biases
due o self-reporting or inaccuracy of study tools/
Can the results be applied to the local population? Although the study shows association between
depression and diabetes, however the study cannot
be applied in local population as the research was
done only in a tertiary hospital. Hence, the sample
and setting is not representative of the whole
population.
Do the results of this study fit with other available
evidence?
The study is consistent with other research papers
as the researcher cited papers that reported

10HEALTH CARE
prevalence of depression among diabetes.
References:

11HEALTH CARE
Akpalu, J., Yorke, E., Ainuson-Quampah, J., Balogun, W., & Yeboah, K. (2018). Depression and
glycaemic control among type 2 diabetes patients: a cross-sectional study in a tertiary
healthcare facility in Ghana. BMC psychiatry, 18(1), 357.
Howe, C. J., Cole, S. R., Lau, B., Napravnik, S., & Eron Jr, J. J. (2016). Selection bias due to
loss to follow up in cohort studies. Epidemiology (Cambridge, Mass.), 27(1), 91.
Islam, S. M. S., Ferrari, U., Seissler, J., Niessen, L., & Lechner, A. (2015). Association between
depression and diabetes amongst adults in Bangladesh: a hospital based case–control
study. Journal of global health, 5(2).
Kyriacou, D. N., & Lewis, R. J. (2016). Confounding by indication in clinical
research. Jama, 316(17), 1818-1819.
Lewis, S. (2015). Qualitative inquiry and research design: Choosing among five
approaches. Health promotion practice, 16(4), 473-475. s
Meng, R., Liu, N., Yu, C., Pan, X., Lv, J., Guo, Y., ... & Pan, A. (2018). Association between
major depressive episode and risk of type 2 diabetes: A large prospective cohort study in
Chinese adults. Journal of affective disorders, 234, 59-66.
Woldetensay, Y. K., Belachew, T., Tesfaye, M., Spielman, K., Biesalski, H. K., Kantelhardt, E.
J., & Scherbaum, V. (2018). Validation of the Patient Health Questionnaire (PHQ-9) as a
screening tool for depression in pregnant women: Afaan Oromo version. PloS one, 13(2),
e0191782.