p u b l i c h e a l t h 1 5 6 ( 2 0 1 8 ) 8 e1 4

Available online at www.sciencedirect.com

Public Health

journal homepage: www.elsevier.com/puhe

Original Research

Impact of health literacy on diabetes outcomes:

a cross-sectional study from Lahore, Pakistan

H. Saeed*, Z. Saleem, R. Naeem, I. Shahzadi, M. Islam

Section of Pharmacy Practice, University College of Pharmacy, University of the Punjab, Lahore, Pakistan

article info abstract

Article history: Objectives: To evaluate the functional health literacy of patients with type 2 diabetes in
Received 18 April 2017 Lahore and its impact on glycaemic control.
Received in revised form Study design: A six-month cross-sectional study.
12 July 2017 Methods: Health literacy in 204 patients with diabetes was evaluated using a validated
Accepted 6 December 2017 questionnaire (Short Test of Functional Health Literacy [s-TOFHLA]).
Results: The frequency distribution among various age groups (P ¼ 0.003), education levels
(P ¼ 0.0005), socio-economic status levels (P ¼ 0.0005) and glycated haemoglobin (HbA1C)
Keywords: levels (P ¼ 0.0005) differed significantly with health literacy level. The majority of patients
Diabetes with diabetes (86.1%) with poor glycaemic control (HbA1C >9%) had inadequate health lit-
Health literacy eracy and were more likely to have retinopathy (odds ratio ¼ 13.1, P ¼ 0.003). Health literacy
Retinopathy levels were not significantly different when compared for antidiabetic therapies (P ¼ 0.234).
Lahore Significant associations were observed between predictors of glycaemic control (s-TOFHLA
Glycaemic control score [P ¼ 0.0005], education status [P ¼ 0.0005] and disease risks [P ¼ 0.005]) and HbA1C,
Glycated haemoglobin level. However, after adjusting for basic characteristics, only s-TOFHLA score had a sig-
nificant association with HbA1C level (P ¼ 0.001).
Conclusions: These data suggest that inadequate health literacy is potentially associated
with poor glycaemic control, and microvascular and macrovascular complications,
particularly retinopathy. As such, educational and training programmes should be intro-
duced to improve functional health literacy of patients with diabetes for better glycaemic
control.
© 2017 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.

much attention and is considered a risk factor for poor

Introduction medication adherence, improper drug usage and adverse
outcomes.1e3 Studies have shown that patients with poor
Health literacy (i.e. a constellation of skills required to obtain, health literacy have difficulties in reading drug labels, dosing
process, understand and communicate health-related infor- schedules, educational brochures related to health, and
mation to make informed health decisions) has received

* Corresponding author. Section of Pharmacy Practice, University College of Pharmacy, University of the Punjab, Allama Iqbal Campus,
54000, Lahore, Pakistan. Tel.: þ92304 880 1243.
E-mail address: hamid.pharmacy@pu.edu.pk (H. Saeed).
https://doi.org/10.1016/j.puhe.2017.12.005
0033-3506/© 2017 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.
 p u b l i c h e a l t h 1 5 6 ( 2 0 1 8 ) 8 e1 4 9

informed consent forms; interpreting self-managed labora- hospitals based on clinician reports. Type 2 diabetes was
tory values; processing oral communication; and con- defined using the criteria of the American Diabetes Associa-
ceptualising disease and therapy risks.4,5 Similarly, low health tion (i.e. HbA1c level of 6.5% or higher, fasting plasma glucose
literacy has been documented as a stronger predictor of a of 126 mg/dl or higher and random plasma glucose of 200 mg/
person's health than age, education, socio-economic status dl or higher). Data were collected from four tertiary care
and employment status affecting health care and disease hospitals: Services Hospital Lahore, Mayo Hospital, Sheikh
management outcomes.6,7 In this context, poor health literacy Zayed Hospital and Nobel Hospital Lahore.
is observed more frequently in minority populations, people
with English as a second language, elderly people, and people Study population
with a low income and poor education.7,8
Interestingly, the same populations (i.e. elderly, low in- Potential subjects were identified from the hospital clinical
come and poorly educated) with low health literacy carry the databases. Subjects were screened using the inclusion and
highest burden of chronic conditions, such as diabetes and exclusion criteria for suitability for inclusion in this study.
hypertension.5,8 Diabetes is a chronic disease with profound
complexity requiring rigorous self-care, education and man- Inclusion criteria
agement, often relying on printed education materials All patients with type 2 diabetes (controlled or uncontrolled),
regarding diet and self-care practices complemented by verbal older than 30 years, speaking Punjabi, Urdu or English, with or
instructions, and therefore requiring advanced health literacy without complications and without any mental health issues
skills.9 Studies in patients with diabetes have shown that were included in the study. As per the database records, the
limited health literacy coupled with poor knowledge of the participant should have visited a primary care physician at
disease is associated with a poorer outcome.10,11 Similarly, least once in the six months preceding their enrolment in the
studies have shown that adherence to diet, exercise and study in an attempt to ensure that subjects were routinely and
pharmacotherapy is pivotal for optimal glucose control in uniformly managed by healthcare professionals.
patients with diabetes and is related to patient knowledge
about self-care, self-efficacy and self-management of dia- Exclusion criteria
betes.12,13 More precisely, differences in self-care and self- Participants with end-stage renal disease, blindness,
management are associated with inconsistencies in treat- compromised mental health and any health condition that
ment outcomes.14 can hinder or interfere with the accurate measurement of
In developing countries, such as Pakistan, where the ma- health literacy were excluded from the study. In addition,
jority of people have low incomes and do not have easy access participants who were unable or unwilling to provide
to education, health literacy is an unexplored entity, created informed consent were excluded from the study.
and refined by the developed world. The literature indicates Socio-economic status was classified as follows:
that the prevalence of diabetes is higher in South Asians Lower: lived in rural area, informal occupation, no formal
compared with Caucasians, and the estimated prevalence of education, and annual income of 0e299,999 PKR.
diabetes in Pakistan is 7.1%, placing this in the seventh high- Middle: lived in urban area, early-stage professionals,
est position globally.15e17 However, to the authors' knowledge, formal education (college/university), and annual income of
no studies to date have evaluated the association between 300,000e999,999 PKR.
functional health literacy and diabetes outcomes in Pakistan. Upper: lived in urban area, inherited income-generating
Therefore, the authors used Short Test of Functional Health assets, minimum bachelor's degree from top-tier university
Literacy (s-TOFHLA) to assess the impact of health literacy on or foreign university, a certain lifestyle and annual income of
glycaemic control of patients with diabetes in Lahore, 1000,000 PKR.
Pakistan, and estimated the relationships between patient
characteristics and glycated haemoglobin (HbA1C) level. Data collection

Between August 2015 and January 2016, three Pharm D stu-

Methods dents, adept in data collection and with comprehensive
knowledge of the disease and the project, administered the s-
Ethical approval TOFHLA to assess the health literacy levels of patients.18 For
Urdu-speaking participants, the English version of the s-
Ethical approval for this study was obtained from the Ethics TOFHLA (36 items) was translated into Urdu (the national
Committee of University College of Pharmacy, University of language of Pakistan) by a language qualified exegete and was
the Punjab (Reference No. EC/UCP/092/2015) and Hospital translated back into English for content verification and to
Committee of Ethics on Human Research. attain concordance in meaning between English and Urdu
versions. A pilot study was conducted to estimate the level of
Study design concordance by comparing data obtained from both the En-
glish and Urdu versions. No major differences were observed
Two hundred and four patients with diabetes were enrolled in the information obtained from the two versions of the
into this six-month cross-sectional study from hospitals in questionnaire, so the data provided have a concordance level
Lahore, Pakistan. Patients with diabetes of more than 12 exceeding 97%. The s-TOFHLA is administered as a timed
months' duration were identified from four tertiary care reading comprehension test with four multiple choice
10 p u b l i c h e a l t h 1 5 6 ( 2 0 1 8 ) 8 e1 4

options. It is scored using a scale of 0e36, with 0e16 indicating middle (58.1%) and upper socio-economic status (38.7%) (Table
inadequate health literacy, 17e22 indicating marginal health 1). As expected, almost 92% of patients with inadequate health
literacy, and 23e36 indicating adequate health literacy. A literacy were in the third and fourth HbA1C quartiles (8.6% and
questionnaire was also administered regarding demographic >10.6%). On the other hand, 29% of patients with adequate
information, duration of disease, presence of risk factors for health literacy were in the lowest HbA1C quartile (6.5%)
diabetes, and complications, along with current blood pres- (Fig. 1D). Moreover, no significant differences in frequency
sure and HbA1C values. distribution were noted between patients on oral hypo-
glycaemic agents (OHAs), insulin and a combination (i.e.
Data analysis OHA þ insulin) when assessed by health literacy level (i.e.
adequate, marginal and inadequate) (P ¼ 0.234) (Table 1).
All data were analysed using Graphpad (Prism5) and SPSS,
version 21 (IBM Corp., Armonk, NY, USA), unless otherwise Association of influencing variables and relationships
reported. Health literacy was analysed as a continuous vari- between patient characteristics and HbA1C level
able. Regression analysis was performed to assess the asso-
ciation between health literacy score (s-TOFHLA) and HbA1C Associations between patient characteristics (predictors) and
level. Linear regression was performed to explore the re- HbA1C level were examined. No relationship was found be-
lationships between dependent variables and influencing in- tween HbA1C level and several predictors of glycaemic control
dependent variables after controlling for age, sex, education, (i.e. age, sex, employment, socio-economic status and dura-
employment status and socio-economic status. An alpha tion of disease) (Table 2). Only s-TOFHLA score, disease risks
value of less than 0.05 was considered to indicate statistical and education were found to be significantly associated with
significance. HbA1C level (Table 2). After adjustment for age, sex, education,
employment and socio-economic status, only s-TOFHLA score
was independently associated with HbA1C level (Table 2).
Results Interestingly, 29% of patients with adequate health literacy
had tight glycaemic control (HbA1C  6.5%) (odds ratio [OR] ¼
Six hundred forty-two patients were identified from the dia- 0.017, 95% confidence interval [CI] ¼ 0.0e0.15, P ¼ 0.0005)
betes clinics of four tertiary care hospitals. Of these, 229 pa- compared with 0.73% of patients with inadequate health lit-
tients had not visited their physician in the preceding year, eracy (Table 1 and Fig. 1D). On the other hand, almost 92% of
137 patients did not meet the inclusion and exclusion criteria, patients with inadequate health literacy had poor glycaemic
27 patients did not have type 2 diabetes and 24 patients were control, while none of the patients with adequate health lit-
no longer alive. In total, 225 subjects were approached eracy had very poor glycaemic control (i.e. third and fourth
regarding enrolment in the study, and 21 did not provide quartiles) (Table 1 and Fig. 1D).
informed consent. As such, 204 eligible patients were evalu-
ated in terms of health literacy and its impact on diabetes Health literacy level and diabetes complications
outcomes.
Next, diabetes complicationsdmacrovascular (peripheral ar-
Basic demographic and clinical characteristics of patients tery disease [PAD] ischaemic heart disease [IHD]) and micro-
vasculardwere evaluated against inadequate and adequate
Table 1 shows the basic demographic characteristics of pa- health literacy. As shown in Table 3, 47.4% of patients with
tients. Overall, 67.2% of patients had inadequate health liter- inadequate health literacy had retinopathy (OR ¼ 13.1, 95% CI ¼
acy (s-TOFHLA score 0e16), 17.6% had marginal health literacy 3e57.03, P ¼ 0.003) compared with only 6.5% of subjects with
(s-TOFHLA score 17e22) and 15.2% had adequate health lit- adequate health literacy. However, the wide CIs could be due to
eracy (s-TOFHLA score 23e36) (Table 1). The majority of pa- the small sample size, which may have affected the precision
tients with inadequate health literacy were aged more than 45 of the gathered information. Likewise, patients with inade-
years (54.7%), female (54%), primary (21.2%) and higher sec- quate health literacy were more likely to have microvascular
ondary (45.3%) education, lower socio-economic status (46%), (nephropathy: OR ¼ 3.5, P ¼ 0.274; neuropathy: OR ¼ 1.3,
middle socio-economic status (44.5%) and disease duration of P ¼ 0.992) and macrovascular complications (PAD: OR ¼ 3.5,
less than 10 years (78.1%) (Table 1). Subjects with inadequate P ¼ 0.523; IHD: OR ¼ 4.8, P ¼ 0.741) than patients with adequate
health literacy were more likely to have HbA1C level above health literacy (Table 3), although this difference was not sig-
8.6%, two disease-related risk factors (78.1%), more than two nificant. When assessed for overall differences in complica-
(5.8%) disease-related risk factors and diabetes complications tions, no differences in microvascular (OR ¼ 0.9, P ¼ 0.99) and
(microvascular [54.7%], macrovascular [16.1%] and combina- macrovascular (OR ¼ 1.8, P ¼ 0.972) complications were
tion of both [29.2%]) (Fig. 1AeD). Likewise, of all reported observed between patients with inadequate and adequate
complications, the highest reported frequencies of diabetes health literacy (Fig. 1A). However, patients with inadequate
complications (microvascular and macrovascular) were health literacy were more likely to report a combination of both
observed in subjects with inadequate health literacy (Fig. 1B). microvascular and macrovascular complications (OR ¼ 12.4,
On the other hand, the majority of patients with adequate P ¼ 0.10) than patients with adequate health literacy (Fig. 1A).
health literacy were younger (<45 years, 67.7%), female (71%), Moreover, no self-reported complications were observed in
higher secondary education (48.4%), employed (71%), and patients with adequate health literacy (Fig. 1A).
 p u b l i c h e a l t h 1 5 6 ( 2 0 1 8 ) 8 e1 4 11

Table 1 e Patient characteristics and health literacy.

Characteristics Health literacy level
Inadequate (n ¼ 137) Marginal (n ¼ 36) Adequate (n ¼ 31) P-value
Age (years)
<45 61 (44.5%) 26 (72.2%) 21 (67.7%) 0.003*
>45 75 (54.7%) 10 (27.8%) 10 (32.3%)
Sex
Male 63 (46%) 13 (36.1%) 9 (29%) 0.170
Female 74 (54%) 23 (63.9%) 22 (71%)
Education
Illiterate 25 (18.2%) 0 0 0.0005**
Primary 29 (21.2%) 2 (5.6%) 0
Secondary 19 (13.9%) 6 (16.7%) 0
Higher secondary 62 (45.3%) 24 (66.6%) 15 (48.4%)
College/university 2 (1.5%) 4 (11.1%) 0
Employment status
Employed 86 (62.8%) 24 (66.6%) 22 (71%) 0.665
Unemployed 51 (37.2%) 12 (33.3%) 9 (29%)
Socio-economic status
Lower 63 (46%) 11 (30.5%) 1 (3.2%) 0.0005**
Middle 61 (44.5%) 20 (55.5%) 18 (58.1%)
Upper 13 (9.5%) 5 (14%) 12 (38.7%)
Genetic predisposition
Yes 20 (14.6%) 8 (22.2%) 8 (25.8%) 0.245
No 117 (85.4%) 28 (77.7%) 23 (74.2%)
Therapeutic options
OHA 28 (20.4%) 6 (16.7%) 9 (29%) 0.234
Insulin 61 (44.1%) 16 (44.4%) 7 (22.6%)
Insulin þ OHA 48 (35%) 14 (38.9%) 15 (48.4%)
Duration (years)
<10 107 (78.1%) 30 (83.3%) 26 (84%) 0.656
>10 30 (21.9%) 6 (16.7%) 5 (16%)

OHA, oral hypoglycaemic agents.

*
P-values 0.05e0.002.
**
P-values <0.002.

patients with diabetes in Lahore, suggesting that the s-

Discussion TOFHLA score was significantly associated with patient edu-
cation. Moreover, most health-related information is
These data demonstrated that inadequate health literacy is disseminated in English, which is the second language for the
likely to be associated with poor glycaemic control, with majority of patients enrolled in this study, which makes it
higher odds of diabetes complications, particularly retinop- even more difficult for patients to extricate and assimilate
athy. Furthermore, only patients with adequate health liter- relevant health- or disease-related information that is vital for
acy reported no complications and had higher socio-economic self-monitoring and self-care practices; this is corroborated by
status. It appears that the socio-economic status of patients a previous report.7 Numerous studies have shown that low
might be one independent variable that could affect health health literacy is associated with poor glycaemic control, yet
literacy, directly via education and indirectly via access to the association did not sustain when adjusted for diabetes
digital medical content using smartphones and tab- knowledge.24 The present study did not evaluate diabetes
letsdunaffordable for many in Pakistan. knowledge, and it is therefore not possible to support or
Poor disease knowledge due to limited health literacy, negate these claims, which seem quite pertinent; however,
common in patients with chronic diseases, has been associ- data from previous studies did not support this notion.10
ated with poorer outcomes.10,19 In patients with diabetes Diabetes care and management require efficient transfer
requiring continual self-care activities, differences in self-care and conceptualisation of several skills (e.g. self-monitoring of
behaviours can contribute towards inconsistency in treat- blood glucose, managing multiple medicines, multiple pro-
ment outcomes, as demonstrated by several studies.5,14,20 An viders, foot care, adherence to diet plan and physical exercise)
adult's literacy level has been shown to correlate positively deftly conveyed by a healthcare professional and need to be
with health literacy, making it easier to relay important health practised and performed by the patient. Therefore, any issues
information to people with higher literacy levels.21,22 Unfor- or problems in communication and knowledge empowerment
tunately, in Pakistan, the adult literacy rate is just 55% (men affecting self-monitoring and care by patients could lead to
67% and women 42%), hindering the effective delivery of improper diabetes care and management.25 It appears that the
health education regarding disease and its management.23 factors that govern the quality of diabetes care are complex
The data from the present study support these findings in and multiple, entailing scrupulous interactions at various
12 p u b l i c h e a l t h 1 5 6 ( 2 0 1 8 ) 8 e1 4

Fig. 1 e Population frequency distribution of health literacy levels in patients with type 2 diabetes in Lahore. (A) Diabetes
complications: microvascular, macrovascular, combination (microvascular þ macrovascular) and no complications.
(B) Frequency distribution of microvascular complications (neuropathy, nephropathy and retinopathy) and macrovascular
complications (diabetic foot and ischaemic heart disease) according to health literacy level. (C) Diabetes risks: single
(obesity, age, genetic predisposition, hypertension, sedentary lifestyle and insulin resistance) and multiple (combination of
options mentioned as single risks). (D) Health literacy level according to glycated haemoglobin (HbA1C) quartile (<6.5%,
6.6e8.5%, 8.6e10.6%, >10.6%).

levels (patients, physicians, health system, family and evaluate any of these interactions, particularly self-care
community).26e28 Thus, minor issues at any of these levels practices, to determine the association of these self-care
during interactions with healthcare providers would result in practices with functional health literacy. Besides, in
suboptimal self-monitoring and care practices. Furthermore, Pakistan, where literacy is defined by one's ability to produce
studies have shown that patients with poor health literacy are his/her signature, it is highly unlikely that the majority of the
less likely to interpret and act on self-monitoring results, population deemed to be literate would be able to compre-
further implicated by intricate multiplicity of self-care activ- hend highly advanced technical knowledge about the disease
ities demanding knowledge empowerment via a timely and practice self-care skills to ensure proper self-monitoring
transferable source.10,29 However, the present study did not and care. Additionally, as reported previously,5 the present

Table 2 e Associations between patient characteristics and glycated haemoglobin (HbA1C) level.
Predictors Unadjusted Adjusteda
Coefficients (95% CI) P-value Coefficients (95% CI) P-value
**
s-TOFHLA score 0.712 (0.17 to 0.07) 0.0005 0.0387 (0.125 to 0.032) 0.001**
Age 0.096 (0.20 to 0.86) 0.195 0.002 (0.42 to 0.41) 0.975
Sex 0.005 (0.65 to 0.66) 0.950 0.041 (0.30 to 0.61) 0.475
Education 0.387 (1 to 0.5) 0.0005** 0.105 (0.41 to 0.018) 0.072
Employment 0.035 (0.55 to 0.84) 0.676 0.072 (0.19 to 0.78) 0.226
Socio-economic status 0.074 (0.21 to 0.63) 0.312 0.058 (0.47 to 0.13) 0.272
Duration 0.393 (0.83 to 0.55) 0.695 0.008 (0.45 to 0.53) 0.870
Risks 0.181 (0.10 to 1.31) 0.005* 0.079 (0.082 to 1.1) 0.091

s-TOFHLA, Short Test of Functional Health Literacy in Adults; CI, confidence interval.
*
P-values 0.05e0.002.
**
P-values <0.002.
a
Adjusted for age, sex, education, employment and socio-economic status.
 p u b l i c h e a l t h 1 5 6 ( 2 0 1 8 ) 8 e1 4 13

findings with a population of similar geographic, socio-

Table 3 e Odds of diabetes complications: inadequate vs
economic and ethnic background.
adequate health literacy.
In conclusion, these data from patients with type 2 diabetes,
Complications Frequency of Odds ratio P-value
reporting to diabetes centres of tertiary care hospitals of
complications (95% CI)
Lahore, Pakistan, showed that health literacy was potentially
Microvascular complications
associated with HbA1C level. Furthermore, inadequate health
Retinopathy 65 13.1 (3e57.03) 0.003*
literacy was an independent predictor of poor glycaemic con-
Nephropathy 55 3.5 (1.26e9.64) 0.274
Neuropathy 67 1.3 (0.61e2.91) 0.992 trol and was associated with a lower tendency to attain tight
Macrovascular complications glycaemic control. Similarly, inadequate health literacy was
Peripheral artery 37 3.5 (0.99e12.4) 0.523 associated with higher odds of retinopathy and other compli-
disease cations compared with subjects with adequate health literacy.
Ischaemic heart 19 4.8 (0.62e37.54) 0.741 Therefore, health professionals, particularly pharmacists
disease
(the most underutilised group of health professionals in
CI, confidence interval. Pakistan) must be involved in disseminating disease-related
*
P-values 0.05e0.002. education and counselling to improve the functional health
literacy of patients in terms of chronic disease self-monitoring
and care practices in hospital and community settings.
data showed a significant association between lower health
literacy and retinopathy. Although this association was not
significant, patients with poor health literacy had a higher Author statements
likelihood of nephropathy, neuropathy, PAD and IHD
compared with patients with adequate health literacy. In Acknowledgements
Pakistan, lower health literacy, diabetes care and manage-
ment can be affected by socio-economic and cultural motifs, The authors wish to thank the study participants and hospital
such as influential social gatherings affecting disease man- staff for their cooperation.
agement, alternative medicine options (suggested by friends
Ethical approval
and family members) and financial issues, duly supported by
the findings of the present study (i.e. higher frequency of
Ethical approval for the study was obtained from the Ethics
inadequate health literacy in subjects of lower socio-
Committee of University College of Pharmacy, University of
economic status). Surprisingly, most of the patients of
the Punjab (Reference No. EC/UCP/092/2015) and Hospital
higher socio-economic status agreed that digital medical
Committee of Ethics on Human Research.
content accessibility via smartphones and tablets helped
them to improve their health literacy regarding diabetes. Funding
Therefore, further studies should be conducted to evaluate
self-care practices of patients with diabetes and pertinent None declared.
interacting factors between patient and healthcare providers
Competing interests
with regards to the functional health literacy of patients to
determine plausible barriers and missing links that affect not None declared.
only the transfer of health information, affecting knowledge
empowerment, but also patients' self-care practices and
references
management. Additionally, s-TOFHLA is a comprehension-
based questionnaire and did not test the patient's ability to
recall disease management and is therefore unlikely to be
1. Davis TC, Wolf MS. Health literacy: implications for family
affected by patient's recall bias. medicine. Fam Med 2004;36:595e8.
This study has several limitations. First, it was not possible 2. Paasche-Orlow MK, Parker RM, Gazmararian JA, Nielsen-
to examine the effect of inadequate health literacy over an Bohlman LT, Rudd RR. The prevalence of limited health
extended period of time due to the cross-sectional design of literacy. J Gen Intern Med 2005;20:175e84.
this study, and it was not possible to evaluate the knowledge 3. Safeer RS, Keenan J. Health literacy: the gap between
of diabetes among subjects enrolled in the study and relate physicians and patients. Am Fam Physician 2005;72:463e8.
4. Mayeaux Jr EJ, Murphy PW, Arnold C, Davis TC, Jackson RH,
this to overall health literacy, medication adherence and gly-
Sentell T. Improving patient education for patients with low
caemic control. Moreover, the study did not take into account literacy skills. Am Fam Physician 1996;53:205e11.
the field of education and access to digital content via the 5. Schillinger D, Grumbach K, Piette J, Wang F, Osmond D,
Internet, which might affect the overall health literacy levels Daher C, et al. Association of health literacy with diabetes
of the enrolled subjects. Furthermore, lipid profiles and sub- outcomes. JAMA 2002;288:475e82.
categorisation of the complications were not documented 6. Berkman ND, Sheridan SL, Donahue KE, Halpern DJ, Crotty K.
Low health literacy and health outcomes: an updated
routinely and were not included in the data. Moreover, it is not
systematic review. Ann Intern Med 2011;155:97e107.
possible to generalise the findings of this study due to the
7. Al Sayah F, Majumdar SR, Williams B, Robertson S,
limited sample size and the fact that all patients lived in one Johnson JA. Health literacy and health outcomes in diabetes:
city (i.e. Lahore). Unfortunately, no other studies from a systematic review. J Gen Intern Med 2013;28:
Pakistan are available, so it is not possible to compare these 444e52.
14 p u b l i c h e a l t h 1 5 6 ( 2 0 1 8 ) 8 e1 4

8. Sudore RL, Mehta KM, Simonsick EM, Harris TB, Newman AB, 19. Gazmararian JA, Williams MV, Peel J, Baker DW. Health
Satterfield S, et al. Limited literacy in older people and literacy and knowledge of chronic disease. Pat Educ Counsel
disparities in health and healthcare access. J Am Geriatr Soc 2003;51:267e75.
2006;54:770e6. 20. Williams MV, Baker DW, Honig EG, Lee TM, Nowlan A.
9. White RO, Wolff K, Cavanaugh KL, Rothman R. Addressing Inadequate literacy is a barrier to asthma knowledge and self-
health literacy and numeracy to improve diabetes education care. Chest J 1998;114:1008e15.
and care. Diabet Spect 2010;23:238e43. 21. Zarcadoolas C, Pleasant A, Greer DS. Understanding health
10. Williams MV, Baker DW, Parker RM, Nurss JR. Relationship of literacy: an expanded model. Health Promot Int 2005;20:195e203.
functional health literacy to patients' knowledge of their 22. McCray AT. Promoting health literacy. J Am Med Inform Assoc
chronic disease: a study of patients with hypertension and 2005;12:152e63.
diabetes. Arch Intern Med 1998;158:166e72. 23. Ayub RA, Jaffery T, Aziz F, Rahmat M. Improving health
11. Norris SL, Engelgau MM, Narayan KV. Effectiveness of self- literacy of women about iron deficiency anemia and civic
management training in type 2 diabetes a systematic review responsibility of students through service learning. Educ
of randomized controlled trials. Diabetes Care 2001;24:561e87. Health 2015;28:130.
12. Iannotti RJ, Schneider S, Nansel TR, Haynie DL, Plotnick LP, 24. Bains SS, Egede LE. Associations between health literacy,
Clark LM, et al. Self-efficacy, outcome expectations, and diabetes knowledge, self-care behaviors, and glycemic
diabetes self-management in adolescents with type 1 control in a low income population with type 2 diabetes.
diabetes. J Dev Behav Pediatr 2006;27:98e105. Diabetes Technol Ther 2011;13:335e41.
13. Inoue M, Takahashi M, Kai I. Impact of communicative and 25. Anderson RM, Funnell MM. Compliance and adherence are
critical health literacy on understanding of diabetes care and dysfunctional concepts in diabetes care. Diabetes Educ
self-efficacy in diabetes management: a cross-sectional study 1999;26:597e604.
of primary care in Japan. BMC Fam Pract 2013;14:1. 26. Von Korff M, Gruman J, Schaefer J, Curry SJ, Wagner EH.
14. Kim S, Love F, Quistberg DA, Shea JA. Association of health Collaborative management of chronic illness. Ann Intern Med
literacy with self-management behavior in patients with 1997;127:1097e102.
diabetes. Diabetes Care 2004;27:2980e2. 27. Fisher L, Chesla CA, Bartz RJ, Gilliss C, Skaff MA, Sabogal F,
15. Sabir S, Hassan F. Improving health literacy in Pakistan e 'a et al. The family and type 2 diabetes: a framework for
new oil in old lanterns'. J Pak Med Assoc 2013;63:539. intervention. Diabetes Educ 1998;24:599e607.
16. Mather HM, Keen H. The Southall Diabetes Survey: 28. McCulloch DK, Price MJ, Hindmarsh M, Wagner EH.
prevalence of known diabetes in Asians and Europeans. BMJ Improvement in diabetes care using an integrated
1985;291:1081e4. population-based approach in a primary care setting. Dis
17. Akhter J. The burden of diabetes in Pakistan: the national Manag 2000;3:75e82.
diabetes survey. J Pak Med Assoc 1999;49:205. 29. Schillinger D, Piette J, Grumbach K, Wang F, Wilson C,
18. Parker RM, Baker DW, Williams MV, Nurss JR. The test of Daher C, et al. Closing the loop: physician communication
functional health literacy in adults. J Gen Intern Med with diabetic patients who have low health literacy. Arch
1995;10:537e41. Intern Med 2003;163:83e90.