Edited by: Vesna Bjegovic-Mikanovic, University of Belgrade, Serbia
Reviewed by: Margo Bergman, University of Washington Tacoma, United States; Tetine Sentell, University of Hawaii, United States; Lijana Zaletel-Kragelj, University of Ljubljana, Slovenia
This article was submitted to Public Health Education and Promotion, a section of the journal Frontiers in Public Health
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Health status is largely influenced by health behavior through complex processes involving multi-channel multi-layer interactions between individual socio-demographic factors and numerous external factors, influenced and modeled by cognitive skills and internal motivation and attitude toward certain behaviors (
Maintaining the levels of blood glucose within normal range is important to health. As the levels of blood glucose fluctuate greatly during everyday life activities (
After the collapse of the communist regime in 1990, Albania is undergoing deep political and socioeconomic reforms which are also associated with changes in the epidemiological profile and health characteristics (
Scientific research has reported significant inverse associations between HL and mean hemoglobin A1c (HbA1c) and mean plasma glucose levels, in both diabetic (
However, some other studies did not find significant associations of HL with plasma glucose levels and glycemic control (
In Albania and other Western Balkan countries there are no studies addressing comprehensively the association between HL and population fasting plasma glucose levels. A previous population-based study conducted in Albania has reported a significant association between HL and body mass index (BMI), irrespective of socio-demographic factors and socio-economic characteristics (
This was a cross-sectional study carried out during the period September 2012—February 2014.
The study population included a population-drawn simple random sample of 1,500 adult men and women (≥18 years), selected from the registries of family physicians operating in primary health care services of the city of Tirana, which is the capital of Albania. Further details about the study population and the sampling technique are reported elsewhere (
Data collection process included measurement of glucose level, BMI and administration of a structured questionnaire to all study participants.
Glucose level was measured by use of a calibrated glucometer (finger stick method, a rapid glucose test kit). Glucose was measured in a fasting state in all study participants (
Distribution of blood glucose level among study participants.
Mean | 102.67 |
Standard deviation | 28.76 |
Median | 97 |
Interquartile Range | 87–110 |
Normal (<100 mg/dl) | 478 (55.8) |
Pre-diabetes (100–125.9 mg/dl) | 285 (33.3) |
Diabetes (≥126 mg/dl) | 94 (11.0) |
No diabetes | 763 (89.1) |
Diabetes | 94 (11.0) |
The original full version of the HLS-EU-Q instrument (
Each of the HL items assessed the self-perceived difficulty of performing selected health-related tasks on a 4-point scale ranging from very easy (one) to very difficult (four) (
Anthropometrics included weight (with precision of 100 g) which was measured in light clothes using a calibrated beam balance and height (with precision of 1 mm) which was measured using a tape attached to the wall with subjects not wearing shoes. BMI was calculated as weight (in kg) / height (in m2).
Furthermore, data on a full range of socio-demographic and socioeconomic characteristics was collected for all study participants, including age (in the analysis, categorized into: ≤25, 26–45, 46–65, and ≥66 years), sex (men vs. women), marital status (in the analysis, dichotomized into: married vs. single, divorced, and widowed), employment status (dichotomized into: unemployed vs. employed and/or retired), educational attainment (0–8, 9–12, and ≥13 years), economic status (trichotomized into: very bad/bad, average, and good/very good), and social status (low, middle, and high).
All participants signed an informed consent form after being explained the aims and procedures of the survey. The study was approved by the Albanian Committee of Bio-Medical Ethics on 23 July 2012.
For 297 individuals there was missing information on glucose level, BMI or other covariates. These cases were excluded from the analysis; hence, the statistical analysis consisted of 1,154–297 = 857 individuals.
Measures of central tendency and dispersion were calculated for the glucose level.
The chi-square test was used to compare the distribution of glucose levels (no diabetes vs. diabetes) according to socio-demographic characteristics (age, sex, marital status, employment, education, and economic and social status), BMI and HL of study participants.
General Linear Model was used to calculate mean values of glucose level among individuals distinguished by different HL categories (inadequate, problematic, sufficient and excellent). Initially, crude (unadjusted) mean values, their 95% confidence intervals (95%CIs) and
In addition, binary logistic regression was used to assess the independent association of glucose level (dependent variable, dichotomized into: diabetes vs. no diabetes) with HL levels (inadequate, problematic, sufficient, and excellent). Crude (unadjusted) odds ratios (ORs), their respective 95% confidence intervals (95%CIs) and
In all cases, a
Statistical Package for Social Sciences (SPSS, version 19.0) was used for all the statistical analyses.
On the whole, mean age in this study sample was 45.5 ± 16.4 years; 57% of participants were women; mean years of formal schooling were 12.6 years; about 82% of study participants perceived themselves as middle class, and about two-thirds (64%) reported an average economic status (
Overall, mean glucose level among study participants was 103±29 mg/dl (
Men had a higher prevalence of measured diabetes than women (13 vs. 10%), but this finding was not statistically significant (
Distribution of blood glucose level by socio-demographic characteristics, HL and BMI of study participants.
0.187 |
||||
Men | 373 (43.5) |
326 (87.4) | 47 (12.6) | |
Women | 484 (56.5) | 437 (90.3) | 47 (9.7) | |
<0.001 | ||||
≤ 25 years | 113 (13.2) | 113 (100.0) | 0 (–) | |
26-45 years | 249 (29.1) | 241 (968) | 8 (3.2) | |
46-65 years | 388 (45.3) | 328 (84.5) | 60 (15.5) | |
≥66 years | 107 (12.5) | 81 (75.7) | 26 (24.3) | |
0.139 | ||||
0-8 years | 125 (14.6) | 107 (85.6) | 18 (14.4) | |
9-12 years | 442 (51.6) | 390 (88.2) | 52 (11.8) | |
≥13 years | 290 (33.8) | 266 (91.7) | 24 (8.3) | |
0.257 | ||||
Unemployed | 160 (19.2) | 145 (90.6) | 15 (9.4) | |
Employed and/or retired | 672 (80.8) | 594 (88.4) | 78 (11.6) | |
0.008 | ||||
Not married |
293 (34.8) | 272 (92.8) | 21 (7.2) | |
Married | 549 (65.2) | 477 (86.9) | 72 (13.1) | |
0.216 | ||||
Low | 89 (11.1) | 80 (89.9) | 9 (10.1) | |
Middle | 649 (81.2) | 570 (57.8) | 79 (12.2) | |
High | 61 (7.6) | 58 (95.1) | 3 (4.9) | |
0.047 | ||||
Very bad/bad | 106 (13.2) | 90 (84.9) | 16 (15.1) | |
Average | 527 (65.4) | 463 (87.9) | 64 (12.1) | |
Good/very good | 173 (21.5) | 162 (93.6) | 11 (6.4) | |
<0.001 | ||||
Normal | 288 (33.6) | 272 (94.4) | 16 (5.6) | |
Overweight | 377 (44.0) | 337 (89.4) | 40 (10.6) | |
Obese | 192 (22.4) | 154 (80.2) | 38 (19.8) | |
<0.001 | ||||
Inadequate | 160 (20.0) | 128 (80.0) | 32 (20.0) | |
Problematic | 152 (19.0) | 129 (84.9) | 23 (15.1) | |
Sufficient | 213 (26.7) | 190 (89.2) | 23 (10.8) | |
Excellent | 274 (34.3) | 259 (94.5) | 15 (5.5) |
There was evidence of a mild positive linear association between glucose level and BMI (Spearman's rho = 0.27,
Mean unadjusted glucose levels were significantly lower among participants with excellent HL levels compared with those with inadequate HL levels (98 vs. 111 mg/dl) (
Association of blood glucose level with HL (General Linear Models).
Inadequate | 110.9 | 106.4–115-5 | <0.001 |
Problematic | 105.9 | 101.2–110.5 | 0.006 |
Sufficient | 102.2 | 98.3–106.1 | 0.095 |
Excellent | 97.7 | 94.3–101.2 | reference |
Inadequate | 105.3 | 100.8–109.8 | 0.012 |
Problematic | 103.1 | 98.5–107.7 | 0.069 |
Sufficient | 101 | 97.1–104.9 | 0.222 |
Excellent | 97.9 | 94.3–101.5 | reference |
Inadequate | 105.8 | 101.2–110.3 | 0.015 |
Problematic | 103.6 | 98.9–108.2 | 0.086 |
Sufficient | 101 | 97.1–105.0 | 0.347 |
Excellent | 98.6 | 94.9–102.2 | reference |
Inadequate | 105.2 | 99.4-111.0 | 0.034 |
Problematic | 103.2 | 91.1-109.2 | 0.123 |
Sufficient | 100.4 | 94.5-106.3 | 0.472 |
Excellent | 98.4 | 92.9-103.7 | reference |
Inadequate | 106 | 100.2–111.8 | 0.039 |
Problematic | 103.7 | 97.6–109.7 | 0.161 |
Sufficient | 101.1 | 95.2–107.0 | 0.514 |
Excellent | 99.3 | 93.9–104.7 | reference |
Conversely, in unadjusted binary logistic regression models (
Association of HL with blood glucose level; multivariable-adjusted odds ratios (ORs: diabetes vs. no diabetes) from binary logistic regression.
Inadequate | 4.32 | 2.26-8.26 | <0.001 |
Problematic | 3.08 | 1.55-6.10 | 0.001 |
Sufficient | 2.09 | 1.06-4.11 | 0.033 |
Excellent | 1 | Reference | - |
Inadequate | 2.73 | 1.38-5.44 | 0.004 |
Problematic | 2.44 | 1.21-4.95 | 0.013 |
Sufficient | 1.88 | 0.94-3.77 | 0.074 |
Excellent | 1 | Reference | - |
Inadequate | 2.77 | 1.39-5.49 | 0.004 |
Problematic | 2.46 | 1.21-4.99 | 0.013 |
Sufficient | 1.84 | 0.92-3.69 | 0.086 |
Excellent | 1 | Reference | - |
Inadequate | 2.6 | 1.26-5.37 | 0.010 |
Problematic | 2.23 | 1.08-4.62 | 0.031 |
Sufficient | 1.67 | 0.82-3.41 | 0.159 |
Excellent | 1 | Reference | - |
Inadequate | 2.62 | 1.26-5.44 | 0.01 |
Problematic | 2.15 | 1.03-4.48 | 0.041 |
Sufficient | 1.65 | 0.80-3.38 | 0.174 |
Excellent | 1 | Reference | - |
In the General Linear Models, adjustment for age (
In age-adjusted binary logistic regression models (
In this population-based sample of adult men and women in transitional Albania, we obtained evidence of a significant inverse association between measured blood glucose level and HL, which was assessed based on a well-established international instrument already validated in Albania (
Our findings, in general, are in line with previous reports from international research.
A study among 228 individuals aged 30 years or older and seeking care at emergency department of a hospital in Georgetown, Guyana, reported that mean blood glucose level was higher among lower HL subjects (128.3 mg/dl) compared to high HL subjects (117.1 mg/dl), but the difference was not statistically significant (
In a population-based study of elderly people aged 70–79 years, there was a reverse significant association between health literacy and mean fasting blood glucose among women, but not in men (
Our findings are also compatible with another population-based study including 1817 Japanese individuals (
Evidence of significant and inverse associations between health literacy and fasting plasma glucose is provided by another recent paper reporting on the associations of HL with an array of laboratory parameters among individuals aged 23–88 years receiving health checkup in Taipei, Taiwan (
Previous publications about this study population (among 1,154 individuals) in Albania have revealed inverse and significant associations of mean health literacy score with age, education and low social and economic status (
Besides fasting plasma glucose levels we also reported about the prevalence of prediabetes (33.3%) and diabetes (11%) in this study population group. We found that diabetes prevalence (including prediabetes and diabetes individuals) was higher among married individuals and it was significantly and positively associated with age, but inversely associated with economic status, body mass index (BMI) and health literacy. Prediabetes is considered a high-risk condition for progression to full diabetes as one-quarter of affected persons will develop it in the next 5 years and more than two-thirds will do so in their lifetime (
According to Centers for Disease Control and Prevention (CDC) the prevalence of prediabetes among American adults aged ≥18 years was 33.9% in 2015 (
Prevalence rate of prediabetes and diabetes vary according to different factors including education level, social and economic status, prevalence of overweight and obesity, access to healthcare services, etc. (
In our study population we reported that being married was significantly associated with a higher prevalence of diabetes, a finding supported by international research (
The association of inadequate HL and diabetes is also supported by international literature (
Given the unfavorable position of low HL diabetic patients, it is necessary to pinpoint strategies and interventions that would help overcome this barrier. Such efforts include specifically tailored toolkits, guides or use of new online technologies intended to facilitate patient-provider communication (
Policymakers, decision makers and health professionals need to be aware about these effective and beneficial strategies, the application of which could facilitate the interaction of low health literacy patients and individuals with the health system.
There are some potential limitations of this study including the possibility of selection bias, information bias, timeline of the study and its cross-sectional design. On the whole, the response rate of this study was high (89%), but for 297 individuals there was no valid information on covariates including blood glucose level. In addition, at best, findings from this study can be generalized to the adult population of Tirana only given the fact that the survey was confined to this region. Blood glucose level and BMI were objectively assessed in all participants, which is reassuring. However, one measurement is not sufficient to establish the diagnosis of diabetes and, usually, other parameters are used (2-h Oral Glucose Tolerance Test [OGTT] and HbA1c testing)—posing another limitation of this study. Also, the instrument for assessment of HL was based on a well-developed and standardized tool (
In conclusion, regardless of these potential limitations, we obtained evidence of a strong and significant inverse relationship between measured blood glucose level and HL, independent of many socio-demographic characteristics and measured BMI in a population-based sample of adult men and women in a post-communist country.
In the context of current uncertainty prevailing in the international literature about the relationship between HL and blood glucose level, the present survey conducted in transitional Albania adds to the body of literature that supports a significant association between HL and fasting plasma glucose levels.
The datasets generated for this study are available on request to the corresponding author.
The studies involving human participants were reviewed and approved by Albanian Committee of Bio-Medical Ethics. The patients/participants provided their written informed consent to participate in this study.
GB, HB, and ET contributed to the study conceptualization and design, analysis, and interpretation of the data. QD wrote the first draft of the article. MD and DT commented comprehensively on the manuscript. All authors have read and approved the submitted manuscript.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The reviewer TS declared a past co-authorship with one of the authors GB to the handling Editor.