An observational cross-sectional study was conducted over 5 months (June–October 2022), targeting previously diagnosed patients with diabetes from 2 endocrinology clinics in the South of Lebanon. The study protocol, survey, and consent form were reviewed and approved by the institutional review board of the faculty of pharmacy at the Lebanese University.

Patients were recruited from 2 external endocrinology clinics located next to 2 hospitals in the South of Lebanon. No criteria were based on sex, nationality, age, and ethnicity. Only those previously diagnosed with a medical file at the endocrinologists were invited to participate based on their clinic availability for routine follow-up. The sample size was determined using the Epi Info 7 software. A recent national survey in Lebanon reported that the prevalence of diabetes among adults is 7.95%.11 Therefore, the total number of potential participants was calculated considering this prevalence. As no similar studies related to cognitive function in Lebanon, the calculation assumed that the probability of having good mental and physical health was 50%. Considering a 95% confidence interval and a 5% acceptable margin of error, at least 384 participants were required. The sample included 318 patients. G power software calculated the corresponding power (70.7%).

Two pharmacists approached the patients during their internship time (Monday–Friday from 9 am to 4 pm), each in an external private endocrinology clinic. They explained the study's aims orally and invited them to participate by completing an online survey. The first page of the survey included the study's aims (written), the estimated time to answer the questionnaire (12–15 min), statements on the anonymity and confidentiality of their responses, and consent to use their data for research purposes only. Participation in the survey was strictly voluntary. Data related to the medical history and laboratory tests were retrieved from the patients' medical files after the approval of the endocrinologist and the patient.

After an extensive literature review, a uniform questionnaire was used for data collection. It was administered in Arabic (the official language in Lebanon). The first part included questions about the general characteristics of the participants (sex, age (categorized into <45, 45–65 or >60), height and weight (used to generate the Body Mass Index (BMI)), marital status (single/divorced or married), level of education, work status, perceived economic situation, and if they had medical insurance. This part also included information regarding their lifestyle, such as smoking status (current, former, or non-smoker), alcohol consumption, physical activity, following a special diet, and if they were current car drivers. The second part asked participants about their medical history, including other comorbidities (hypertension, dyslipidemia, cardiovascular diseases, chronic kidney disease, psychological disorders, and dysthyroidism). They were also asked about the management of their diabetes: the duration of the disease (<2, 2–5 or >5 years), if they had a controlled glycemia, type of medication used (tablets, injections, or both), adherence to medications, and doing regular laboratory tests or glycemic monitoring at home. The last part comprised the assessment of the cognitive function of the patients. The Arab version of the Mini-Mental State Examination (MMSE) test was used.15 It included 11 items exploring 6 cognitive domains (orientation, recording of information, attention and calculation mind, word recall, language, and constructive praxis). Patients with a total score of less than 24 were considered at high risk of cognitive impairment. Other information related to laboratory tests, such as glycated hemoglobin A1c (HBA1c), total cholesterol, low-density lipoproteins (LDL), high-density lipoprotein (HDL), triglycerides, uric acid, hematocrit, thyroid-stimulating hormone (TSH), and vitamin B12 was retrieved from the patient's files. During the visit, the systolic and diastolic blood pressures were noted based on an examination performed by the endocrinologist. Values were then dichotomized based on the normal values adopted in the hospital.

Statistical analyses were performed using Statistical Package for Social Sciences (SPSS Inc., Chicago, Illinois) Version 29. Categorical variables are presented through frequencies and percentages. These variables included the general characteristics of the patients and the questions related to their medical history and laboratory test results. The age, Body Mass Index (BMI), the mean of the laboratory results, and the cognitive function score are presented through means and standard deviations. Individual MMSE items were recorded, summed, and transformed as recommended.16 One-way analysis of variance (ANOVA) was performed to assess the relationship between the cognitive function score and the general characteristics, medical history, and laboratory test variables. A multivariate analysis was performed using linear regression models to evaluate the combined effect of predictors on the cognitive function score. The confusion variables (or covariates) were considered if they had a statistically significant P-value (<.05) in the ANOVA analysis.

The institutional review board of the Lebanese University faculty of pharmacy reviewed and approved the study protocol, tool, and consent form. Written informed consent was obtained from each patient. They were acknowledged that their participation was voluntary and that they could withdraw it at any point of the study with only provided answers registered. This study received no funding and no financial incentives were provided. Results were considered for research purposes only, with no potential conflict of interest to declare.

Overall, 339 potential participants were approached, and 318 agreed to participate in the study (93.8%). The general characteristics of the study participants are presented in Table 1. The sample was approximately equally distributed between males (49.7%) and females (50.3%). The mean age of the participants was 59.8 (10.7) years, with 57.5% between 45 and 65 years and 32.4% above 65 years of age. Over half of the patients (51.6%) were obese, 33.3% were overweight, and only 15.1% had a normal BMI. Most participants (77.4) were married; the rest (22.6%) were single, widowed, or divorced. Among others, 15.7% were illiterate, 28.6% finished primary school, and 13.2% had a university degree. Almost 59% did not work or were retired, and 41.5% were employed. Around 45% reported an average economic situation, while 36% had a below average situation. Around 55.0% of the participants reported having medical insurance, particularly public one. Forty-eight percent were current smokers, 12.9% were former smokers, and the remaining (39.3%) were non-smokers. Furthermore, 37.7% said they practiced sports regularly, 18.6% followed a special diet suggested by their doctor, and 56.6% reported being current car drivers.

Table 2 shows the patients' medical history and latest laboratory test results. Dyslipidemia (54.2%) and hypertension (51.6%) were the most reported comorbidities among the participants, followed by cardiovascular diseases (22.3%) and psychological disorders (12.3%). Around half of the patients had diabetes for over 5 years (52.5%), and only 35.1% had their glycemia controlled. Regarding medication prescribed, most patients (86.5%) used oral tablets, and only 1.6% used injectable insulin. Treatment adherence was reported among the majority (82.4%), with 68.2% reporting regular laboratory analyses and 78.3% doing a regular glycemic check at home. The mean HBA1c value was 7.9 (1.5) %, with 77.7% of the patients having values ≥7%. Almost 70.0% of the sample had total cholesterol values <200 mg/dL, 52.2% had an LDL value above 100 mg/dL, and 80.8% had an HDL value below 60 mg/dL. Based on the doctor's examination during the day of the visit, 63.2% of the patients had a systolic blood pressure <139 mmHg and 75.2% had a diastolic blood pressure below 80 mmHg. As regards the other laboratory tests collected, 74.8% had blood uric acid values below 6 mg/dL, 47.2% had hematocrit values below 40%, 91.9% had their TSH values below 5 μU/L, and 81.7% had vitamin B12 values between 160 and 950 pg/mL.

Based on the MMSE test results, the patients' mean score was 21.6 (5.3), distributed as follows: 217 patients (68.2%) had possible cognitive impairment, and 101 (31.8%) had a normal cognitive function. Table 3 shows the association between the characteristics of the participants and the cognitive function scores. Significantly lower scores were noted among females (20.0) than males (23.2; P < .001). The increase in age was found to be significantly affecting the cognitive function scores of patients, with higher scores among those below 45 (24.1) and lower ones among those between 45 and 65 years (22.4) and above 65 (19.4; P < .001). The level of education also directly impacted the cognitive function scores, where illiterate had significantly lower scores (15.0) compared to those with advanced education (secondary school (24.5) and university graduates (24.9; P < .001)). A higher cognitive function score increased per increase in perceived economic situation with the lowest score for those with a below average situation (19.3 vs. 22.5 and 24.3; P < .001). Furthermore, those employed (23.4 vs. 20.3; P < .001) and those with medical insurance (22.4 vs. 20.7; P = .004) had significantly lower scores than others. Participants with diabetes for more than 5 years (20.8) and those with uncontrolled glycemia (20.8) had lower cognitive function than patients diagnosed more recently (22.2; P = .021) or those with controlled glycemia (23.1; P < .001). Moreover, those reporting adherence to treatment (21.9; P = .010) and those with an HBA1c value below 7% (23.2; P = .003) had higher scores than others. As regards the association between the participants' cognitive function and other laboratory tests, a cognitive decline was found among those with total cholesterol ≥200 mg/dL (20.7 vs. 21.9; P = .044), a systolic blood pressure of 140 mmHg or more (20.0 vs. 22.5; P < .001), and hematocrit levels below 40% (20.6 vs. 22.4; P = .002).

After adjusting for covariates, a significantly lower cognitive function score was found per increase of 1 year of age (b=−0.120, CI (−0.13, −0.03); P = .038). Females had a score of 0.186 less than males (CI (−3.30, −0.35); P = .016), and an increase of a factor of 0.342 was found between the levels of education (CI (0.71–2.04); P < .001). Higher economic situation was associated with a significantly better cognitive function (b = 0.112, CI (0.06, 1.83); P = .046).

Patients with uncontrolled glycemia had a lower score of 0.120 than those with controlled glycemia (CI (−2.78, −0.29); P = .023), and those with HBA1c values ≥7% had significantly lower cognitive function scores (b=−0.166 (CI: −3.57, −0.26); P = .023). Moreover, having a systolic blood pressure ≥140 mmHg decreased the cognitive function score by a factor of 0.147 (CI (−2.98, −0.01); P = .049) (Table 4). A significant regression analysis was found (F (1,82)=8.511, P < .001) with an R2 of 0.386. The corresponding equation is presented below: