Age-dependent effects of lockdown and post-lockdown periods on HbA1c during the COVID-19 pandemic: A 3-year longitudinal cohort study Authors: Vehbi Şirikçi, Cem Onur Kiraç, Hüseyin Avni Findikli Published: 2024 ABSTRACT The impact of COVID-19 lockdown on glycemic control in patients with diabetes mellitus (DM) remains unclear. This study aimed to investigate the effects of lockdown measures on the metabolic parameters of patients with DM, with particular emphasis on geriatric populations. In this retrospective, longitudinal cohort study, 1224 patients were analyzed. Three periods were identified to examine the effects of the lockdown: pre-lockdown, lockdown, and post-lockdown. Each period spanned 1-year. Within each 1-year period, at least 2 measurements were taken at least 3 months apart, and their arithmetic mean was calculated. Only patients who presented to the hospital for DM management during all 3 periods were included in the study. While HbA1c levels significantly increased in patients over 65 years old during the lockdown period (P = .017), we observed a significant decrease in HbA1c levels in patients under 65 years old (P = .014). Upon further stratification of patients over 65 by age groups, HbA1c levels increased the most among those aged 75 to 85 years, with a significant rise also observed in those aged 65 to 75 years during the lockdown. However, there was no change in HbA1c levels for patients over 85 years old during the lockdown. These findings highlight the need for careful monitoring of elderly patients with DM during lockdown periods, facilitated via home care or telehealth services. Structured diet and exercise programs should also be provided for at home adherence. 1. Introduction The emergence of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) in late 2019 in Wuhan, China, led to the global pandemic of Coronavirus Disease 2019 (COVID-19), profoundly impacting public health systems worldwide.[1] Beyond its respiratory effects, COVID-19 affects multiple organ systems, including the kidneys, heart, gastrointestinal tract, and endocrine system, which may be linked to vasculitis-related inflammation.[2–5] Many nations implemented strict lockdown measures to curb the virus’s spread. Lockdown measures restricted daily activities and access to essential healthcare services, driven by government policies and fear of COVID-19.[6,7] The lockdown severely limited people’s movements, leading to substantial changes in lifestyle, mental health, and behavior, such as modifications in diet, exercise routines, and work-related activities. These alterations might substantially influence the metabolic health of the population, particularly in those with diabetes mellitus (DM), where diet and exercise are key components of disease management.[8] However, studies investigating the impact of lockdown on patients with DM have reported inconsistent findings, varying not only across countries but also within regions. Some studies highlight a worsening of glycemic control during or after lockdown, while others report improvements or no significant change.[9–12] The aim of our study, in light of this information, is to evaluate the metabolic parameters of patients with DM in our city before, during, and after the lockdown, with a specific focus on geriatric age groups. The motivation for focusing on this demographic stems from its higher vulnerability to the challenges posed by the pandemic and the associated lockdown measures. Older adults with DM are particularly susceptible to exacerbations in health and disease management during crises like the lockdown. This study provides critical insights into the impact of lockdown measures on this vulnerable population, aiming to guide future strategies for managing similar challenges. 2. Methods This study was approved by both the Ministry of Health of the Republic of Turkey and the ethics committee of Kahramanmaras Sutcu Imam University Faculty of Medicine (decision no:05, dated:14.12.2022). All clinical investigations were conducted in accordance with the principles of the Declaration of Helsinki. In this retrospective, longitudinal cohort study, 1224 patients were analyzed (Fig. 1). Data were obtained retrospectively from the hospital’s patient database. The study included patients aged 18 years and older, diagnosed with type 1 or type 2 DM, and followed up at the hospital between January 1, 2019, and July 1, 2022. Patients were included if they had at least 2 hospital visits per period, with a minimum interval of 3 months between visits. Patients with incomplete measurements in any period were excluded from the analysis. No imputation methods were applied for missing data. Although this approach reduced the sample size, it was chosen to maintain data integrity and enhance the reliability of the results. Patient enrollment process chart. Three main periods were identified in Turkey to examine the effects of the COVID-19 pandemic, forming 3 groups: pre-lockdown, lockdown, and post-lockdown. The pre-lockdown period spans from January 1, 2019, to March 10, 2020, representing the period before the onset of the COVID-19 pandemic in Turkey. This period serves as a reference for evaluating the normal health and social conditions before the pandemic. The lockdown period extends from March 11, 2020, to March 1, 2021. This period began with the detection of the first officially confirmed COVID-19 case in Turkey on March 11, 2020.[13] The post-lockdown period covers March 2, 2021, to July 1, 2022. March 1, 2021, marks the start of the controlled normalization period, during which restrictions were gradually eased,[14] while July 1, 2021, is the date when most restrictions were lifted, transitioning to full normalization.[15] Each period spans approximately 1-year. For each period, at least 2 measurements were taken at least 3 months apart within each period, and the arithmetic mean of these measurements was calculated to reduce the influence of measurement variability across. To evaluate periodic changes in parameters, delta (Δ) values were calculated as the differences between measurements in each period. The difference between the lockdown and pre-lockdown periods was termed the “Lockdown period change,” while the difference between the post-lockdown and lockdown periods was termed the “Post-Lockdown period change.” Variables with calculated differences were denoted using the delta symbol (Δ). 2.1. Statistical analysis 2.1.1. Statistical analysis Statistical analyses were conducted using the Python and R programming languages. In Python, the Pandas, NumPy, SciPy, Statsmodels, and Matplotlib libraries were utilized, while in R, the dplyr, ggplot2, car, and robustbase packages were employed. Continuous variables were described using means, standard deviations, and medians (interquartile range), and categorical variables were expressed as frequencies (n) and percentages. The normality of data distribution was assessed using the Kolmogorov-Smirnov and Shapiro–Wilk tests. Based on these tests, appropriate statistical tests such as the Chi-square test, Fisher exact test, and the Mann–Whitney U test were applied. To evaluate repeated measurements, the Friedman test followed by the Wilcoxon signed-rank test was used. The effects of different periods on the dependent variable were analyzed using robust regression analysis with Huber’s T-norm method. Robust regression analysis was chosen to mitigate the influence of outliers, with Huber’s T-norm method applied for weighting. Parameters with high multicollinearity or those not meeting regression assumptions were excluded from the analysis to avoid potential bias in regression estimates. Multicollinearity was assessed using the Variance Inflation Factor (VIF), revealing a high correlation between blood glucose (BG) and HbA1c. Consequently, BG was excluded from the final model due to its strong association with HbA1c. Initially, univariate analyses were performed to examine the effect of each independent variable on the dependent variable. In this selection process, variables with P < .1 were considered for inclusion as covariates. However, statistical significance across all primary analyses was defined as P < .05 to ensure rigorous interpretability. Fixed factors were subsequently incorporated to assess the effects on the dependent variable while controlling for potential confounding influences. This methodological framework was designed to enhance the robustness and validity of the findings, ensuring that the results accurately reflect the biochemical impacts of different periods and groups. 3. Results 3.1. Participant demographics and baseline characteristics In this study, a total of 1224 patients were examined; 40% (n = 490) were male and 60% (n = 734) were female. The overall mean age was 62.50 ± 18.36 years, ranging from 18 to 119 years. Female participants ranged from 18 to 118 years, with a mean age of 60.94 ± 17.63 years. Male participants ranged from 20 to 119 years, with a mean age of 64.85 ± 19.17 years. In terms of age distribution, 671 individuals (54.8%) were under 65 years old, while 553 individuals (45.2%) were 65 years or older. Among female participants, 57.5% (n = 422) were under 65 years old, and 42.5% (n = 312) were 65 years or older. Among male participants, 50.8% (n = 249) were under 65 years old, and 49.2% (n = 241) were 65 years or older. The relationship between gender and age groups was statistically significant (Pearson Chi-square = 5.289, DF = 1, P = .021), indicating a higher proportion of males in the 65 years and older group. 3.2. Analysis of repeated measures by age groups To evaluate the impact of the COVID-19 pandemic on various parameters, repeated measures analysis was performed on individuals aged 65 and older as well as those under 65 years old. Comparing the lockdown period to the pre-lockdown period, significant increases in HbA1c, alanine aminotransferase (ALT), and creatinine (CR) levels (P < .05) and significant decreases in High-Density Lipoprotein (HDL) levels (P < .05) were observed in individuals aged 65 and older. Increases in BG, Low-Density Lipoprotein (LDL), and hemoglobin (HGB) levels approached statistical significance but were not significant (P > .05). Increases in triglyceride (TG) and white blood cell (WBC) levels and decreases in platelet (PLT) levels were also not statistically significant (P > .05). In the post-lockdown period, significant decreases in HbA1c levels (P < .05) and significant increases in HDL levels (P < .05) were observed. Decreases in ALT, LDL, and CR levels approached statistical significance but were not significant (P > .05). Additionally, decreases in BG, TG, HGB, and PLT levels and increases in WBC levels were not statistically significant (P > .05). In individuals under 65 years old, comparing the Lockdown period to the pre-lockdown period, significant decreases in HbA1c and TG levels (P < .05) and significant increases in HDL levels (P < .05) were observed. Decreases in BG and LDL levels and increases in PLT levels approached statistical significance but were not significant (P > .05). Decreases in ALT and CR levels and increases in WBC and HGB levels were also not statistically significant (P > .05). In the post-lockdown period, increases in HbA1c levels and decreases in HDL levels approached statistical significance but were not significant (P > .05). Decreases in ALT, WBC, and PLT levels and increases in BG, CR, and HGB levels were not statistically significant (P > .05). These findings indicate significant changes in biochemical parameters during different periods of the COVID-19 pandemic in both older and younger populations. Detailed results are presented in Table 1. Table 1Comparison of repeated measures of biochemical parameters across 3 time points in different age groups.VariablePre-lockdownLockdownPost-lockdownP-value (pre vs lockdown)P-value (lockdown vs post) Age ≥ 65 group HbA1c8, 7 (8, 1–9, 5)8, 8 (7, 9–10, 1)8, 72 (7, 7–9, 91) .017 .040 ALT24, 3 (22, 1–26, 1)25.1 (21, 7–28, 2)24, 9 (21, 7–28, 3) .023 .065BG197 (181–221)209 (184–241)201 (162–254).080.142HB13, 4 (12, 1–14, 4)13, 5 (12, 4–14, 5)13, 5 (12, 6–14, 6).063.226PLT259 (216–308)261 (209–304)255 (222–298).330.647WBC7, 4 (6, 4–8, 9)7, 7 (6, 6–9, 1)7.6 (6.5–8.9).106.376TG162 (123–227)175 (125–232)164 (120–240).268.140HDL46 (40–54)44 (38–53)46 (38–54) .027 .039 LDL129 (101–157)132 (96–164)130 (100–155).053.084CR0, 87 (0, 72–1, 02)0, 89 (0, 74–1, 09)0, 86 (0, 71–1, 05) .034 .061 Age < 65 group HbA1c8, 3 (7, 75–9, 47)8, 3 (7, 72–9, 22)8, 4 (7, 6–9, 41) .014 .085ALT19, 4 (14, 4–28, 4)19 (14–26)19 (13–27).120.161BG199 (154–257)193 (145–250)202 (149–255).083.150HB13, 6 (12, 4–14, 9)13, 6 (12, 4–14, 8)13, 7 (12, 5–14, 8).647.411PLT271 (228–320)271 (231–322)269 (222–320).076.208WBC7, 63 (6, 48–8, 86)7, 58 (6, 54–9)7, 48 (6, 4–8, 8).109.099TG163 (115–242)160 (117–237)166 (96–231) .035 .044 HDL42 (37–50)44 (37.6–51)43 (37–51) .030 .067LDL120 (95–148)118 (89–149)123 (94–152).062 .028 CR0, 73 (0, 62–0, 88)0, 72 (0, 62–0, 87)0, 73 (0, 62–0, 90).103.116Bold values indicate statistically significant results (P < .05).ALT = alanine aminotransferase (U/L), BG = blood glucose (mg/dL), CR = creatinine (mg/dL), HB = hemoglobin (g/dL), HbA1c = glycated hemoglobin (%), HDL = high-density lipoprotein (mg/dL), LDL = low-density lipoprotein (mg/dL), PLT = platelet count (103/µL), TG = triglycerides (mg/dL), WBC = white blood cells (103/µL). Comparison of repeated measures of biochemical parameters across 3 time points in different age groups. Bold values indicate statistically significant results (P < .05). ALT = alanine aminotransferase (U/L), BG = blood glucose (mg/dL), CR = creatinine (mg/dL), HB = hemoglobin (g/dL), HbA1c = glycated hemoglobin (%), HDL = high-density lipoprotein (mg/dL), LDL = low-density lipoprotein (mg/dL), PLT = platelet count (103/µL), TG = triglycerides (mg/dL), WBC = white blood cells (103/µL). 3.3. Analysis of parameter changes (Δ) by age groups during lockdown and post-lockdown periods This study evaluated parameter changes during the lockdown and post-lockdown periods in individuals aged 65 and older and those under 65 years old. ΔHbA1c levels were significantly higher in the Lockdown period for the 65 and older group (P < .05) and in the post-lockdown period for the under 65 group (P < .05). ΔALT levels were significantly higher in the Lockdown period for the 65 and older group (P < .05) but did not show a significant difference in the post-lockdown period (P > .05). ΔCR levels were significantly higher in the Lockdown period for the 65 and older group (P < .05) but did not show a significant change in the post-lockdown period (P > .05). ΔWBC, ΔHGB, and ΔPLT levels did not show significant differences between the groups during the Lockdown and post-lockdown periods (P > .05). ΔTG levels were significantly higher in the Lockdown period for the 65 and older group (P < .05) and in the post-lockdown period for the under 65 group (P < .05). ΔHDL levels were significantly higher in the Lockdown period for the under 65 group (P < .05) and in the post-lockdown period for the 65 and older group (P < .05). ΔLDL levels were significantly higher in the Lockdown period for the 65 and older group (P < .05) but did not show a significant difference in the post-lockdown period (P > .05). ΔBG levels were significantly higher in the Lockdown period for the 65 and older group (P < .05) and in the post-lockdown period for the under 65 group (P < .05). These findings underscore the importance of age-related parameter changes. Detailed results are presented in Table 2. Table 2Comparison of biochemical parameter changes (Δ) by age group during lockdown and post-lockdown periods.VariableLockdown periodPost-lockdown periodAge < 65Age ≥ 65P-valueAge < 65Age ≥ 65P-valueΔHbA1c−0.0002 (−0.52, 0.38)0.06 (−0.65, 0.85) <.001 0.03 (−0.38, 0.51)−0.05 (−0.9–0.6) .007 ΔALT0.007 (−4.41, 3.42)0.09 (−3.08, 4.69) .006 0.001 (−4.5, 3.5)−0.09 (−0.68, 0.54).779ΔCR−0.01 (−0.09, 0.08)0.02 (−0.08, 0.09) .005 0.002 (−0.09, 0.1)−0.03 (−0.9, 0.07) .025 ΔBG−10 (−66.4, 56.7)6.2 (−39.5, 49.6) .035 10 (−58.6, 66.5)−10.1 (−64.5, 50) .042 ΔWBC0.047 (−0.66, 0.80)0.011 (−0.66, 0.84).941−0.11 (−0.82, 0.71)0.035 (−0.87, 0.64).695ΔHB0.0 (−0.47, 0.45)0.05 (−0.4, 0.5).1090.025 (−0.50, 0.6)0.0 (−0.5, 0.4).147ΔPLT2.0 (−16, 19)−2.0 (−26.5, 20).054−2.5 (−21.0, 17.25)1.0 (−18.5, 22.5).200ΔTG−0.89 (−40.5, 27.7)6.41 (−28.9–32.2) .030 0.43 (−30.1, 44.9)−4.46 (−33.1, 29.5) .015 ΔHDL0.25 (−4, 5)−0.5 (−5.12–3.5) .002 −1 (−5, 5)1 (−3, 4.16) .005 ΔLDL−5.17 (−29.6–25.1)2.0 (−9, 10) .023 1.19 (−24.2, 35.1)−1.9 (−4.48, 8.41).126Bold values indicate statistically significant results (P < .05).ΔALT = alanine aminotransferase (U/L), ΔBG = blood glucose (mg/dL), ΔCR = creatinine (mg/dL), ΔHB = hemoglobin (g/dL), ΔHbA1c = glycated hemoglobin (%), ΔHDL = high-density lipoprotein (mg/dL), ΔLDL = low-density lipoprotein (mg/dL), ΔPLT = platelet count (103/µL), ΔTG = triglycerides (mg/dL), ΔWBC = white blood cells (103/µL). Comparison of biochemical parameter changes (Δ) by age group during lockdown and post-lockdown periods. Bold values indicate statistically significant results (P < .05). ΔALT = alanine aminotransferase (U/L), ΔBG = blood glucose (mg/dL), ΔCR = creatinine (mg/dL), ΔHB = hemoglobin (g/dL), ΔHbA1c = glycated hemoglobin (%), ΔHDL = high-density lipoprotein (mg/dL), ΔLDL = low-density lipoprotein (mg/dL), ΔPLT = platelet count (103/µL), ΔTG = triglycerides (mg/dL), ΔWBC = white blood cells (103/µL). 3.4. Univariate analysis of variables The individual effects of various factors on ΔHba1c levels were evaluated using univariate robust regression analysis, as summarized in Table 3. During the lockdown period, being aged 65 and over (β = 0.1898, OR = 1.209, 95% CI: 1.091–1.339, P < .001) and increased ΔALT levels (β = 0.0112, OR = 1.011, 95% CI: 1.006–1.017, P < .001) were significant determinants of ΔHba1c. These findings suggest that older individuals and those with elevated ΔALT levels experienced greater difficulties in glycemic control during the lockdown period. Additionally, a decrease in ΔHDL levels approached significance (β = −0.0059, OR = 0.994, 95% CI: 0.988–1.000, P = .063), indicating a potential association between changes in ΔHDL levels and glycemic control. Table 3Univariate assessment of factors affecting ΔHbA1c changes in lockdown and post-lockdown periods.VariableLockdown periodPost-lockdown periodβOR95% CIP-valueβOR95% CIP-valueAge ≥ 650.18981.209(1.091, 1.339)<.001*−0.12040.88650.7963, 0.9867.0240**Sex0.05141.053(0.951, 1.165).322−0.01450.98560.8751, 1.1101.7844ΔCR0.01491.015(0.853, 1.208).867−0.00970.99030.8235, 1.1905.9245ΔALT0.01121.011(1.006, 1.017)<.001*−0.01210.98800.9814, 0.9946.0005**ΔTG0.00011.000(1.000, 1.001).716−0.00210.99790.9974, 0.9985.0000**ΔHDL−0.00590.994(0.988, 1.000).063*0.00831.00841.0020, 1.0148.0109**ΔLDL0.00051.000(0.999, 1.002).428−0.00160.99840.9972, 0.9996.0067**ALT = alanine aminotransferase (U/L), CR = creatinine (mg/dL), HDL = high-density lipoprotein (mg/dL), LDL = low-density lipoprotein (mg/dL), TG = triglycerides (mg/dL). Variables with * and ** (P < .1) were included in the multivariate analysis. The reference category for “Sex” is male. Univariate assessment of factors affecting ΔHbA1c changes in lockdown and post-lockdown periods. ALT = alanine aminotransferase (U/L), CR = creatinine (mg/dL), HDL = high-density lipoprotein (mg/dL), LDL = low-density lipoprotein (mg/dL), TG = triglycerides (mg/dL). Variables with * and ** (P < .1) were included in the multivariate analysis. The reference category for “Sex” is male. In the post-lockdown period, significant determinants of ΔHba1c included ΔALT (β = −0.0121, OR = 0.988, 95% CI: 0.9814–0.9946, P = .0005), ΔTG (β = −0.0021, OR = 0.9979, 95% CI: 0.9974–0.9985, P < .0001), increased ΔHDL levels (β = 0.0083, OR = 1.0084, 95% CI: 1.002–1.0148, P = .0109), and decreased ΔLDL levels (β = −0.0016, OR = 0.9984, 95% CI: 0.9972–0.9996, P = .0067). These results indicate that biochemical changes continued to affect glycemic control after the lockdown period, and that alterations in ΔALT, ΔTG, ΔHDL, and ΔLDL levels are significant potential risk factors. Decreases in ΔALT, ΔTG, and ΔLDL levels were associated with improved glycemic control, while increases in ΔHDL levels also demonstrated a positive effect. In conclusion, changes in biochemical parameters during both the lockdown and post-lockdown periods have significant impacts on glycemic control. Particularly, changes in age, ΔALT, ΔTG, ΔHDL, and ΔLDL levels are critical for maintaining glycemic control. Careful monitoring and management of these parameters could be an essential strategy in diabetes management. 3.5. Adjusted age group impacts In the univariate analyses, all variables with a significance level of P < .1 for changes in ΔHba1c were included as covariates. Age groups were categorized as follows: “group [1]” (age < 65, reference category), “group [2]” (65 ≤ age < 75), “group [3]” (75 ≤ age < 85), and “group [4]” (age ≥ 85). These categories were included as fixed factors in the multivariate analysis. The analyses were conducted using robust regression analysis with Huber’s T-norm. The relationship between ΔHba1c and age groups was assessed, controlling for the effects of covariates. During the lockdown period, the effects of age groups on ΔHba1c were analyzed, and the results are presented in Figure 2. For group [2], the odds ratio (OR) was 1.186 (CI = 1.045–1.346), β = 0.1706, P = .008. This result indicates a significant increase in ΔHba1c for individuals aged 65 to 75. For group [3], the OR was 1.314 (CI = 1.117–1.546), β = 0.2730, P = .001, indicating that the increase in ΔHba1c was highest in individuals aged 75 to 85. For group [4], the OR was 1.151 (CI = 0.974–1.360), β = 0.1406, P = .098, but this increase was not statistically significant. The most pronounced increase in ΔHba1c was observed in group [3] (Fig. 3). In addition to these changes in ΔHba1c, significant changes in biochemical parameters such as ΔHDL and ΔALT were also observed during the lockdown period. There was a decrease in ΔHDL (OR = 0.993, CI = 0.987–1.000, β = −0.0067, P = .04) and an increase in ΔALT (OR = 1.011, CI = 1.006–1.016, β = 0.0109, P < .001). Forest plot of ΔHba1c changes by age group during lockdown, adjusted for covariates. The reference categories are as follows: “group [2],” “group [3],” and “group [4]” with “group [1]” as the reference category. “group [1]”: age < 65, “group [2]”: 65 ≤ age < 75, “group [3]”: 75 ≤ age < 85, “group [4]”: age ≥ 85. Forest plot of ΔHba1c changes by age group post-lockdown, adjusted for covariates. The reference categories are as follows: “group [2],” “group [3],” and “group [4]” with “group [1]” as the reference category. “group [1]”: age < 65, “group [2]”: 65 ≤ age < 75, “group [3]”: 75 ≤ age < 85, “group [4]”: age ≥ 85. In the post-lockdown period, the effects of age groups on ΔHba1c were examined. For group [2], the OR was 0.925 (CI = 0.810–1.058), β = −0.0774, P = .2566, indicating no significant change in ΔHba1c for this group. For group [3], the OR was 0.782 (CI = 0.660–0.927), β = −0.2455, P = .0047, indicating a significant decrease in ΔHba1c for individuals aged 75 to 85. Biochemical parameters also showed changes; there was a decrease in ΔTG (OR = 0.998, CI = 0.997–0.999, β = −0.0020, P < .0001) and ΔALT (OR = 0.988, CI = 0.981–0.995, β = −0.0122, P = .0004). No significant changes in ΔHba1c were observed for group [4] (Fig. 3). Overall, a decrease in ΔHba1c values was observed during the post-lockdown period. The most affected group was group [3], with a significant reduction in ΔHba1c. In conclusion, the impact of COVID-19 on biochemical parameters varied across different age groups during and after the pandemic. During the lockdown period, group [3] was the most affected, while group [4] was the least affected. In the post-lockdown period, group [3] showed the most improvement, with no significant changes observed in group [4]. These findings help us understand the differential biochemical impacts of the pandemic across age groups (Fig. 4). Beta coefficient plot of multivariate analysis of ΔHba1c changes by age group, adjusted for covariates including reference category. The reference categories are as follows: group [2]: 65 ≤ age < 75, group [3]: 75 ≤ age < 85, group [4]: age ≥ 85, with group [1]: age < 65 as the reference category. Note: In this analysis, the beta coefficients for the reference age group (group [1]) were calculated, and the beta coefficients for the other groups were adjusted relative to the beta value of the reference group. 4. Discussion Based on the data from the World Health Organization, COVID-19 has caused the deaths of over 7 million people worldwide since the outbreak began in late 2019.[16] Research has indicated that Angiotensin-Converting Enzyme II (ACE-II) served as the entry point for SARS-CoV-2 infection.[17] ACE-II receptors were widely present in alveolar type I and type II epithelial cells and were also detected in various organs, including the pancreas, which leads to inflammation and cytokine storm.[18,19] A meta-analysis revealed that DM was the leading cause of death among hospitalized COVID-19 patients, associated with a 1.85 times higher risk of mortality.[20] Additionally, uncontrolled hyperglycemia in patients with DM results in various microvascular complications, such as nephropathy, neuropathy, and retinopathy.[21–23] Since COVID-19 has impacted the entire world for an extended period, numerous studies investigating lockdown effects on patients with DM have been published from various countries. The results identified in the studies were not consistent with each other. This inconsistency could have emerged from several factors, including the lockdown policies in the countries, the accessibility of healthcare services for patients, the design of the studies, and the habits of the patients in these countries before the lockdown. Ojo et al published a comprehensive meta-analysis that reviewed 8 studies conducted in 7 countries, assessing the impact of lockdown on the glycemic and metabolic values of patients diagnosed with DM.[9] The studies revealed differing effects of the COVID-19 lockdown on HbA1c levels, with some reporting a significant increase in patients with DM,[24–26] others no significant change,[27–29] and some indicating an overall improvement in glycemic control.[30,31] The meta-analysis’s statistical review, however, indicated that the lockdown was responsible for an increase in HbA1c levels.[9] Our study design differed from the previously mentioned studies. We assessed patients by dividing them into age groups and evaluated the impact of both the lockdown period and the post-lockdown period on BG regulation. The reason for other studies may not have included such a time frame could be the very short duration of the lockdown in some countries. In our study, while HbA1c levels significantly increased in patients over 65 years old during the lockdown period, we observed a decrease in HbA1c levels in patients under 65 years old. This difference may be attributed to the fact that lockdown restrictions in our country were applied more strictly to individuals over 65 years old. Starting from March 2020, public places such as cinemas, theaters, cafes, sports arenas, and entertainment venues were closed to all age groups, and restrictions were imposed on domestic and international travel, while a full lockdown was implemented for individuals over 65 years old, effectively banning them from leaving their homes. When we further categorized patients over 65 years old, some notable patterns emerged. HbA1c levels increased the most among those aged 75 to 85 years during the lockdown, with a significant rise also observed in those aged 65 to 75 years. However, there was no change in HbA1c levels for patients over 85 years old during the lockdown. The reason for BG regulation in patients over 85 years old was unaffected by the lockdown might be that they already led a sedentary lifestyle prior to the lockdown. There are not many studies investigating the impact of COVID-19 lockdown on patients with DM in the geriatric age group. In the study conducted by Giandalia et al in Italy on patients with DM over 65 years old, it was noted that there was no deterioration in BG regulation during the COVID-19 lockdown.[32] The authors attributed this to the support provided to patients through telemedicine during the lockdown. In a study conducted by Falcetta et al in Italy on patients with DM over 60 years old, it was noted that the lockdown period had a negative impact on HbA1c levels, particularly in those over 80 years old.[33] As observed, even within our study, subgroup analysis revealed significant differences in HbA1c levels among various age groups. In another study conducted by Eren et al in our country, it was observed that the number of patients with DM attending follow-ups decreased during the lockdown, and HbA1c levels of these patients increased. Additionally, the average age of patients included in the study indicated that a younger group of patients attended follow-ups during the lockdown.[34] These studies show that even within the same country, patients in different regions experienced varying effects from the lockdown. After the lockdown period ended, when HbA1c levels were assessed by age group, we observed a significant decrease in HbA1c levels in patients over 65 years old that had increased during the lockdown. In contrast, there was no significant change in HbA1c levels in patients under 65 years old. The significant rise in HbA1c observed in the 75 to 85 age group during the lockdown notably decreased after the lockdown period. HbA1c levels in patients over 85 years old remained unchanged even after the lockdown period. In the study conducted by Alyahya et al in Jordan, similar to our results, an improvement in HbA1c levels of patients with DM was observed after the lockdown period.[35] In another study conducted in Taiwan, it was shown that BG regulation of the patients with DM was better after the lockdown compared to before the COVID-19 pandemic.[36] The authors attribute this improvement to the fact that the lockdown measures in Taiwan were implemented in a way that did not adversely affect the healthcare system. It is well known that the lipid profile of patients with DM has a significant impact on cardiovascular risk.[37] Therefore, in this study, we also evaluated changes in the patients’ lipid levels during the lockdown period and observing differences according to age groups. During the lockdown period, we observed that HDL levels decreased in patients over 65 years old, while HDL levels increased in those under 65 years old. We believe that the deterioration in HDL levels in patients over 65, similar to the changes in HbA1c, may be due to lifestyle changes during the lockdown period. Recent surveys indicate that around 22% of people reported weight gain during quarantine, accompanied by decreased physical activity and poorer eating habits during the COVID-19 lockdown.[38,39] In a study conducted by Falcetta et al, similar to our findings, it was observed that HDL levels of the patients decreased during the lockdown. This study also showed that patients experienced weight gain during the lockdown, which was associated with increased insulin resistance and worsening of metabolic syndrome criteria.[33] In the same study, it was observed that, similar to our findings, TG levels were not affected by the lockdown. In our study, while no changes in TG levels were observed in patients aged 65 and older during both the lockdown and post-lockdown periods, a decrease in TG levels was noted during the lockdown period in patients younger than 65 years, followed by an increase in TG levels after the lockdown. Unfortunately, due to the retrospective design of our study, data about the change in dietary habits and physical activity during lockdown were not available. In the study conducted by Huang et al in Taiwan, it was observed that there were no changes in the lipid profiles of the patients during the lockdown period.[36] This may be due to the fact that the lockdown measures implemented in Taiwan lasted only a short period of about 2 months. One of the reasons for the increased cardiovascular risk in diabetic patients is the deterioration of kidney function.[40] Therefore, in this study, we also evaluated changes in the creatinine levels of the patients. Our study also found a statistically significant increase in creatinine levels among patients aged 65 and older during the quarantine period. However, this increase did not reach the level of acute renal failure in any of the patients. Since this is a retrospective study, we were unable to assess the weight changes of the patients. It is known that elevated ALT levels are closely associated with weight gain, especially in patients with DM.[41] As with creatinine levels, a statistically significant increase in ALT levels was observed in patients over 65 years old during the lockdown period; however, this increase was not of clinical significance. This result may suggest that there was weight gain in patients over 65 years old during the lockdown period. There are several limitations to this study that should be acknowledged. First, this was a retrospective chart review conducted at a single center in Kahramanmaraş, Turkey, which may limit the generalizability of our findings to other populations and settings. The dataset’s limitations also prevented us from conducting subgroup analyses for diabetes-related chronic complications, restricting our ability to fully evaluate the impact of the pandemic period on patient subgroups with varying comorbidities. Additionally, the dataset did not allow for classification of diabetes type (type 1 or type 2 DM), a significant limitation given that previous studies suggest different impacts of lockdown on various types of DM. Furthermore, the retrospective nature of the study precluded data collection on potential confounding variables such as weight changes, modifications in diet and exercise routines, and COVID-19-related stress and anxiety, all of which could affect BG regulation. We were also unable to gather reliable information on the type of glucose-lowering treatment, the duration of DM diagnosis, or any adjustments to DM or dyslipidemia treatment during or after the lockdown. These limitations may have affected our ability to comprehensively assess the biochemical parameters in patients with DM, and we acknowledge these as constraints within the study. 5. Conclusions As a result of our study, a notable deterioration was observed in BG and lipid profiles, especially in the geriatric age group of patients with DM, during the lockdown period. In studies on telemedicine conducted during the lockdown period, it was observed that when patients with DM received both medical and psychological support through telemedicine, their BG regulation was not adversely affected.[32,42,43] These findings highlight the importance of diligent monitoring of patients with DM during lockdown periods. This can be facilitated through home care or telehealth services. Additionally, patients should be provided with structured diet and exercise programs that they can adhere to at home. While we believe the results of our study are primarily related to lifestyle changes during the lockdown, larger prospective studies are needed to more clearly identify the factors influencing the metabolic parameters of patients with DM during such periods. Author contributions Conceptualization: Vehbi Şirikçi, Hüseyin Avni Findikli. Data curation: Vehbi Şirikçi, Hüseyin Avni Findikli. Formal analysis: Hüseyin Avni Findikli. Investigation: Vehbi Şirikçi. Methodology: Cem Onur Kiraç. Project administration: Cem Onur Kiraç, Hüseyin Avni Findikli. Resources: Vehbi Şirikçi, Cem Onur Kiraç. Software: Hüseyin Avni Findikli. Writing – original draft: Cem Onur Kiraç. Writing – review & editing: Vehbi Şirikçi, Cem Onur Kiraç.