Effect of Glucose Variability on the Mortality of the Very Old People During the First Year of the COVID-19 Pandemic

doi:10.21203/rs.3.rs-3413808/v1

Download PDF

Research Article

Effect of Glucose Variability on the Mortality of the Very Old People During the First Year of the COVID-19 Pandemic

https://doi.org/10.21203/rs.3.rs-3413808/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background

To our knowledge, only one study has examined the association between glucose variability (GV) and mortality in the elderly population with diabetes. GV was assessed by HbA1c, and a J-shaped curve was observed in the relationship between HbA1c thresholds and mortality. No study of GV was conducted during the COVID-19 pandemic and its lockdown. This study aims to evaluate whether GV is an independent predictor of all-cause mortality in patients aged 75 years or older with and without COVID-19 who were followed during the first year of the COVID-19 pandemic and its lockdown measures.

Methods

This was a retrospective cohort study of 407,492 patients from the AGED-MADRID dataset aged 75 years and older; 63.2% were women, and 29.3% had diabetes. GV was measured by the coefficient of variation of fasting plasma glucose (CV-FPG) over 6 years of follow-up (2015–2020). The outcome measure was all-cause mortality in 2020. Four models of logistic regression were performed, from simple (age, sex) to fully adjusted, to assess the effect of CV-FPG on all-cause mortality.

Results

During follow-up, 34,925 patients died (14,999 women and 19,926 men), with an all-cause mortality rate of 822.3 per 10,000 person-years (95% confidence interval (CI), 813.7 to 822.3) (739 per 10,000; 95% CI 728.7 to 739.0 in women and 967.1 per 10,000; 95% CI 951.7 to 967.2 in men). The highest quartile of CV-FPG was significantly more common in the deceased group (40.1% vs. 23.6%; p < 0.001). The fully adjusted model, including basal FPG, showed an odds ratio for mortality that ranged from 2.48 to 2.88 according to different sensitivity analyses.

Conclusions

GV has clear implications for clinical practice, as its assessment as a risk prediction tool should be included in the routine follow-up of the elderly and in a comprehensive geriatric assessment. Electronic health records can incorporate tools that allow its calculation, and with this information, clinicians will have a broader view of the medium- and long-term prognosis of their patients.

Blood glucose

Age

Mortality

Follow-up studies

COVID-19

Several studies have analysed mortality in elderly individuals, with cardiovascular and oncological as the most frequent causes of death in the developed world [1]. The crude mortality rate for the elderly in Spain was 3,824 deaths per 100,000 inhabitants aged 75–84 (2017) [2]. Factors such as socioeconomic level [3], frailty [4], physical activity [5) (6], self-rated health status [7], social and family support [8], chronic diseases [9, 10], multimorbidity [11], nutritional status [12], body mass index [11], and cognitive function [13] may influence mortality rates. However, it is unknown whether any homeostatic factor that influences the degree of control of chronic diseases may be associated with an increased all-cause mortality in the elderly population.

Day-to-day glucose variability (GV) is considered a homeostatic phenomenon defined as the oscillation of blood glucose levels outside the normal range, which is a predictor of microvascular and macrovascular diseases and all-cause mortality, particularly in patients with type 2 diabetes mellitus (T2DM) [14, 15]. Our initial findings confirmed a clear association between GV, as measured by the coefficient of variation of fasting plasma glucose (FPG), and all-cause mortality in patients with T2DM and additionally in individuals with prediabetes or normoglycemia [16].

In 2020, the SARS-CoV-2 lockdown and social distancing motivated changes in the general population's daily routines, specifically in elderly people, favouring an unbalanced diet, less physical activity, unavailability of some medications [17], and significantly less contact with health care professionals than usual [18]. We hypothesize that this situation could worsen the control of chronic diseases in elderly patients with a history of GV and favour an increase in all-cause mortality.

This study aims to evaluate whether GV, mediated by oxidative stress and other factors associated with senescence, is an independent predictor of all-cause mortality in patients aged 75 years or older with and without COVID-19, followed during the first year of the COVID-19 pandemic and its lockdown measures in Spain.

A retrospective cohort study was carried out in the Aged-Madrid Study, a new data analytics platform in Madrid (Spain) created to address urgent COVID-19-related questions. We used routinely collected electronic health records (EHRs) from primary care practices using AP-Madrid software, covering 424 practices (3,881 general practitioners) and 100% of the population in Madrid, linked to the Office of National Statistics death registrations (INDEF). We included all adults (aged 75 years or over) alive and under follow-up on 1 January 2020 and with at least five years of continuous EHRs in primary care before this date. We ensured that baseline data could be adequately captured (n = 587,603). We excluded individuals with missing values for age, sex or less than three FPG values between 2015 and 2020. We directly compared both all-cause mortality and COVID-19 mortality with survivors to identify the covariates to include in the models. The variable of most interest was GV, which was used to test the main hypothesis.

The CV-FPG was obtained in those patients with at least three values of FPG along a follow-up of six years and calculated as the ratio of the standard deviation to the mean FPG multiplied by 100. Patients were categorized according to the quartiles of CV-FPG. The values of these quartiles were Q1: ≤5.2635; Q2: 5.2636 to 7.8577; Q3: 7.8578 to 12.3739; and Q4: ≥12.3740.

We considered SARS-CoV-2 death when it was registered as such in the clinical chart of a hospitalized patient or when the death recorded in the INDEF occurred 15 days from a first confirmed diagnosis of SARS-CoV-2 infection. The INDEF collects all deaths occurring in Spain, but it does not record their cause.

Covariates considered in the analysis included age, sex, cardiovascular risk factors, morbidities, and medication prescriptions until December 31, 2019, and were obtained from EHRs. We recorded morbidities according to the International Classification of Primary Care (ICPC-2). We specifically registered the presence of any previous cardiovascular disease (either myocardial infarction, angina, stroke, or peripheral artery disease), any cancer active during the previous five years (except nonmelanoma skin cancer), chronic kidney disease (CKD), congestive heart failure, chronic obstructive pulmonary disease (COPD), atrial fibrillation, dementia, hypertension, and diabetes. We also gathered information about tobacco consumption from EHRs.

All blood analyses and anthropometric measurements performed between January 1, 2015 and December 31, 2020 were available for the study. However, FPG values and other biochemical parameters measured during hospitalization were not used to avoid artifacts in the results, as their values may depend on the reason for hospitalization or the cause of death at the time of hospitalization. A total of 180,111 patients were excluded for having < 3 FPG measurements during follow-up, and this analysis was performed on 407,492 patients (Fig. 1).

Medications prescribed for chronic medical conditions were obtained from the Electronic Pharmacy Database (Módulo Unico de Prescripción, MUP) of Madrid, integrated in EHRs.

There were no losses to mortality follow-up because regardless of whether the patient moved to a new city, the mortality registry is at the national level and is based on the patient's identification data, including the national identity card number, which is unique for each Spanish citizen.

We have validated the quality of the EHRs in primary care for research use [19) (20) (21], and the database has been widely employed to study the epidemiology of cardiovascular risk factors in older patients [22].

Statistical analysis

We used unpaired Student's t-test or one-way analysis of variance for continuous variables and chi-squared test for categorical variables for comparisons between/among subgroups. Cohen's d was used to measure the effect size for comparing the means between living and dead individuals. Cohen's d values of 0.2 or less reflect a small effect size, approximately 0.50 reflects a moderate effect size, and 0.80 and greater reflect a large effect size. Cramer's V was used to determine the effect size (small = 0.1, medium = 0.3, large = 0.5) of categorical variables between the two groups.

Mortality rates were calculated per 10,000 person-years with their 95% confidence interval (CI). These mortality rates were stratified by quartiles of CV-FPG and glycemic status.

Univariate and multivariate survival analyses were performed by logistic regression. In the first analysis for all-cause mortality, the odds ratios (ORs) and 95% CIs were calculated based on model 1: adjusted for age and sex; model 2: adjusted for age, sex, history of cardiovascular disease, heart failure, COVID-19 infection, and cancer; model 3: adjusted for variables in model 2 plus glycemic status, hypertension, atrial fibrillation, COPD, CKD, use of tobacco, dyslipidemia, use of antiplatelet drugs, statins, and dual blockade of the renin–angiotensin–aldosterone system (RAAS); and model 4: model 3 plus mean FPG.

Finally, two sensitivity analyses were performed. The first excluded participants with cancer in the previous two years to avoid its possible influence on mortality. The second we excluded patients who died of COVID-19.

Analyses were performed with SPSS version 21.0 (IBM Corp., Chicago, IL) and Epidat 4.1 software; a 2-sided p value < 0.05 was considered statistically significant.

Table 1 shows the baseline sociodemographic, anthropometric, and clinical factors of 407,492 participants. There were significant differences between women and men in age, Barthel Index, smoking habits, cardiovascular risk factors (dyslipidemia, hypertension, diabetes mellitus), cardiovascular diseases (myocardial infarction, stroke, peripheral artery disease), and heart failure.

Table 1

Baseline characteristics of the population aged 75 and over (1st January 2020)
	All (N = 407,492)	Women (N = 257,401)	Men (N = 150,091)	p value
Age (years), mean (SD)	83.5 (5.7)	83.9 (5.8)	82.7 (5.3)	< 0.001
Age group
75–84 y, n (%)	244.946 (60.1)	146,377 (56.9)	98,569 (65.7)	< 0.001
85–94 y, n (%)	146.228 (35.9)	98,343 (38.2)	47,885 (31.9)
> 94 y, n (%)	16,318 (4)	12,681 (4.9)	3,637 (2.8)
BMI, mean (SD)	28.6 (4.7)	28.8 (5.1)	28.4 (4)	< 0.001
Barthel index, mean (SD)	78.8 (21.1)	78.1 (21.3)	80.4 (20.7)	< 0.001
Barthel group
Barthel: Total dependence, n/N (%)	6,559/159,720 (4.1)	4,623/111,950 (4.1)	1,936/47,770 (4.1)	< 0.001
Barthel: severe dependence, n/N (%)	6,343/159,720 (4)	4,721/111,950 (4.2)	1,622/47,770 (3.4)
Barthel: moderate dependence, n/N (%)	14,054/159,720 (8.8)	10,589/111,950 (9.5)	3,465/47,770 (7.3)
Barthel: mild dependence, n/N (%)	132,760/159,720 (83.1)	92,015/111,950 (82.2)	40,745/47,770 (85.3)
Current Smoking, n (%)	20,867 (5.1)	7,137 (2.8)	13,730 (9.1)	< 0.001
Baseline SBP, mean (SD)	132.7 (16.2)	133 (16.3)	132.3 (16.1)	< 0.001
Baseline DBP, mean (SD)	73.8 (9.5)	74 (9.4)	73.5 (9.5)	< 0.001
Dementia (Alzheimer), n (%)	34.100 (8.4)	24.501 (9.5)	9,599 (6.4)	< 0.001
Dyslipidemia, n (%)	244,365 (60)	162,711 (63.2)	81,654 (54.4)	< 0.001
Hypertension, n (%)	301,608 (74)	196,780 (76.4)	104,828 (69.8)	< 0.001
Type 1 Diabetes mellitus, n (%)	1,676 (0.4)	999 (0.4)	677 (0.5)	0.002
Type 2 Diabetes mellitus, n (%)	117.880 (28.9)	66,824 (26)	51,166 (34)	< 0.001
Type 2 DM + Hypertension, n (%)	96,070 (23.6)	57,598 (22.4)	38,472 (25.6)	< 0.001
CVD, n (%)	60,573 (14.9)	26,637 (10.4)	33,936 (22.6)	< 0.001
1 vascular bed	54,952 (13.5)	25,157 (9.8)	29,795 (19.9)	< 0.001
2 vascular beds	5,353 (1.3)	1,433 (0.6)	3,920 (2.6)
3 vascular beds	268 (0.1)	47 (0.0)	221 (0.1)
Heart failure, n (%)	29,110 (7.1)	19,096 (7.4)	10,014 (6.7)	< 0.001
CKD, n (%)	124,529 (30.6)	76,887 (29.9)	47,642 (31.7)	< 0.001
Number of glucose measurements, mean (SD)	5.89 (2.7)	5.94 (2.7)	5.81 (2.7)	< 0.001
FPG, mean (SD) [when at least 3 FPG measurements]	103.3 (24.3)	101.4 (23.8)	106.4 (24.7)	< 0.001
T2DM patients	129 (26.9)	128.5 (27.7)	129.6 (25.9)	< 0.001
T1DM patients	137.4 (35.7)	138.5 (36.3)	135.8 (34.8)	0.114
Non-DM patients	92.5 (11.7)	91.7 (10.5)	94.2 (11.9)	< 0.001
CV-FPG, mean (SD) [when at least 3 FPG measurements]	10.5 (8.7)	10.2 (8.6)	10.8 (9)	< 0.001
T2DM patients	17.20 (11.53)	17.46 (11.69)	16.86 (11.31)	< 0.001
T1DM patients	27.55 (14.73)	28.47 (15.02)	26.20 (14.20)	0.002
Non-DM patients	7.61 (4.9)	7.60 (4.86)	7.61 (4.97)	0.696
FPG < 60 mg/dl at least 20% of measurements (%)*	2,790 (0.5)	1,592 (0.6)	794 (0.5)	< 0.001
T2DM patients, n/N (%)	1,316/117,880 (1.1)	826/66,824 (1.2)	490/51,056 (1)	< 0.001
T1DM patients, n/N (%)	78/1,676 (4.7)	51/999 (5.1)	27/677 (4)	0.345
Non-DM patients, n/N (%)	992/287,936 (0.3)	715/189,578 (0.4)	277/98,358 (0.3)	< 0.001
HbA1c (%), mean (SD) [when at least 3 HbA1c measurements]**	6.38 (0.9)	6.35 (0.9)	6.43 (0.9)	< 0.001
T2DM patients	6.88 (0.9)	6.89 (0.9)	6.86 (0.9)	< 0.001
T1DM patients	7.54 (1)	7.64 (1)	7.42 (1)	0.002
Non-DM patients	5.77 (0.4)	5.77 (0.4)	5.78 (0.4)	0.204
CVD, cardiovascular disease (myocardial infarction, stroke, or peripheral artery disease); CKD, chronic kidney disease (CKD-EPI < 60 ml/min/1.73 m² and/or albumin/creatinine ratio ≥ 30 mg/g (≥ 3 mg/mmol)); FPG, fasting plasma glucose; CV-FPG, coefficient of variation of fasting plasma glucose
* If the relative frequency of each patient with FPG below 60 mg/dl was 20% or more during 2015–2020.
** Mean HbA1c was calculated as the mean of all HbA1c measurements if at least three measurements were taken between 2015 and 2020.

During follow-up, 34,925 patients died (14,999 women and 19,926 men), with an all-cause mortality rate of 822.3 per 10,000 person-years (95% CI, 813.7 to 822.3) (739 per 10,000; 95% CI 728.7 to 739.0 in women and 967.1 per 10,000; 95% CI 951.7 to 967.2 in men). Mortality rates by glycemic status are shown in Fig. 2.

Compared with survivors, people who died were more likely to be male, older, ex-smokers, hypertensive, and had a more frequent history of Alzheimer's disease, cancer, diabetes, chronic obstructive pulmonary disease (COPD), COVID-19 disease, cardiovascular disease (CVD), heart failure, atrial fibrillation, chronic kidney disease (CKD), and use of aspirin, anticoagulants, beta-blockers, and insulin. They also had lower mean body mass index (BMI), lower mean systolic and diastolic blood pressure, and higher mean FPG (Table 2). The highest quartile of CV-FPG was significantly more common in the deceased group (40.1% vs. 23.6%; p < 0.001).

Table 2

Baseline sociodemographic and anthropometric measurements and clinical factors of all study participants and by survival status.
Variables	Survivors (N = 372,567)	Deceased (N = 34,925)	p value	Effect size
Age, mean (SD)	83.1 (5.4)	87.9 (6.1)	< 0.001	0.833*
Sex male, n (%)	135,092 (36.3)	14,999 (42.9)	< 0.001	0.039†
Barthel index, mean (SD)	80.5 (19.9)	68.2 (25.2)	< 0.001	0.542*
Grouped Barthel Index
Independent (100)	4/137,403 (0)	0/22,317 (0)	< 0.001	0.211†
Slight dependence (61–99)	118,003/137,403 (85.9)	14,757/22,317 (66.1)
Moderate dependence (41–60)	10,398/137,403 (7.6)	3,656/22,317 (16.4)
Severe dependence (21–40)	4,257/137,403 (3.1)	2,086/22,317 (9.3)
Complete dependence (< 20)	4,741/137,403 (3.5)	1,818/22,317 (8.1)
Alcohol consumption, n (%)	4,233 (1.1)	522 (1.5)	< 0.001	0.009†
Current smoking, n (%)	19,049 (5.1)	1,818 (5.2)	< 0.001	0.001†
BMI, mean (SD)	28.7 (4.6)	28.1 (5)	< 0.001	0.125*
Baseline SBP, mean (SD)	132.9 (16.1)	131.3 (17.2)	< 0.001	0.096*
Baseline DBP, mean (SD)	74 (9.4)	72 (9.8)	< 0.001	0.208*
T2DM patients, n (%)	105,992 (28.4)	11,888 (34)	< 0.001	0.037†
T1DM patients, n (%)	1,403 (0.4)	273 (0.8)
Normoglycemia patients, n (%)	265,172 (71.2)	22,764 (65.2)
History of CVD, n (%)	52,280 (14)	8,293 (23.7)	< 0.001	0.076†
Heart failure, n (%)	23,049 (6.2)	6,061 (17.4)	< 0.001	0.121†
Atrial Fibrillation, n (%)	57,412 (15.4)	9,305 (26.6)	< 0.001	0.085†
Hypertension, n (%)	275,381 (73.9)	26,227 (75.1)	< 0.001	0.008†
Dyslipidemia, n (%)	226,255 (60.7)	18,110 (51.9)	< 0.001	0.051†
Solid cancer, n (%)	16,785 (4.5)	3,450 (9.9)	< 0.001	0.069†
Myeloma, n (%)	674 (0.2)	186 (0.5)	< 0.001	0.022†
Leukemia, n (%)	1,246 (0.3)	294 (0.8)	< 0.001	0.023†
Lymphoma, n (%)	1,937 (0.5)	309 (0.9)	< 0.001	0.014†
Alzheimer, n (%)	26,245 (7)	7,855 (22.5)	< 0.001	0.156†
COVID-19, n (%)	10,828 (2.9)	3,863 (11.1)	< 0.001	0.122†
COPD, n (%)	39,629 (10.6)	4,898 (14)	< 0.001	0.030†
CKD, n (%)	109,793 (29.5)	14,763 (42.2)	< 0.001	0.078†
Statin use, n (%)	208,308 (55.9)	13,592 (38.9)	< 0.001	0.096†
ACEI or ARB use, n (%)	234,428 (62.9)	17,102 (49)	< 0.001	0.080†
Antiplatelet drug use, n (%)	97,336 (26.1)	10,075 (28.8)	< 0.001	0.017†
Anticoagulant use, n (%)	66,669 (17.9)	9,643 (27.6)	< 0.001	0.070†
Beta-blocker use, n (%)	80,557 (21.6)	8,460 (24.2)	< 0.001	0.018†
Calcium antagonist use, n (%)	91,759 (24.6)	7,028 (20.1)	< 0.001	0.029†
Metformin use, n (%)	76,489 (20.5)	5,877 (16.8)	< 0.001	0.026†
Insulin use, n (%)	22,361 (6)	3,360 (9.6)	< 0.001	0.042†
Baseline FPG level, mean (SD)	104.4 (29.2)	105.7 (35.3)	< 0.001	0.041*
CV-FPG, mean (SD)	10.1 (8.4)	14 (11.5)	< 0.001	0.387*
CVD, cardiovascular disease (myocardial infarction, stroke, or peripheral artery disease); CKD, chronic kidney disease (CKD-EPI < 60 ml/min/1.73 m² and/or albumin/creatinine ratio ≥ 30 mg/g (≥ 3 mg/mmol)); FPG, fasting plasma glucose; CV-FPG, coefficient of variation of fasting plasma glucose; BMI: body mass index; SBP: systolic blood pressure; DBP: diastolic blood pressure; T2DM: type 2 diabetes mellitus; T1DM: type 1 diabetes mellitus; ACEI: angiotensin-converting enzyme inhibitor; ARB: angiotensin II receptor blocker; * Cohen´s d; † Cramer´s V

The distribution of key clinical characteristics across quartiles of CV-FPG showed a higher burden of disease and vascular disease in the more extreme quartiles (3 and 4). Obviously, participants with T1DM and T2DM were more likely to be in quartile 4, and this circumstance could explain, at least in part, the high burden of vascular disease (Table 3). This fact justifies an adjustment for basal FPG in the multivariate models, as will be seen later.

Table 3

Baseline factors of 407,492 subjects grouped by quartiles of the coefficient of variation of FPG levels.
Variables	Quartiles of CV-FPG
	1 (lowest)	2	3	4 (highest)	p value
Quartiles range	≤ 5,2635	5.2636–7.8577	7.8578–12.3739	≥ 12.3740
N	101,878	101,868	101,868	101,878
Anthropometric and clinical variables
Male sex, n (%)	35,803 (35.1)	36,077 (35.4)	37,660 (37)	40,548 (39.8)	< 0.001*
Age, mean (SD)	82.9 (5.5)	83.2 (5.6)	83.7 (5.7)	84.1 (5.8)	< 0.001
Current smoking, n (%)	4,400 (4.3)	4,876 (4.8)	5,456 (5.4)	6,135 (6.0)	< 0.001*
Alcohol consumption, n (%)	929 (0.9)	1,065 (1)	1,252 (1.2)	1,509 (1.5)	< 0.001
BMI, mean (SD)	28.2 (4.4)	28.4 (4.5)	28.7 (4.7)	29.1 (4.9)	< 0.001
Baseline SBP, mean (SD)	132.4 (16.0)	132.5 (16.0)	132.7 (16.1)	133.3 (16.6)	< 0.001
Baseline DBP, mean (SD)	74.2 (9.3)	74.1 (9.4)	73.9 (9.4)	73.1 (9.6)	< 0.001
Hypertension, n (%)	70,539 (69.2)	73,477 (72.1)	76,989 (75.6)	80,596 (79.1)	< 0.001*
Dyslipidemia, n (%)	60,345 (59.2)	61,692 (60.6)	61,339 (60.2)	60,983 (59.9)	< 0.001
Glycemic status
Normoglycemia, n (%)	95,344 (93.6)	88,404 (86.8)	71,956 (70.6)	32,227 (31.6)	< 0.001*
Type 2 DM, n (%)	6,507 (6.4)	13,414 (13.2)	29,768 (29.2)	68,188 (66.9)
Type 1 DM, n (%)	27 (0.0)	49 (0.0)	144 (0.1)	1,456 (1.4)
Cardiovascular status
Previous CVD, n (%)	11,601 (11.4)	13,003 (12.8)	15,216 (14.9)	20,752 (20.4)	< 0.001*
Previous Heart failure, n (%)	4,349 (4.3)	5,523 (5.4)	7,543 (7.4)	11,695 (11.5)	< 0.001*
Atrial Fibrillation, n (%)	13,032 (12.8)	14,875 (14.6)	17,652 (17.3)	21,157 (20.8)	< 0.001*
CKD, n (%)	24,065 (23.6)	27,827 (26.8)	31,022 (30.5)	42,153 (41.4)	< 0.001*
Other clinical conditions
COVID-19 during 2020, n (%)	2,926 (2.9)	3,176 (3.1)	3,853 (3.8)	4,736 (4.6)	< 0.001*
COPD, n (%)	9,407 (9.2)	10,508 (10.3)	11,835 (11.6)	12,777 (12.5)	< 0.001*
Solid Cancer, n (%)	4,369 (4.3)	4,615 (4.5)	5,227 (5.2)	5,984 (5.9)	< 0.001*
Medication profile
Statin, n (%)	50,988 (50.0)	53,737 (52.8)	55,884 (54.9)	61,285 (60.2)	< 0.001*
ACEI or ARB, n (%)	59,143 (58.1)	61,404 (60.3)	63,555 (62.4)	67,422 (66.2)	< 0.001*
Antiplatelet drug, n (%)	21,920 (21.5)	24,049 (23.6)	26,801 (26.3)	34,639 (34.0)	< 0.001*
Anticoagulant, n (%)	15,304 (15.0)	17,120 (16.8)	20,038 (19.7)	23,847 (23.4)	< 0.001*
Diuretics, n (%)	41,458 (40.7)	45,051 (44.2)	48,956 (48.1)	53,874 (52.9)	< 0.001*
Loop diuretics, n (%)	12,118 (11.9)	14,786 (14.5)	18,689 (18.3)	26,022 (25.5)	< 0.001*
Beta-blocker use, n (%)	18,474 (18.1)	20,467 (20.1)	23,084 (22.7)	26,989 (26.5)	< 0.001*
Calcium antagonist, n (%)	20,720 (20.3)	23,024 (22.6)	25,218 (24.8)	29,822 (29.3)	< 0.001*
Metformin, n (%)	4,646 (4.6)	9,792 (9.6)	22,014 (21.6)	45,911 (45.1)	< 0.001*
DPP4-Inhibitor	1,219 (1.2)	3,064 (3.0)	8,986 (8.8)	31,792 (31.2)	< 0.001*
SGLT2-Is	171 (0.2)	441 (0.4)	1,332 (1.3)	5,371 (5.3)	< 0.001*
Insulin use, n (%)	228 (0.0)	500 (0.5)	1,762 (1.7)	23,230 (22.8)	< 0.001*
*p value for linear trend across quartiles of CV-FPG
FPG-CV: coefficient of variation of fasting plasma glucose; BMI: body mass index; SBP: systolic blood pressure; DBP: diastolic blood pressure; Type 2 DM: type 2 diabetes mellitus; Type 1 DM: type 1 diabetes mellitus; CVD: cardiovascular disease (myocardial infarction, stroke, and peripheral artery disease); CKD: chronic kidney disease; COVID-19: coronavirus disease 2019; COPD: chronic obstructive pulmonary disease; ACEI: angiotensin-converting enzyme inhibitor; ARB: angiotensin II receptor blocker; DPP4-inhibitor: dipeptidyl peptidase-4 inhibitor; SGLT2-Is: sodium-glucose cotransporter-2 inhibitor.

The adjusted effect of FPF variability on all-cause mortality was examined with 4 models, as shown in Table 4. Compared with the first quartile, the second, third and fourth quartiles showed a statistically significant increase in all-cause mortality from model 1 (adjusted for age and sex) to model 4 (full model). The highest mortality risk was for the fourth quartile in model 1 (OR, 2.54; 95% CI, 2.45 to 2.62), followed by model 4 (OR, 2.48; 95% CI, 2.38–2.57).

Table 4

The odds ratios (ORs) of all-cause mortality grouped by quartiles of the coefficient of variation of FPG among 407,492 subjects with at least three FPG measurements.
	Quartiles of CV of FPG
	1 (lowest)	2	3	4 (highest)
N	101,878	101,868	101,868	101,878
Deaths for All causes, n (%)	5,348 (5.2)	6,514 (6.4)	9,052 (8.9)	14,011 (13,8)
Model 1	1	1.19 (1.14–1.24) *	1.60 (1.54–1.66) *	2.54 (2.45–2.62) *
Model 2	1	1.17 (1.12–1.21) *	1.52 (1.46–1.58) *	2.29 (2.21–2.37) *
Model 3	1	1.17 (1.13–1.22) *	1.53 (1.47–1.58) *	2.32 (2.23–2.41) *
Model 4	1	1.18 (1.13–1.22) *	1.55 (1.49–1.61) *	2.48 (2.38–2.57) *
N: number of persons included in the analysis for each group; FPG: fasting plasma glucose; CV: coefficient of variation
Model 1: adjusted for age and sex. Model 2: adjusted for age, sex, history of cardiovascular disease, heart failure, COVID-19 infection, and solid cancer. Model 3: adjusted for variables in model 2 plus glycemic status, hypertension, atrial fibrillation, COPD, CKD, use of tobacco, dyslipidemia, use of antiplatelet drugs, use of statins, and SRA. Model 4: Model 3 plus mean FPG. *p < 0.001

A first sensitivity analysis of the logistic regression excluding persons with a history of cancer is shown in Table 5. As seen, the ORs overlap with those of the main analysis, with a similar magnitude of association in each quartile and model. The second sensitivity analysis (Table 6) showed that the results were consistent, as minimal variations in the OR were observed.

Table 5

The odds ratios (ORs) of all-cause mortality grouped by quartiles of the coefficient of variation of FPG excluding patients with a history of cancer.
	Quartiles of CV of FPG
	1 (lowest)	2	3	4 (highest)
N	82,638	81,720	80,440	79,425
Deaths for All causes, n (%)	3,887 (4.7)	4,760 (5.8)	6,572 (8.2)	10,198 (12,8)
Model 1	1	1.21 (1.15–1.26) *	1.62 (1.55–1.69) *	2.59 (2.49–2.69) *
Model 2	1	1.19 (1.13–1.24) *	1.55 (1.48–1.61) *	2.35 (2.25–2.44) *
Model 3	1	1.19 (1.13–1.24) *	1.55 (1.48–1.62) *	2.38 (2.27–2.49) *
Model 4	1	1.19 (1.14–1.25) *	1.57 (1.51–1.65) *	2.56 (2.44–2.68) *
N: number of persons included in the analysis for each group; FPG: fasting plasma glucose; CV: coefficient of variation
Model 1: adjusted for age and sex. Model 2: adjusted for age, sex, history of cardiovascular disease, heart failure, and COVID-19 infection. Model 3: adjusted for variables in model 2 plus glycemic status, hypertension, atrial fibrillation, COPD, CKD, use of tobacco, dyslipidemia, use of antiplatelet drugs, use of statins, and SRA. Model 4: Model 3 plus mean FPG. *p < 0.001

Table 6

The odds ratios (ORs) of all-cause mortality grouped by quartiles of the coefficient of variation of FPG excluding patients with a history of cancer and those who died from COVID-19.
	Quartiles of CV of FPG
	1 (lowest)	2	3	4 (highest)
N	81,750	80,442	78,890	77,190
Deaths for All causes, n (%)	2,897 (3.6)	3,582 (4.5)	5,149 (6.5)	8,139 (10,5)
Model 1	1	1.22 (1.15–1.28) *	1.70 (1.62–1.78) *	2.76 (2.64–2.89) *
Model 2	1	1.20 (1.14–1.26) *	1.64 (1.56–1.72) *	2.55 (2.44–2.67) *
Model 3	1	1.20 (1.14–1.26) *	1.65 (1.57–1.74) *	2.63 (2.50–2.77) *
Model 4	1	1.21 (1.15–1.27) *	1.69 (1.61–1.77) *	2.88 (2.74–3.03) *
N: number of persons included in the analysis for each group; FPG: fasting plasma glucose; CV: coefficient of variation
Model 1: adjusted for age and sex. Model 2: adjusted for age, sex, history of cardiovascular disease, heart failure, and COVID-19 infection. Model 3: adjusted for variables in model 2 plus glycemic status, hypertension, atrial fibrillation, COPD, CKD, use of tobacco, dyslipidemia, use of antiplatelet drugs, use of statins, and SRA. Model 4: Model 3 plus mean FPG. *p < 0.001

Finally, we analysed the effect of GV on all-cause mortality among participants infected with SARS-CoV-2 to determine whether there were differences with respect to the global population analysis. As shown in Table 7, the more adjusted model (3 and 4) did not show significant associations between GV and mortality risk.

Table 7

Odds ratios (ORs) for all-cause mortality grouped by quartiles of the coefficient of variation of FPG in patients with COVID-19 infection and without a history of cancer.
	Quartiles of CV of FPG
	1 (lowest)	2	3	4 (highest)
N	2,325	2,512	2,987	3,646
Deaths for All causes, n (%)	544 (23.4)	562 (22.4)	737 (24.7)	1,065 (29,2)
Model 1	1	0.94 (0.82–1.08)	1.02 (0.89–1.16)	1.28 (1.13–1.45) *
Model 2	1	0.94 (0.82–1.07)	1.01 (0.88–1.14)	1.23 (1.09–1.39) *
Model 3	1	0.92 (0.80–1.05)	0.95 (0.83–1.09)	1.08 (0.94–1.24)
Model 4	1	0.92 (0.80–1.05)	0.95 (0.83–1.08)	1.05 (0.91–1.21)
N: number of persons included in the analysis for each group; FPG: fasting plasma glucose; CV: coefficient of variation
Model 1: adjusted for age and sex. Model 2: adjusted for age, sex, history of cardiovascular disease, and heart failure. Model 3: adjusted for variables in model 2 plus glycemic status, hypertension, atrial fibrillation, COPD, CKD, use of tobacco, dyslipidemia, use of antiplatelet drugs, use of statins, and SRA. Model 4: Model 3 plus mean FPG. *p < 0.001

It is well established that long-term GV is an independent predictor of all-cause mortality in patients with DM [23]. However, there is still insufficient evidence in the population without DM, at least when using quartiles of CV-FPG [14]. In this regard, a prospective cohort analysis in the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT) showed a similar effect of GV in those participants without DM and higher CV-FPG when used as a continuous variable, with a hazard ratio for all-cause mortality of 1.032 (95% CI, 1.014–1.049) in the most fully adjusted model [24].

Early analyses by our group showed that in individuals with prediabetes or T2DM, the fourth quartile of CV-FPG had a significant association with all-cause mortality after simple and full adjustment [16].

To our knowledge, only one study has examined the association between GV and mortality in the elderly population. This was a retrospective cohort study using the Health Improvement Network (THIN) database, which included 54,803 individuals aged 70 years and older in 587 UK primary care practices. All were diagnosed with diabetes, and GV was assessed by HbA1c variability over time. Higher HbA1c variability was associated with mortality, and a J-shaped curve was observed in the relationship between HbA1c thresholds and mortality [25].

Our study was conducted in the entire population aged 75 years and older living in Madrid (Spain) who had at least three FPG measurements during the follow-up period (2015–2020), and 29.3% were diagnosed with DM at baseline.

Patients with type 1 diabetes had a higher mortality rate for quartiles 1 and 2 of CV-FPG, with a mortality similar to that of patients with normoglycemia for the most extreme quartiles of CV-FPG, and higher than that of participants with T2DM. The latter may be because T2DM patients are more likely (Table 3) to take medications such as DPP-4 inhibitors and ISGLT-2, which have been shown to reduce vascular complications (e.g., heart failure and CKD). These findings also highlight the importance of glycemic variability, including small CV FPG, in both normoglycemic and diabetes mellitus participants, as we had previously found.

The fully adjusted model, which included the basal value of FPG, showed an OR for mortality that ranged from 2.48 to 2.88, according to different sensitivity analyses. Our results suggest that GV may be an important indicator and possibly an independent predictor of mortality in older people with and without diabetes.

With respect to physiopathology, higher GV has been associated with high protein expression of markers such as Wnt1 [26]. The Wnt signaling pathway causes at least two factors associated with mortality: first, it favours vascular calcification and regulates key aspects of vascular disease [27], as this calcification is more prevalent in the elderly than in the young and is highly associated with cardiovascular disease mortality [28]; second, the Wnt signaling pathway causes susceptibility to cancer [29]. On the other hand, elderly people tend to develop mitochondrial dysfunction, which increases oxidative damage during aging and metabolic diseases [30]. Additionally, GV “per se” has been associated with oxidative stress in patients with T2DM and hypertension [31], which increases the inflammatory response, vascular calcification [32], and endothelial damage, all of which lead to vascular complications and mortality.

Data from patients with acute injury, such as intracerebral hemorrhage, analysed by a recent meta-analysis [33], showed that those who had a higher category of standard deviation of blood glucose were associated with a higher risk of mortality (RR: 2.39, 95% CI: 1.79 to 3.19, p < 0.001). Other injuries, such as SARS-CoV-2 infection with acute respiratory distress syndrome, have shown similar results in a recent study of intensive care unit (ICU) patients: CV-FPG measured daily showed an adjusted OR for mortality of 12.83 (95% CI, 1.24-132.58) [34]. In our study, in patients with SARS-CoV-2 infection, GV was associated with a lower effect on all-cause mortality. This finding could be because in our case, we included a broad spectrum of patients with COVID-19: nonhospitalized, admitted to the ICU and hospitalized in beds outside the ICU.

Our study has several strengths, including its robust design to minimize bias and the large number of patients with diabetes, diabetes plus hypertension, and normoglycemia. In addition, to our knowledge, our study is the first to examine the relationship between variability in FPG and all-cause mortality in elderly patients with differences in glycemic status in southern European countries. This aspect is especially relevant, given the possible lower effect of GV on all-cause mortality in countries with healthier lifestyles [35] and better glycemic control than other countries participating in the EUROASPIRE IV survey [36].

This study has some limitations. First, we included patients with differences in glycemic status, and the analyses could not be adjusted for variables such as mean HbA1c, duration of diabetes, diabetic nephropathy, diabetes treatments, or microalbuminuria, as in other studies. Second, we did not have information on the cause of death, which would have enabled us to verify that mortality is primarily accounted for by cardiovascular disease, given the known association between GV and macrovascular complications. Third, we did not record hypoglycemia episodes and could not assess their association with mortality. Fourth, we could not study GV measured with CV-HbA1c, given that few persons with normoglycemia or IGT had at least three HbA1c measurements during follow-up. Fifth, given the observational nature of the present study, individuals with higher GV and lower GV were dissimilar. Therefore, adjusting for differences in both groups in the multivariate analysis was necessary to obtain an accurate picture of the association between all-cause mortality and GV. Propensity score matching (PSM) would be an appropriate alternative that would yield less biased results than standard methods such as logistic regression. However, given that propensity scores can only control for observed confounders, they cannot be counted upon to balance unobserved covariates.

GV has clear implications for clinical practice, as its assessment as a risk prediction tool should be included in the routine follow-up of the elderly and in a comprehensive geriatric assessment. Electronic medical records can incorporate tools that allow its calculation, and with this information, clinicians will have a broader view of the medium- and long-term prognosis of their patients.

CKD

Chronic kidney disease

COPD

Chronic obstructive pulmonary disease

COVID-19

Coronavirus disease 2019

EHRs

Electronic health records

Glucose variability

T2DM

Type 2 diabetes mellitus

T! DM

Type 1 diabetes mellitus

ICPC-2

the International Classification of Primary Care (ICPC-2).

FPG

Fasting plasma glucose

INDEF

Office of National Statistics death registrations

Coefficient of variation

CV-FPG

Coefficient of variation of fasting plasma glucose

Ethics approval and consent to participate.

The study was conducted according to the tenets of the Declaration of Helsinki and approved by the Regional Institutional Review Board of Madrid (Ref. CV_AGED-COVID-01-20). The consent of the participants was not required, as the data were de-identified before analysis for the purpose of confidentiality, according to the Spanish Organic Law 3/2018, of December 5, on the Protection of Personal Data and the Guarantee of Digital Rights.

Consent for publication

Not applicable in this study.

Availability of data and materials

The datasets used or analysed during the current study are available from the corresponding author upon reasonable request.

Competing interests

The authors declare no competing interests.

Funding

This study was funded by Instituto de Salud Carlos III (ISCIII) through the projects "PI18/01025”, “PI19/01569” and “PI22/01499” and was cofunded by the European Union. Additionally, complementary funding from the Sociedad Española de Arteriosclerosis and Regional Ministry of Health of the Community of Madrid through nonrefundable grants from the credits awarded to the Community of Madrid by the Spanish Government Fund COVID-19, included in Order HAC/667/2020, dated 17 July, which provides extraordinary credit in budget application 32.01.94°459.00 “Fondo COVID-19”.

Author’s contributions

MASF: Conceptualization, funding acquisition, conducted analyses, supervision and was a major contributor in writing the manuscript.

JSR: Conceptualization, supervision and reviewed the manuscript.

JCV: Collected data, made the visualization, interpreted data, and drafted the manuscript.

JM: Funding acquisition, interpreted data and reviewed the manuscript

CL, FRA, PGC, PVP, JMR, VIC and RJC: Interpreted data and reviewed the manuscript.

JMMY: Interpreted data and revised the final version of the manuscript.

All authors reviewed, read, and approved the final manuscript.

Acknowledgements

The Aged-Madrid Research group constituted by Gutiérrez-Misis A (Medicine School, UAM, Madrid), Carrillo de Santa Pau E (IMDEA Food, Madrid), Castell-Alcalá MV (Dr. Castroviejo Health Center, Madrid), Álvarez-Embarba B (Cruz Roja Nursing School, Madrid), Behzadi-Koochani N (SUMMA 112, Madrid), de Burgos-Lunar C (San Carlos Clinic Hospital, Madrid), Regueiro-Toribio P (Mar Báltico Health Center, Madrid), Gijón-Conde T (Cerro del Aire Health Center, Majadahonda, Madrid).

Noale M, Limongi F, Maggi S. Epidemiology of Cardiovascular Diseases in the Elderly. En: Veronese N, editor. Frailty and Cardiovascular Diseases [Internet]. Cham: Springer International Publishing; 2020 [citado 21 de septiembre de 2023]. p. 29–38. (Advances in Experimental Medicine and Biology; vol. 1216). Disponible en: http://link.springer.com/10.1007/978-3-030-33330-0_4.
Altman DG, Bland JM. Statistics notes: the normal distribution. BMJ 4 de febrero de. 1995;310(6975):298.
Yang L, Martikainen P, Silventoinen K. Effects of Individual, Spousal, and Offspring Socioeconomic Status on Mortality Among Elderly People in China. J Epidemiol 5 de noviembre de. 2016;26(11):602–9.
Leme DEdaC, Thomaz RP, Borim FSA, Brenelli SL, de Oliveira DV, Fattori A. Survival of elderly outpatients: effects of frailty, multimorbidity and disability. Cienc Saude Coletiva enero de. 2019;24(1):137–46.
Fortes C, Mastroeni S, Sperati A, Pacifici R, Zuccaro P, Francesco F, et al. Walking four times weekly for at least 15 min is associated with longevity in a cohort of very elderly people. Maturitas marzo de. 2013;74(3):246–51.
Koolhaas CM, Dhana K, Schoufour JD, Lahousse L, van Rooij FJA, Ikram MA, et al. Physical activity and cause-specific mortality: the Rotterdam Study. Int J Epidemiol 1 de octubre de. 2018;47(5):1705–13.
Wu Xguang, Tang Z, Fang X, hua, Liu H jun, Diao L jun, Xiang M. [Evaluation of predictive effect of some health-related indices on deaths among ageing residents through a 8-years’ follow-up study in Beijing]. Zhonghua Liu Xing Bing Xue Za Zhi Zhonghua Liuxingbingxue Zazhi abril de. 2004;25(4):325–8.
Liu H, Xiao Q, Cai Y, Li S. The quality of life and mortality risk of elderly people in rural China: the role of family support. Asia Pac J Public Health marzo de. 2015;27(2):NP2232–2245.
Lee J, Kim HS, Song KH, Yoo SJ, Han K, Lee SH et al. Risk of Cause-Specific Mortality across Glucose Spectrum in Elderly People: A Nationwide Population-Based Cohort Study. Endocrinol Metab Seoul Korea. 7 de septiembre de 2023;.
Song MS, Choi YJ, Kim H, Nam MJ, Lee CW, Han K, et al. Relationship between blood pressure levels and ischemic stroke, myocardial infarction, and mortality in very elderly patients taking antihypertensives: a nationwide population-based cohort study. BMC Geriatr 2 de noviembre de. 2021;21(1):620.
Huh Y, Kim DH, Jung JH, Park YG, Roh YK, Kim SM, et al. Risk Assessment of Mortality in Elderly Individuals: A Nationwide Cohort Study. Gerontology. 2022;68(11):1266–75.
Ahmed N, Choe Y, Mustad VA, Chakraborty S, Goates S, Luo M, et al. Impact of malnutrition on survival and healthcare utilization in Medicare beneficiaries with diabetes: a retrospective cohort analysis. BMJ Open Diabetes Res Care. 2018;6(1):e000471.
Connors MH, Sachdev PS, Kochan NA, Xu J, Draper B, Brodaty H. Cognition and mortality in older people: the Sydney Memory and Ageing Study. Age Ageing noviembre de. 2015;44(6):1049–54.
Barzegar N, Ramezankhani A, Tohidi M, Azizi F, Hadaegh F. Long-term glucose variability and incident cardiovascular diseases and all-cause mortality events in subjects with and without diabetes: Tehran Lipid and Glucose Study. Diabetes Res Clin Pract agosto de. 2021;178:108942.
Klimontov VV, Saik OV, Korbut AI. Glucose Variability: How Does It Work? Int J Mol Sci 21 de julio de. 2021;22(15):7783.
Salinero-Fort MA, San Andrés-Rebollo FJ, Cárdenas-Valladolid J, Mostaza JM, Lahoz C, Rodriguez-Artalejo F, et al. Glycemic variability and all-cause mortality in a large prospective southern European cohort of patients with differences in glycemic status. PLoS ONE. 2022;17(7):e0271632.
Saqib MAN, Siddiqui S, Qasim M, Jamil MA, Rafique I, Awan UA, et al. Effect of COVID-19 lockdown on patients with chronic diseases. Diabetes Metab Syndr. 2020;14(6):1621–3.
Kendzerska T, Zhu DT, Gershon AS, Edwards JD, Peixoto C, Robillard R, et al. The Effects of the Health System Response to the COVID-19 Pandemic on Chronic Disease Management: A Narrative Review. Risk Manag Healthc Policy. 2021;14:575–84.
de Burgos-Lunar C, Del Cura-Gonzalez I, Cárdenas-Valladolid J, Gómez-Campelo P, Abánades-Herranz JC, Lopez-de-Andres A, et al. Validation of diagnosis of acute myocardial infarction and stroke in electronic medical records: a primary care cross-sectional study in Madrid, Spain (the e-MADVEVA Study). BMJ Open 12 de junio de. 2023;13(6):e068938.
de Burgos-Lunar C, Del Cura-González I, Cárdenas-Valladolid J, Gómez-Campelo P, Abánades-Herranz JC, López-de Andrés A et al. Real-world data in primary care: validation of diagnosis of atrial fibrillation in primary care electronic medical records and estimated prevalence among consulting patients’. BMC Prim Care. 4 de enero de 2023;24(1):4.
de Burgos-Lunar C, Salinero-Fort MA, Cárdenas-Valladolid J, Soto-Díaz S, Fuentes-Rodríguez CY, Abánades-Herranz JC et al. Validation of diabetes mellitus and hypertension diagnosis in computerized medical records in primary health care. BMC Med Res Methodol. 28 de octubre de 2011;11:146.
Salinero-Fort MA, Mostaza J, Lahoz C, Cárdenas-Valladolid J, Vicente-Díez JI, Gómez-Campelo P, et al. All-cause mortality and cardiovascular events in a Spanish nonagenarian cohort according to type 2 diabetes mellitus status and established cardiovascular disease. BMC Geriatr 18 de marzo de. 2022;22(1):224.
Chen J, Yi Q, Wang Y, Wang J, Yu H, Zhang J, et al. Long-term glycemic variability and risk of adverse health outcomes in patients with diabetes: A systematic review and meta-analysis of cohort studies. Diabetes Res Clin Pract octubre de. 2022;192:110085.
Echouffo-Tcheugui JB, Zhao S, Brock G, Matsouaka RA, Kline D, Joseph JJ. Visit-to-Visit Glycemic Variability and Risks of Cardiovascular Events and All-Cause Mortality: The ALLHAT Study. Diabetes Care marzo de. 2019;42(3):486–93.
Forbes A, Murrells T, Mulnier H, Sinclair AJ. Mean HbA1c, HbA1c variability, and mortality in people with diabetes aged 70 years and older: a retrospective cohort study. Lancet Diabetes Endocrinol junio de. 2018;6(6):476–86.
Zhang L, Sun H, Liu S, Gao J, Xia J. Glycemic variability is associated with vascular calcification by the markers of endoplasmic reticulum stress-related apoptosis, Wnt1, galectin-3 and BMP-2. Diabetol Metab Syndr. 2019;11:67.
Foulquier S, Daskalopoulos EP, Lluri G, Hermans KCM, Deb A, Blankesteijn WM. WNT Signaling in Cardiac and Vascular Disease. Pharmacol Rev enero de. 2018;70(1):68–141.
Wu M, Rementer C, Giachelli CM. Vascular calcification: an update on mechanisms and challenges in treatment. Calcif Tissue Int octubre de. 2013;93(4):365–73.
Taciak B, Pruszynska I, Kiraga L, Bialasek M, Krol M. Wnt signaling pathway in development and cancer. J Physiol Pharmacol Off J Pol Physiol Soc abril de 2018;69(2).
López-Lluch G, Hernández-Camacho JD, Fernández-Ayala DJM, Navas P. Mitochondrial dysfunction in metabolism and ageing: shared mechanisms and outcomes? Biogerontology. diciembre de 2018;19(6):461 – 80.
Ohara M, Kohata Y, Nagaike H, Koshibu M, Gima H, Hiromura M, et al. Association of glucose and blood pressure variability on oxidative stress in patients with type 2 diabetes mellitus and hypertension: a cross-sectional study. Diabetol Metab Syndr. 2019;11:29.
Mody N, Parhami F, Sarafian TA, Demer LL. Oxidative stress modulates osteoblastic differentiation of vascular and bone cells. Free Radic Biol Med. 15 de agosto de 2001;31(4):509–19.
Jiao X, Wang H, Li M, Lu Y. Glycemic Variability and Prognosis of Patients with Intracerebral Hemorrhage: A Meta-Analysis. Horm Metab Res Horm Stoffwechselforschung Horm Metab marzo de. 2023;55(3):176–83.
Hartmann B, Verket M, Balfanz P, Hartmann NU, Jacobsen M, Brandts J, et al. Glycaemic variability is associated with all-cause mortality in COVID-19 patients with ARDS, a retrospective subcohort study. Sci Rep 14 de junio de. 2022;12(1):9862.
Sotos-Prieto M, Ortolá R, Ruiz-Canela M, Garcia-Esquinas E, Martínez-Gómez D, Lopez-Garcia E, et al. Association between the Mediterranean lifestyle, metabolic syndrome and mortality: a whole-country cohort in Spain. Cardiovasc Diabetol 5 de enero de. 2021;20(1):5.
Kotseva K, Wood D, De Bacquer D, De Backer G, Rydén L, Jennings C, et al. EUROASPIRE IV: A European Society of Cardiology survey on the lifestyle, risk factor and therapeutic management of coronary patients from 24 European countries. Eur J Prev Cardiol abril de. 2016;23(6):636–48.

No competing interests reported.

Download PDF

Editorial decision: Revision requested
07 Feb, 2024
Reviews received at journal
04 Feb, 2024
Reviews received at journal
14 Nov, 2023
Reviewers agreed at journal
06 Nov, 2023
Reviewers agreed at journal
01 Nov, 2023
Reviewers invited by journal
25 Oct, 2023
Editor assigned by journal
25 Oct, 2023
Editor invited by journal
18 Oct, 2023
Submission checks completed at journal
18 Oct, 2023
First submitted to journal
05 Oct, 2023

You are reading this latest preprint version

Effect of Glucose Variability on the Mortality of the Very Old People During the First Year of the COVID-19 Pandemic

Status:

Version 1

Abstract

Background

Methods

Results

Conclusions

Figures

Background

Material and Methods

Statistical analysis

Results

Discussion

Conclusion

Abbreviations

Declarations

References

Additional Declarations

Status:

Version 1