No association between glucose at age 30 and aortic diameter at age 65 in men: a population-based study

Abstract

Objectives Impaired glucose metabolism and diabetes mellitus has been linked to a decreased risk for abdominal aortic aneurysm development in men. We evaluated potential relationships between blood glucose levels in 1722 men with mean age 34 years and their aortic diameter measured by ultrasound at age 65 years. Design Prospective cohort study. Results Mean follow-up between baseline glucose investigation and aortic ultrasound was 32.8 ± 1.8 years. There was no correlation between baseline blood glucose and later aortic diameter (r = 0.035, p = 0.146), whereas a weak correlation between body mass index (BMI) and aortic diameter was observed (r = 0.097 p < 0.001). In a partial correlation between aortic diameter and glucose levels adjusted for BMI, the correlation did not change (r = 0.011, p = 0.66). Neither were there any significant differences in mean aortic diameter between men belonging to different quartiles of baseline blood glucose levels (19.5, 19.1, 19.6 and 19.7 mm, p = 0.18). Conclusion Fasting blood glucose in 33-year-old men was not associated with their aortic diameter at age 65 years. Potential pathophysiological processes linking disturbed glucose metabolism to a decreased risk for development of abdominal aortic aneurysm in men are therefore probably not relevant until later in life.

Keywords:

• Abdominal aortic aneurysm
• aortic diameter
• blood glucose

Introduction

Risk factors for abdominal aortic aneurysm (AAA) are partly the same as for atherosclerotic disease in general,[1–5] and many AAA-patients have manifestations of concomitant atherosclerosis. Inflammatory mediators, endothelial function and the coagulation system all interact with genetics and classic risk factors (hypertension, hyperlipidemia and smoking) in the pathogenesis of both conditions. On the other hand, type 2 diabetes mellitus (DM2) has epidemiologically been associated with reduced risk for AAA development and growth.[4–6] This association has recently been confirmed in a meta-analysis of 64 studies together comprising more than three million individuals.[7] AAA mortality has also been demonstrated to be inversely related to obesity,[8] an association termed the obesity paradox. These epidemiological data are supported by experimental results showing that hyperglycemia inhibits experimental AAA formation,[9] which might be related to factors such as glycation of arterial wall proteins,[9] arterial stiffness and vascular aging measured as carotid-femoral pulse wave velocity (c-f PWV),[10] vascular calcification [11] or down-regulation of matrix metalloproteinases (MMPs) and chymases.

Rupture of an AAA (rAAA) has a mortality rate of 80–90%. However, AAA-related mortality can be reduced in a cost-effective manner by ultrasound screening for AAA in 65-year-old men, and preventive operative treatment of screening-detected large AAAs,[12–15] whereas screening in older men does not represent the same benefits.[14] Therefore, population-based screening programs for AAA among 65-year-old men have been launched in several county councils,[16,17] including Malmö, Sweden.[17]

In contrast to data based on incident AAA-cases or cases of AAA rupture as reported in previous studies,[6] screening programs offer an opportunity to evaluate aortic diameter in population-based cohorts of 65-year-old men which can be related to baseline date obtained from the same subjects when studied at an earlier age.

The aim of this observational study was to evaluate if the proposed relationship between blood glucose levels and AAA development [4–7] can be demonstrated by analyzing baseline levels of blood glucose in men at around age 30 years in relation to the aortic diameter at age 65 years, i.e. after more than 30 years of follow-up.

Subjects and methods

Subjects

The Malmö Preventive Project, MPP,[18] which started in 1974 at the Department of Preventive Medicine, Skåne University Hospital, Malmö, was used for this study. The aim of MPP was to create a case-finding program for intervention by screening a large stratum of the middle-aged population for CVD risk factors, impaired glucose tolerance and alcohol abuse. From the 1980s women and risk factors for breast cancer were also included.[18]

Subjects were invited to participate in a health screening, including physical examination, blood sampling for laboratory testing, as well as filling an extensive self-administered questionnaire with 260 questions concerning medical and family history, lifestyle and social background characteristics. Although the questionnaire evolved during the years it also resulted in some inconsistencies. More detailed information concerning the screening procedures and the main results from MPP has been previously published.[18] Between 1974 and 1992, a total of 22 444 male and 10 902 female subjects attended the MPP (71% attendance rate).

After ∼33 years all men aged 65 years in the area were invited to participate in AAA screening program at Skåne University Hospital, launched in 2010. In total, 4300 65-year-old men in Malmö and its surrounding municipalities are invited on an annual basis to undergo ultrasound screening at the Department of Vascular Diseases [17] and to date 17 763 men have been examined. The invitation includes a pamphlet containing information about AAA and the screening procedure, and a scheduled appointment for examination. The examination included ultrasound measurement of the abdominal aorta and a health questionnaire including questions on smoking habits, educational level and concomitant diseases. Subjects participating in the on-going AAA screening who had previously participated also in the MPP study were included in this study (n = 1722).

The study was approved by the Ethical Committee at the Lund University 2014/643.

Physical examination at baseline

All subjects were examined for height (m) without shoes, and weight (kg) in light indoor clothing. The body mass index (BMI) was calculated (kg/m²). Blood pressure (mmHg) was measured twice in the supine position after 10 min rest by use of a sphygmomanometer with a modifiable cuff width, and a mean figure was recorded. Blood samples were drawn after an overnight fast. Serum total cholesterol, serum triglycerides and blood glucose were analyzed, using routine methods at the Department of Clinical Chemistry, Skåne University Hospital.

Ultrasound examination at AAA screening

The ultrasound (US)-based AAA screening was performed by biomedical scientists and registered vascular nurses using the LOGIQ® ultrasound machine and 3.5–12 MHz probes (General Electric Healthcare Inc., Chalfont St. Giles, UK). In patients (1.1%) with inconclusive US results, a computed tomography (CT) scan without contrast was done to measure the diameter of the aorta. The diameter was measured according to the “leading-edge to leading-edge” principle,[19] meaning that the maximal transverse aortic diameter is measured postero-anteriorly between the outer vessel walls. The widest diameter of the infra-renal aorta was measured, and a diameter of ≥ 30 mm was considered as an AAA.

Statistics

Baseline characteristics were examined in total cohort and in quartiles of aortic diameter as mean (standard deviation, SD), median or percentages. The distribution of triglycerides and glucose were skewed and therefore log-transformed. ANOVA for continuous and chi-square for dichotomous variables were used to test differences between mean levels or proportions comparing participants in quartiles of aortic diameter. To examine if there was any relationship between aortic diameter at screening and previous blood glucose in MPP, Spearman’s correlation and a linear regression were performed. We constructed quartiles of baseline blood glucose levels to explore if there were any significant differences in mean value of aortic diameter. A p value <0.05 was considered significant. All analyses were performed using SPSS 22.0 (Chicago, IL).

Results

Baseline characteristics of the total cohort and in quartiles of aortic diameter range are shown in Table 1. Mean follow-up time between baseline investigation and AAA screening was 32.8 ± 1.8 years. There were significant associations between baseline weight, height and BMI and increased aortic diameter. An inverse relationship was noticed for heart rate and increased aortic width (Table 1). In further exploration with linear regression analysis the aortic diameter was significantly related to these variables (Table 2). There was no correlation between baseline blood glucose and aortic diameter at age 65 years (r = 0.035, p = 0.15), whereas a weak correlation was observed between BMI and aortic diameter (r = 0.097 p < 0.001). In a partial correlation between aortic diameter and glucose levels adjusted for BMI the relation did not change (r = 0.011, p = 0.66). Neither were there any significant differences in mean aortic diameter between men belonging to different quartiles of baseline blood glucose (19.5, 19.1, 19.6 and 19.7 mm, p = 0.19, Table 3). There was no significant difference in mean glucose levels between those with aortic diameter wider than 30 mm or not, mean glucose value was 5.0 mmol/L in both groups. During follow-up, in all 177 men developed diabetes, but there was no significant difference in mean aortic diameter between this group and men without diabetes (19.9 ± 4.6 mm versus 19.4 ± 3.4 mm, p value = 0.21).

Table 1. Baseline characteristics of the total cohort and in quartiles of aortic diameter.

	Total cohort	Q1	Q2	Q3	Q4	p Value*
Aortic diameter range (mm)		10–17.5	17.6–19.0	19.1–20.6	20.7–58
mean value		16.2	18.2	19.7	23.8
	N = 1722
Age (years)	33.6 ± 2.7	33.6	33.5	33.6	33.8	0.25
Height (cm)	178.8 ± 6.7	176.9	177.9	179.9	180.6	<0.001
Weight (kg)	75.7 ± 10.8	72.3	74.0	77.4	78.9	<0.001
Body mass index (kg/m^2)	23.7 ± 3.0	23.1	23.4	23.9	24.2	<0.001
Heart rate (bpm)	65.7 ± 9.5	66.8	66.6	65.2	64.3	0.001
Systolic blood pressure (mmHg)	122.6 ± 10.7	122.9	121.9	123.0	122.6	0.98
Diastolic blood pressure (mmHg)	81.7 ± 7.8	81.9	81.2	81.3	82.3	0.36
Total cholesterol (mmol/L)	5.1 ± 0.7	5.22	5.10	5.08	5.17	0.43
Triglycerides (mmol/L)	1.09 (0.8–1.5)	1.24	1.25	1.18	1.32	0.81
Fasting glucose (mmol/L)	5.0 (4.7–5.3)	5.02	4.99	5.04	5.07	0.11
Hemoglobin (g/L)	147.9 ± 9.0	148.4	147.6	147.8	147.9	0.50
Leukocytes (10^9/L)	5.6 ± 1.9	5.7	5.3	5.5	5.9	0.49
Sedimentation rate (mm)	3.0 (2.0–5.0)	3.85	4.09	4.04	4.13	0.79
Creatinine (μmol/L)	93.2 ± 27.9	92.3	95.9	93.4	91.4	0.39
Current smoker (%)	41.0	10.8	8.5	10.0	11.7	0.54

* p Value test for trend.

Table 2. Linear regression model for aortic diameter in relation to associated factors.

	β	(95%CI)	p Value
BMI	0.114	(0.0348 to 0.180)	0.001
Heart rate	−0.039	(−0.059 to −0.019)	<0.001
Smoking	0.687	(0.301 to 1.074)	<0.001

Table 3. Aortic diameter at ultrasound screening at 65 years of age in relation to quartiles of blood glucose levels at the baseline examination.

	Q1	Q2	Q3	Q4
Fasting glucose (mmol/L)	3.0–4.7	4.8–4.9	5.0–5.2	5.3–17.2
n = 1722	477	342	432	471
Aortic diameter (mm)	19.5 ± 4.4	19.1 ± 2.8	19.6 ± 4.0	19.72 ± 4.5

p Value for trend =0.18

Discussion

We could not demonstrate any association between blood glucose levels at around age 30 years and aortic dilatation or aneurysm development at age 65 years in our study. There might be several explanations for our findings, whereof the most plausible is that our subjects were of comparably young age at the baseline examination and with apparently normal glucose metabolism. Many of previous epidemiological studies showing relationships between markers of glucose metabolism and aneurysm prevalence have been cross-sectional cohort studies by design,[4,5,7,20] but DM2 has been associated with a lower aneurysm incidence also during long-term follow-up for between 6 and 30 years.[6,7,21–23] These study subjects were, however, clearly older than the men studied in MPP who later underwent AAA-screening, as shown in our report. In the study by Shah et al. [6] mean ages at baseline for men were 45 years in the group without diabetes and 61 years in the group with diabetes; in Iribarren’s study [21] men were 44 years of age at the beginning of follow-up. The study of Törnwall et al. [22] was restricted to smokers, and had a follow-up of only 6 years. In a study with a comparable follow-up as ours of up to 30 years,[23] Rodin studied subjects aged 40–64 years at baseline with regard to clinically detected AAA during follow-up. Associations between glucose metabolism and AAA have been studied previously in our region; Ohrlander et al. [24] showed that hospitalization for a diabetes-related diagnosis was inversely related to AAA incidence. Again, however, this study addressed different research questions than ours. The authors studied clinically diagnosed cases of diabetes and AAA disease, whereas we assessed subclinical disturbances in glucose metabolism and subtle differences in aortic diameter. It has also been noted that risk markers in general tend to decrease in predictive power after extended follow-up.

Furthermore, we found a weak association between BMI and aortic diameter, as well as an unexpected inverse relation to heart rate at the baseline examination at age 30 and aortic diameter at age 65. Relationships between obesity and AAA are complex; Stackelberg et al. [25] reported a positive association between AAA and waist circumference but not between AAA and BMI. Waist–hip ratio was not assessed at our baseline examination.

Hemoglobin A1_c (HbA_1c) is accepted as an index of glycemic control, and is the best risk marker of diabetic microvascular complications.[26] The question remains, however, if the ambient mean glycemic control over a 2- to 3-month period, would significantly correlate with aortic diameter, however such relationships have not yet been proven,[27] and HbA_1c was not analyzed in the MPP study.

To further explore the relationships between disturbances of glucose metabolism and decreased risk for aortic aneurysm formation, we are currently comparing 65-year-old men with and without aneurysms (>30 mm) concerning variables reflecting both glucose metabolism and aortic stiffness.

One obvious limitation is that we studied only men, and the relatively limited sample size constitutes another limitation of the study. Of the 22 444 men attending the MPP examination at baseline, the majority had already duly celebrated their 65th birthday when screening for AAA was started in 2010, and only 1722 men could therefore be included in this set of analyses. The fact that neither subjects non-compliant with the screening invitation, nor those not surviving to age 65 years could be included in the study, introduces another potential selection bias.

Non-attendance with screening programs is associated with lower socioeconomic status (SES) and higher morbidity,[28] and there is also a relationship between low SES and the presence of DM2.[29] In a previous study, subjects in geographical areas with low SES were shown to have both higher AAA prevalence and lower compliance to AAA screening.[17] Therefore, subjects with severe DM2 complications (foot ulcers and amputations) might also have been less likely to attend screening for AAA.

In conclusion, fasting blood glucose levels in 34-year-old men was not associated with their aortic diameter at ultrasound examination at age 65 years. Potential pathophysiological processes linking impaired glucose metabolism to a decreased risk for development of abdominal aortic aneurysm in men are therefore probably not relevant until later in life.

Disclosure statement

This study was supported by grants from the Ernhold Lundström Foundation, Research Funds at Skåne University Hospital and at Region Skåne, the Hulda Ahlmroth Foundation and from the Swedish Government under the LUA/ALF agreement.