Association between Sleep Duration and Left Ventricular Hypertrophy for Patients with Type 2 Diabetes Mellitus

Objective In this study, we aimed to estimate the impact of sleep duration on left ventricular hypertrophy (LVH) in type 2 diabetes mellitus (T2DM). Methods Consecutive patients with T2DM undergoing transthoracic echocardiography (TTE) in our center from October 2017 to February 2021 were analyzed. The association of the risk of LVH in T2DM patients was evaluated using univariable and multivariable logistic regression analyses. Results This study finally included 2689 adult patients (mean age 51.8 ± 12.5 years, 56.2% men, mean sleep duration 7.6 ± 1.4 hours per day). Of all patients, 655 (24.4%) patients were diagnosed with LVH and 2034 did not have LVH. All patients were adults and were diagnosed with T2DM. In the univariate and multivariate regression analyses, gender, sleep duration, body mass index (BMI), waist, hemoglobin (Hb), blood creatinine (Cr), and high-density lipoprotein cholesterol (HDL-c) were associated with LVH. In the restricted cubic spline (RCS) model, the cut-off points of sleep duration given refer to the group of patients with T2DM and LVH were 8 hours per day. With the cut-off points, the multivariable analysis demonstrated that, for diabetic patients, LVH was significantly correlated with a sleep duration of 8 hours per day, hemoglobin, blood urea nitrogen (BUN), and HDL-c. Conclusion For patients with T2DM, long sleep duration (>8 hours per day), hemoglobin, BUN, and HDL-c were independently associated with LVH. This trial is registered with NCT03811470.


Background
Previous studies have shown an increasing incidence and prevalence of diabetes [1][2][3].As reported in Denmark, the overall prevalence of diabetes in 2016 was 4.4% for type 2 diabetes mellitus (T2DM), with annual increases for T2DM since 1996 of 5.5% [1].In Iceland, the prevalence of T2DM more than doubled in the period of 2005-2018 [2].In China, in 2018, the prevalence of diabetes was about 12.4% and the estimated prevalence of prediabetes was 38.1% [3].
Diabetes has caused a tremendous economic burden in world.Te estimated global direct health expenditure on diabetes in 2019 is USD 760 billion and is expected to grow to a projected USD 825 billion by 2030 and USD 845 billion by 2045 [4].In the United States, the total cost of diabetes in 2017 was USD 327 billion and more than half of the expenditure was attributable to T2DM [5].Patients in the United States with diabetes incurred average medical expenditures of about USD 16,750 per year [4].China had one of the highest estimated total costs of diabetes with USD 109.0 billion [4].
Diabetes is a progressive disorder causing numerous complications, including small vessel or microvascular disease and large vessel or macrovascular disease [6].Microvascular complications could afect the kidney, retina, and peripheral nerves termed as diabetic nephropathy, diabetic retinopathy, and diabetic neuropathy, respectively [7].Te macrovascular complications that afect the heart are identifed as cardiovascular disease, the complications affecting the peripheral arteries are termed as peripheral vascular disease, and those afecting the brain are termed as cerebrovascular disease [8].Cardiac hypertrophy is one of the manifestations of diabetic cardiomyopathy [9].Metabolic disturbances caused by diabetes could promote cardiac remodeling, fbrotic diastolic dysfunction, and, ultimately, decreased ejection fraction [10].
Sleep is a public health concern afecting up to one-third of the population, and a large body of evidence has proved the association between insufcient sleep duration and quality and the risk of obesity, insulin resistance, and T2DM [11].As reported, sleep disorders are highly prevalent among patients with T2DM [12].Previous studies have proved a Ushaped association between sleep duration and incidence of T2DM, reporting 7-8 hours of sleep per night as the lowest risk [13].In observational studies, short sleep duration was associated with both a higher risk of incidence of T2DM and may also predict worse outcomes in those with existing diabetes [12].Experimental and clinical studies have suggested an independent association between obstructive sleep apnea (OSA) and left ventricular mass (LVM), but there are few studies evaluating the impact of sleep duration on the LVM for diabetic patients.
In this study, we aimed to estimate the impact of sleep duration on the LVM for T2DM patients.

2.1.
Participants.Tis study is a large-scale single-center retrospective study performed at the Beijing Luhe Hospital.From October 2017 to February 2021, we retrospectively reviewed 3074 consecutive patients undergoing transthoracic echocardiography (TTE) in the Center for Endocrine Metabolism and Immune Disease.Tis study was registered at clinicaltrials.govwith the registry name "Serological, Metabolomics and Genomics Studies of Metabolic Diseases" and allotted the identifer NCT03811470.Tis study complied with the tenets of the Declaration of Helsinki.All patients gave their informed consent prior to their inclusion in the study.
Te inclusion criteria for this study were as follows: all participants were diagnosed with T2DM.All participants were adults.Participants underwent a cardiac ultrasound, and detailed cardiac measurements were recorded.Participants had detailed health records.
Te exclusion criteria for this study were as follows: patients <18 years of age were excluded.Patients with structural heart disease, valvular heart disease, hypertrophic cardiomyopathy, or undergoing cardiac surgery were excluded.

Data Collection.
All clinical data were acquired via a search of the Hospital Information System.Te information retrieved included age, gender, blood pressure (BP), heart rate (HR), height, weight, body mass index (BMI), diabetes duration, smoke history, alcohol intake history, hemoglobin (Hb), fasting blood glucose, hemoglobin A1C (HbA1c), blood lipid level, aminotransferase level, thyroid function and history of hypertension, hyperlipidemia, hyperuricemia (HUA), gout, coronary heart disease (CHD), and vascular disease.As reported, sleep duration was assessed by asking participants "What time do you usually go to bed?When get up?" and we calculated the sleep duration based on the results [14].Left ventricular hypertrophy (LVH), which can be measured by TTE, was represented by the left ventricular mass index (LVMI).
LVM was calculated according to the following formula [15]: IVSD: interventricular septum diameter; LVEDD: left ventricular end-systolic diameter; LVPWD: left ventricular posterior wall diameter; BSA: body surface area.As recommended, LVH was defned as a LVMI of more than 102 g/m 2 for males or 88 g/m 2 for females [16].

Statistical Analysis.
As shown in Figure 1, the proportion of missing variables were not more than 5%, so participants with missing data were included.Finally, 2689 T2DM patients were included in the study.
Continuous variables were presented as the mean-± standard deviation (SD) or medians (interquartile ranges) and compared by Student's t-tests or Mann-Whitney U tests.Categorical variables were presented as number (proportion) and compared by Pearson's χ 2 test.Comparisons were made between the variables in patients with LVH and the control group.A P value of less than 0.05 was considered signifcant.
Te association of the risk of LVH in patients with T2DM was evaluated using univariable logistic regression analysis.Te variables with a P value of less than 0.1 were entered into the backward regression multivariable analysis.

Results
A total of 2689 T2DM patients were fnally included in the study.Out of the 2689 patients, 655 (24.4%) patients were diagnosed with LVH and 2034 patients did not have LVH, who formed the control group.Baseline demographic information is provided in Table 1.Te mean age was 51.8 ± 12.5 years; 56.2% were men; the mean BMI was 26.7 ± 4.0 kg/m 2 ; the mean waist measurement was 94.3 ± 10.8 cm; the mean HR was 83.9 ± 12.6 bpm; the mean systole blood pressure (SBP) was 132.6 ± 17.7 mmHg; the mean diastolic blood pressure (DBP) was 80.0 ± 11.9 mmHg; the mean sleep duration was 7.6 ± 1.4 hours per day.Of the participants, 1131 (42.1%) patients had hypertension, 1214 (45.1%) patients had hyperlipidemia, 312 (11.6%) had HUA, 59 (2.2%) had gout, and 347 (12.9%) had CHD.Other clinical characteristics of patients are summarized in Table 1.
As shown in Figure 2, a restricted cubic spline (RCS) model was used to obtain the cut-of point of sleep duration for LVH.Te cut-of point of sleep duration was 8 hours.

Discussion
In the present study, we conducted a single-center, observational, retrospective study.A total of 2689 consecutive patients with T2DM, for whom transthoracic echocardiography (TTE) was performed and LVMI was calculated, were included in the study.Compared with the control group, patients with LVH were older and had a longer sleep duration per day.In the RCS model, the cut-of point of sleep duration was 8 hours per day.Multivariable analysis demonstrated that LVH was signifcantly correlated with a sleep duration of 8 hours per day (OR � 1.563, P < 0.001), Hb (OR � 1.009, P � 0.001), BUN (OR � 1.093, P � 0.002), and HDL-c (OR � 0.612, P � 0.001) for T2DM patients.
To our knowledge, this is the frst study evaluating sleep duration for the prediction of LVH.As a complication of common cardiovascular disorders (such as coronary atherosclerosis and myocardial infarction) and metabolism syndrome, LVH, which could progress to heart failure, needs early prevention and treatment [17][18][19].Although many drug therapies, nutraceutical and dietary measures are available for the management of ventricular hypertrophy, most patients nonetheless experience a downhill course [20].Positive changes in lifestyle, a safe and inexpensive

Histogram of missing data
Figure 1: Distribution of missing data in the study.In this study, the proportion of missing variables was not more than 5%.Te variable with the most missing is FT3 (∼5.0%).FT3, serum-free triiodothyronine; RBC, red blood cell; HDL-c, high-density lipoprotein cholesterol; TC, total cholesterol; ALT, alanine aminotransferase; HBA1c, hemoglobin A1C; Cr, creatinine; DBP, diastolic blood pressure; HR, heart rate; DM, diabetes mellitus; HUA, hyperuricemia.
International Journal of Endocrinology measure, may have merit in the management of ventricular hypertrophy, while the impact of sleep duration on ventricular hypertrophy remains to be fully understood.In the present study, we found that sleep duration may not be as long as possible, and a sleep duration longer than 8 hours may be one of the risk factors for ventricular hypertrophy.
Our study has several strengths.First, we conducted the study among a relatively high number of individuals (n � 2689).Second, we used RCS analysis to search the optimal cut-of point rather than the cut-point recommended in the guidelines, as the cut-of point of sleep duration to LVH may be unique.Tird, the model was adjusted for duration of diabetes, as diabetes' duration may be associated with LVH.
Previous studies had proved that a long sleep duration was associated with an increased risk of all-cause mortality and cardiovascular events [21,22].
Several potential mechanisms may contribute to the association between long sleep duration and LVH.First, getting too much sleep can be a risk factor for hypertension, and hypertension is an important risk factor for cardiac hypertrophy.Sleep duration and BP are linked in a complex way.After adjusting for confounders such as age, BMI, and alcohol consumption, a U-shaped curve was reported between sleep duration and the risk of hypertension [23].Sleep duration more than or equal to 9 hours was a risk factor of hypertension, and the same result was found for sleep duration more than or equal to 10 hours [24].Second, long sleep duration may be a representation of OSA, a known 4 International Journal of Endocrinology cause of increased need for sleep and an identifed risk factor for ventricular hypertrophy [25].Patients with OSA may be at risk for LVH because of their poorer quality of sleep, such that they sleep for longer periods of time, more often than not in a refreshing sleep state, and often with an early onset of headache [26].Tird, spending a long time in bed may be linked to sleep fragmentation [27], which was associated with more severe arteriolosclerosis and LVH.Changes in infammatory markers (such as C-reactive protein (CRP), interleukin-6 (IL-6), and tumor necrosis factor-α (TNF-α)) are associated with long sleep duration, which may contribute to ventricular hypertrophy [28].In the present study, we found that HDL-c was negatively associated with LVH.Schillaci et al. found that low HDL-c was an independent predictor of LVM in untreated hypertensive patients [29], which is consistent with our results.Previous studies have proved that premenopausal women, as compared to men, are less prone to develop ventricular hypertrophy, while this protection is lost after menopause [30,31].In our study, being female was an independent predictor of LVH for T2DM patients, as the participants in our study were older, with a mean age of 51.8 ± 12.5 years, and most female patients were postmenopausal.
Several limitations of our study should be acknowledged.First, all periods of sleep duration were self-reported by questionnaire or interview instead of detection using sleep monitoring equipment.Second, factors such as sleep apnea are an independent predictor of the risk of LVH but were not assessed in our study.Tird, this is a single-center study, and any extrapolation of the conclusions may be limited.

Conclusion
For patients with T2DM, a long sleep duration (>8 hours per day), Hb, BUN, and HDL-c were independently associated with LVH.

Figure 2 :
Figure 2: Te cut-of points of sleep duration for LVH.

Table 1 :
Baseline characteristics of patients.LVEF, left ventricular ejection fraction.Te P value refers to the statistical diference between the control group and the studied group.

Table 2 :
Logistic multivariable regression analysis of factors of LVH.
BMI6International Journal of Endocrinology