Exploring the association between sleep insufficiency and self-reported cardiovascular disease among northeastern Greeks

Objective To explore the association of sleep characteristics with cardiovascular disease (CVD) using self-reported questionnaires. Material and Methods 957 adults between 19 and 86 years old were enrolled in this cross-sectional study. The participants were classified into three groups [short (<6h), normal (6-8h), and long (>8h) sleepers] by using multistage stratified cluster sampling. CVD was defined by a positive response to the questions: “Have you been told by a doctor that you have had a heart attack or angina or stroke or have you undergone bypass surgery?”. Sleep quality, utilizing Epworth sleepiness scale, Athens insomnia scale, Pittsburgh sleep quality index and Berlin questionnaire, was also examined. Results Prevalence of CVD was 9.5%. Individuals with CVD exhibited reduced sleep duration by 33 min (p<0.001) and sleep efficiency by 10% (p<0.001). In multivariable logistic regression analysis, adjusting for subjects’ sociodemographic, lifestyle habits and health related characteristics, short sleep duration was almost three times more frequent in patients with CVD (aOR=2.86, p<0.001 in the entire sample; aOR=2.68, p=0.019 in women and aOR=2.57, p=0.009 in men). Furthermore, CVD was significantly associated with excessive daytime sleepiness (aOR=2.02, p=0.026), insomnia (aOR=1.93, p=0.010), poor sleep quality (aOR=1.90, p=0.006) and increased risk of obstructive sleep apnea (aOR=2.08, p=0.003). Conclusion Our study highlights a strong correlation of sleep insufficiency with CVD and promotes early pharmacological or cognitive behavioral interventions in order to protect cardiovascular health.


INTRODUCTION
Sleep represents one of the most natural and inseparable life procedures, occupying a third of our everyday time and allowing us to overcome the daily physical and psychological stress 1,2 . The American Academy of Sleep Medicine and the Sleep Research Society recommend a mean period of six to eight hours of sleep per day for preservation of its beneficial health effects 3 . However, the demanding lifestyle of modern society has extended the working schedule favoring sedentary life and stress 4 . As a result, a third of the general population is affected by sleep disturbances, with almost 30% of individuals reporting chronic sleep problems, such as excessive daytime sleepiness and insomnia and consequent deleterious effects on metabolic, immune and endocrine systems 5 .
Cardiovascular disease (CVD) including ischemic heart disease and stroke, represents a major cause of global morbidity and mortality with increasing prevalence 6,7 . Sleep disorders have gradually become an upcoming and modifiable risk factor for CVD, as a growing number of studies points towards a bidirectional relationship of sleep insufficiency with arterial and pulmonary hypertension, diabetes mellitus, coronary artery disease, heart failure, atrial fibrillation, stroke, and overall mortality [8][9][10] . However, there are conflicting evidence emphasizing shorter (≤6h/day) and less frequently longer (≥8h/day) sleep duration, as being associated with a significant risk of CVD 11,12 . Our research group utilizing self-reported questionnaires has recently exhibited that sleep pathology is associated with increased prevalence of anxiety 13 , depression 14 , diabetes mellitus 15 , hypertension 16 and dyslipidemia 17 . In this paper, we aimed to investigate possible correlations between sleep insufficiency and CVD considering several sociodemographic characteristics, lifestyle habits and health related characteristics of the participants.

Study sample and research design
The study population in this cross-sectional study consisted of 957 participants, 439 (45.9%) males and 518 (54.1%) females, with a mean age of 49.62 ± 14.79 years (range, 19-86 years; median age, 50 years). The sample selection was based on a two-stage stratified sampling scheme on all adults living in the region of Thrace and it was conducted between September 2016 and May 2019. Thrace, the Northeastern prefecture of Greece, is characterized by cultural diversity with various national, ethnolinguistic and religious groups. Its population consists of: a.) the indigenous Christian Orthodox population (65% of the region population), b.) the Muslim minority, which is the dominant minority group (30% of the population in Thrace) including the Pomaks and the Roma-Gypsies, and c.) the descendants of Armenian refugees and a lot of expatriated Greeks from countries of the former Soviet Republics who settled in Thrace (estimated 5% of the population in Thrace). In the first stage of the sampling procedure, the area of Thrace was divided in two strata by the degree of urbanization. The urbanization levels were urban (≥10,000 inhabitants) and semiurban or rural (<10,000 inhabitants) areas. According to the 2011 census, which constituted the sampling frame in our study, the urban population of Thrace accounted for approximately 40% of the total population of this area. In the second stage, subjects were recruited proportionally to each stratum size, through a method of random generation of telephone numbers on the basis of the area code. After the aim of the study was explained to them, the participants agreed to have field researchers visit their home and to complete the questionnaires of the study in an hour-long interview. The overall response rate was 71%. The sampling scheme ensured that the sample was randomly selected and representative of the general population of Thrace.

Ethics
Informed consent was obtained from all participants of the study. All procedures performed in the study were in accordance with the ethical standards of the Democritus University Ethics Committee and with the standards of the Helsinki declaration (1964) and its later amendments. The study protocol was approved by the Institutional Ethics Committee (Protocol Number 42570/294).

Covariates
A structured questionnaire was used to collect: a.) standard sociodemographic characteristics (gender, age, place of residence, education level, marital, cultural, employment, and financial status), b.) lifestyle and dietary habits (smoking status, alcohol consumption, daily coffee consumption, caffeine consumption in the evening, adherence to the Mediterranean diet, time watching TV or using a computer before bedtime, physical activity and nap during the day), and c.) health related characteristics (subjective general health status, body mass index, chronic disease morbidity, anxiety, depression, and use of sleep medication) of the participants (Appendix 1).

Estimation of sleep duration and sleep efficiency
Participants provided information on their nighttime sleep by answering the following sleep questions of the questionnaire: "At what time do you normally go to bed?", "At what time do you normally get up?" and "On average, how many hours do you sleep per day?" Responses were obtained for an average weekday and weekend day over the previous month. Time in bed was calculated as the difference between bedtime and rise time. As a proxy of the overall time in bed or sleep duration on a weekly basis, weighted mean measures were calculated using the following formulas: weighted time in bed = 5/7*(time in bed on a weekday) + 2/7*(time in bed on a weekend day) and weighted sleep duration = 5/7*(sleep duration on a weekday) + 2/7*(sleep duration on a weekend day). Sleep efficiency refers to the percentage of time a person sleeps in relation to the amount of time a person spends in bed and was calculated as the ratio of sleep duration and time in bed X 100. Participants were then classified into the following three sleep categories according to calculated sleep duration: short (<6 hours), normal (6-8 hours) and long sleepers (>8 hours) 18 .

Assessment of sleep quality
Sleep quality was assessed with the Greek versions of Epworth Sleepiness Scale (ESS) 19 , Athens insomnia scale (AIS) 20 , Pittsburgh sleep quality index (PSQI) 21 , and Berlin questionnaire (BQ) 22 that evaluate excessive daytime sleepiness, insomnia, sleep quality and risk of obstructive sleep apnea (OSA), respectively. With regards to insomnia characteristics, participants were asked whether they experienced difficulties initiating or maintaining sleep or early morning awakenings.

Definition of CVD
CVD was defined by a positive response to the following questions: "Have you been told by a doctor that you have had a heart attack, angina (chest pain or exertion that is relieved by medication) or have you undergone bypass surgery?" or "Have you been told by a doctor that you have had a stroke?" 23 .

Statistical analysis
Statistical analysis of the data was performed using IBM Statistical Package for the Social Sciences (SPSS), version 19.0 (IBM Corp., Armonk, NY, USA). The normality of quantitative variables was tested with Kolmogorov-Smirnov test. Quantitative variables were expressed as mean ± standard deviation (SD) and qualitative variables were expressed as absolute and relative (%) frequencies. In particular, mean estimated time of sleep characteristics (i.e., bedtime, rise time, time in bed, and sleep duration) were expressed as HH:MM. We conducted the following analyses: (i) in the univariate analysis, the association of cardiovascular diseases with subjects' characteristics, sleep characteristics and sleep disorders were assessed using the chi-square test and Student's t-test; (ii) multivariable stepwise logistic regression analysis was used to explore the independent risk factors for cardiovascular diseases, controlling for all subjects' characteristics; (iii) for the evaluation of the effect of sleep duration and sleep disorders on the prevalence of cardiovascular diseases, two different logistic regression models were constructed: model 1 (crude, unadjusted) and model 2 (adjusted for subjects' sociodemographic, lifestyle habits and health related characteristics). Odds ratios (OR) with their 95% confidence intervals (CI) were estimated as the measure of the above associations. In all the above mentioned multivariable backward stepwise logistic regression models all sociodemographic characteristics (gender, age, marital status, cultural status, place of residence, education level, working status, financial status), lifestyle habits (smoking status, alcohol consumption, daily coffee consumption, caffeine consumption in the evening, adherence to the Mediterranean diet, time watching TV or using a computer before bedtime, physical activity, nap during the day) and health related characteristics (subjective general health status, BMI, chronic disease morbidity, anxiety, depression and use of sleep medication) were initially entered as potential confounders; in the sequence, variables were discarded at a p-value more than 0.20.
Receiver operating characteristic (ROC) analysis was used to provide the ability of sleep duration to classify subjects with cardiovascular diseases. The area under the ROC curve (AUC), sensitivity and specificity were estimated. The optimal cutoff value of the sleep duration that differentiates subjects with cardiovascular diseases from those without cardiovascular diseases was derived according to Youden index 24 . All tests were two tailed and statistical significance was considered for p-values<0.05.

Participants' characteristics
Subjects' sociodemographic, lifestyle and health related features are outlined in Tables 1 and 2. Mean self-reported sleep duration was 6hrs and 19mins on workdays and 6hrs and 45mins on weekends; 31.7% and 22.9% of the participants reported short sleep duration (<6hrs), while 7.9% and 14.2% reported long sleep duration (>8hrs) on workdays and on weekends, respectively. Sleep related medications were regularly used by 6.9% of the participants of our study.  The prevalence of CVD The prevalence of CVDs was 9.5% (91 subjects; 95%CI=7.8% to 11.5%) and its relation to participants' characteristics is shown in Tables 1 and 2. Multivariable logistic regression analysis showed that the strongest risk factor for CVD was age older than 60 years (aOR=23.44, p<0.001) ( Table 3).

CVD and sleep habits
The association of CVD with subjects' sleep characteristics is shown in Table 4. On weekdays, subjects with CVD used to go to bed earlier (p<0.001) and get up earlier (p=0.019) compared to subjects without CVD. Time in bed was longer in subjects with CVD (p=0.004), while they reported 26 min shorter sleep duration (p=0.001) and they had significantly lower sleep efficiency (p<0.001) than those without CVD.
On weekends, although subjects without CVD reported going to bed 29min later (p<0.001), getting up 58min later (p<0.001) and sleeping 28min more (p<0.001) compared to weekdays, the sleep pattern of subjects with CVD remained essentially unchanged (Table 4). In particular, subjects with Notes: aOR = Adjusted odds ratio; CI = Confidence interval; All subjects' sociodemographic characteristics (gender, age, marital status, cultural status, place of residence, education level, working status, financial status), lifestyle habits (smoking status, alcohol consumption, daily coffee consumption, caffeine consumption in the evening, adherence to the Mediterranean diet, time watching TV or using a computer before bedtime, physical activity, nap during the day) and health related characteristics (subjective general health status, BMI, chronic disease morbidity, anxiety, depression and use of sleep medication) were included in the model; All variables were binary (no, yes); Category "no" forms the reference group. CVD reported 52min shorter sleep duration (p<0.001) and lower sleep efficiency (p<0.001) than those without CVD.
In the sequence the weighted weekly values of time in bed and sleep duration were calculated and compared between the two groups (Table 4); it was noted that, although subjects with CVD spent longer time in bed (p=0.047), they reported 33min shorter sleep duration (p<0.001) and lower sleep efficiency (p<0.001) compared to subjects without CVD. All the above relations between CVD and sleep characteristics remained unchanged among men and women (Table 3). In particular, females with CVD used to sleep 43min less than females without CVD (p<0.001) and males with CVD used to sleep 24min less than males without CVD (p<0.001).
Furthermore, according to the reported sleep duration, participants were categorized into three groups: short (<6h), normal (6-8h) and long (>8h) sleep duration. The association of CVD with sleep duration, which was considered as a categorical variable, is shown in Table 5. CVDs were significantly more frequent (p<0.001) in subjects with short (18.7%) compared to those with normal (6.5%) and long (8.8%) sleep duration. The association of CVD with sleep duration exhibited the same pattern in women (p=0.010) and men (p=0.001). In particular, logistic regression analysis revealed that in subjects with short sleep duration there were more than 3-times higher odds for CVD compared to subjects with normal sleep duration (OR=3.28, p<0.001). A 3.07-fold increase in odds of CVD was Sleep Sci. 2022;15(4):388-398 associated with short sleep duration in females (p=0.004) and males (p=0.015), respectively.

Independent effect of sleep habits on CVD
Two separate multivariable logistic regression models, controlling for the effect of all subjects' sociodemographic, lifestyle and health related characteristics, were constructed in order to assess the independent effect of sleep duration on the prevalence of CVD. When sleep duration was entered in the model as a continuous variable, shorter sleep duration remained a statistically significant independent determinant of increased odds for CVD; in particular, shorter sleep duration by one hour was associated with an 29%-increase in the risk for CVD (aOR=1.29, 95%CI=1.05-1.58).
When sleep duration was entered in the multivariable logistic regression model as a categorical variable, the odds for CVD remained higher for the subjects sleeping shorter than 6 hours with adjusted odds ratios of 2.86 (p<0.001) in the entire sample, 2.68 (p=0.019) in women and 2.57 (p=0.009) in men; sleeping longer than 8 hours showed no significant association with CVD ( Figure 1).

CVD and sleep disorders
According to the Greek versions of Epworth sleepiness scale (ESS), Athens insomnia scale (AIS), Pittsburgh sleep quality index (PSQI) and Berlin questionnaire (BQ) the prevalence of daytime sleepiness was 8.7% (83 subjects), insomnia 18.0% (172 subjects), poor sleep quality 38.5% (368 subjects) and high risk of obstructive sleep apnea 36.4% (348 subjects). The internal consistency of all four questionnaires was very high (Cronbach α coefficient ranged from 0.74 to 0.88). The association of CVD with sleep disorders is shown in Table 6. Univariate statistical analysis showed that CVDs were more frequent in subjects with excessive daytime sleepiness (p<0.001), insomnia (p<0.001), poor sleep quality (p<0.001) and higher risk of OSA (p<0.001). In multivariable logistic regression analysis controlling for all subjects' characteristics, the odds of CVD remained higher in subjects with excessive daytime sleepiness (aOR=2.02, p=0.026), insomnia (aOR=1.93, p=0.010), poor sleep quality (aOR=1.90, p=0.006) and higher risk of OSA (aOR=2.08, p=0.003) (Figure 2).
Among the basic difficulties of sleep patterns, significant increased odds of CVDs were found among subjects who reported difficulties in maintaining sleep (aOR=2.36, p=0.008) and early morning awakenings (aOR=1.64, p=0.046), but not with difficulties initiating sleep (aOR=0.77, p=0.303).

DISCUSSION
Our research was designed in order to evaluate the possible associations of sleeping habits and disorders with CVD using a representative population-based sample from the rural region of Thrace, in northeastern Greece. The prevalence of CVD was higher among older men, smokers, with sedentary life, low educational and financial status inhabiting the countryside, along with previous studies 25 . Another interesting finding of our study was the changing pattern of CVD distribution in our sample based upon the marital status, with the widowed patients holding the lion's share, as also concluded in the research of Marzieh et al. (2021) 26 . The overall results revealed a strong correlation of CVD with shorter sleep duration and impaired sleep efficiency, but also high prevalence of excessive daytime sleepiness, insomnia, poor sleep quality and increased risk of obstructive sleep apnea. With regards to insomnia, patients with CVD reported difficulties in maintaining sleep and early morning awakenings, but not difficulties initiating sleep.  In our study, patients with CVD demonstrated significantly shorter sleep duration and lower sleep efficacy compared to individuals without CVD. In particular, mean sleep duration was reduced by 33 min, mean sleep efficiency by 10% and short sleep duration was 3.07-times more frequent in patients with CVD. These results are consistent with the longitudinal study of Covassin et al. (2016) 27 , on a sample of 71,617 participants where CVD presented a 1.39-fold higher prevalence in women reporting ≤5 hours/night compared to those sleeping 8 hours/night. Apart from that, the epidemiologic study of Liu et al. (2013) 28 with the participation of 54,269 adults pointed out that the prevalence of coronary heart disease was higher in the population with sleep duration less than 6 hours/ night. Similarly, according to the meta-analysis by Holliday et al. (2013) 29 , sleep duration of less than 6 hours was significantly associated with increased risk by 30% for type 2 diabetes. Furthermore, Cappuccio et al. (2011) 30 have proven that the incidence of fatal and non-fatal events of coronary heart disease was almost 1.5 times more frequent in the population with sleep duration of less than 7 hours, acknowledging its prognostic role on cardiovascular disease 5,30 . However, He et al. (2017) 31 support that long sleep duration is responsible for the impairment of cardiovascular health through possible prothrombotic pathways, thus favoring the risk of stroke.
We have also demonstrated that sleep duration of less than 5:33 hours could be a potential risk factor for CVD, mainly for females, while most literature emphasizes on sleep duration of less than 6 hours as harmful for the cardiovascular burden 32 . Similar results have been shown in the large national cohort by Shankar et al. (2008) 33 where self-reported sleep less than 5 hours/night by postmenopausal women induced an augmented risk of 25% for coronary heart disease. Kronholm et al. (2011) 34 also concluded that sleep duration of less than 5 hours/night is an independent risk factor for CVD mortality and morbidity in women.
Concerning sleep quality, our results are in accordance with available studies making use of the aforementioned sleep quality scales 35 39 has highlighted the importance of obstructive sleep apnea as a risk factor for cardiac arrhythmias and sudden cardiac arrest. Our study has also noted statistical significance for excessive daytime sleepiness as a predisposing factor for cardiovascular disease, a finding corresponding to the findings of the recent study of Xie et al. (2018) 40 .
Another interesting finding of our study was that long sleep duration was not associated with cardiovascular disease. The pathophysiologic mechanisms underlying the connection between sleep disturbances with CVD although not yet totally understood are based mainly on experimental evidence depicting an interaction between brain and heart 46 . The overstimulation of the sympathetic nervous system is suggested to be a major contributor to cardiovascular disease. In particular, subjects with repeated sleep interruptions exhibited higher nocturnal blood pressure accompanied by dampened nocturnal dipping effect 47 and an enhanced morning rise 48 . A possible explanation resides in the increased cardiac sympathetic drive and cardiovascular over-responsiveness to stress 49 as estimated by heart rate variability measurements 50 . Another important mechanism impairing the cardiovascular system is the hypothalamic-pituitary-adrenal axis (HPAaxis) 46 . Indeed, this theory is based on increased levels of plasma and urinary norepinephrine and cortisol, leading to stress overload 51 . As a result, data from human population's link sleep deprivation with aggravation of arterial stiffness 52 , coronary microcirculation 53 and endothelial function 54 , which may advance atherosclerosis and may cause myocardial damage. Furthermore, the proinflammatory and procoagulant potential of sleep insufficiency is reflected by increased levels of TNF-a, IL-1, IL-6, IL-17, CRP, D-dimers, and fibrinogen 55 . Finally, sleep deficiency affects the metabolic pathways by favoring insulin resistance and weight gain 56 .
Our study overall presents sufficient points of strength which include the following. Firstly, the data of our research derive from a large and representative sample of a regional Greek population, in Thrace. Additionally, the methodology of our sample selection was random, thus reassuring the representation of the general population of this area. The extensive and careful use of diagnostic tools and questionnaires offered acceptable estimates of sleep quality, quantity and cardiovascular disease. The main limitations of our analysis reside in the character of the crosssectional study, the non-investigation of our subjects' medication history and the recall bias of self-reported sleep duration instead of techniques such as polysomnography or actigraphy. The use of self-reported questionnaires may affect CVD prevalence as well as overestimate sleep duration and quality especially in the pattern of a rural population accompanied by a low level of education. Despite this restriction, self-report assessments of sleep have been proven to be reliable measures when compared to quantitative sleep assessments with actigraphy 57 .

CONCLUSION
Our research revealed the increased prevalence of the cardiovascular burden in a regional elder Greek population and the interaction of impaired sleep duration and quality with CVD. Moreover, CVD may induce excessive daytime sleepiness, insomnia, poor sleep quality and increased risk of obstructive sleep apnea. As a result, a balanced sleep duration of 6-8 hours accompanied by a healthy lifestyle is pivotal for the cardiovascular health.

ACKNOWLEDGMENTS
The contributions of all the participants, patient advisers and interviewers are gratefully acknowledged.

SOURCES OF FUNDING
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

CONFLICTS OF INTEREST
The authors report no conflicts of interest.