Association between sleep–wake habits and use of health care services of middle-aged and elderly adults in China

Introduction

Aphorisms such as “Early to bed and early to rise makes a man healthy, wealthy, and wise” and “An apple a day keeps the doctor away,” are considered the crystallization of practical experience and wisdom of broad masses of people. Some aphorisms, such as “Early to bed and early to rise…” have come to symbolize health and healthy habits, serving to guide people’s lifestyles for centuries—or even thousands of years—around the world [1–5]. To this day, almost all school education programs throughout the world emphasize the importance of “early to bed and early to rise” to students, especially for children [6–8].

Indeed, previous studies on adolescents have shown that late bedtimes and late wake-up times are associated with poorer diet quality, unfavorable activities, and weight status profile, and these effects are independent of sleep duration, age, sex, household income, and geographical remoteness [9, 10]. However, studies of adult patients and elderly people have shown that being “early to bed and early to rise” is not associated with mortality, income, or educational attainment [11, 12]. With regards to the older adults, it is necessary to examine whether being “early to bed and early to rise” reduces physician visits, translating into lower health care use—a goal sought by policy makers, individuals, families, and societies. To our knowledge, the association between sleep–wake habits and the use of health care services has never been rigorously examined.

Although some proverbs are not fully supported by empirical evidence, we cannot deny all of the practical experience and wisdom of our predecessors [1]. After all, the theory of modern medicine is derived from the practice of our ancestors and builds on their work, therefore serving as a link between the past and the future. Therefore, in-depth research is necessary to eliminate the false and retain the true, separating out what is useful and essential among the aphorisms. In China, an aphorism closely related to “Early to bed and early to rise…” is “Do not seek an immortality prescription; look for a sleep prescription.” If these aphorisms hold true, sleep prescription could potentially improve health outcomes and thus reduce individual and national health care spending.

We used data from the Dongfeng-Tongji (DFTJ) cohort study to examine the association of sleep–wake habits with the use of health care services (physician visits and hospitalization). We aimed to provide more in-depth, precise evidence to test whether “Early to bed and early to rise…” are associated with better health of middle-aged and older adults.

Results

The maximum age of the participants was 94 years, the youngest age was 37 years, and the average age was 64.7 ± 8.6 years. Table 1 presents the characteristics of participants by sleep–wake habits. Those who were “early to bed” (including 43.0% Early-bed/Early-rise (EE) and 5.7% Early-bed/Late-rise (EL)) and “late to bed” (including 39.5% Late-bed/Early-rise (LE) and 11.8% Late-bed/Late-rise (LL)) each made up roughly half of the participants. The demographic characteristics and lifestyles of participants in the other three groups were significantly different from the EE group, except for sex and drinking status for the EL group (P =0.481 and 0.109, respectively). The average age of the participants by sleep–wake habits was 66.5 ± 8.2 years for EE, 65.2 ± 10.1 years for EL, 63.5 ± 8.1 years for LE, and 61.5 ± 8.9 years for LL. Compared with EE participants, the “late to bed” (including LE and LL) group included more women, highly educated people, and passive smokers, and fewer people with self-reported chronic diseases, former drinkers, and former smokers (all P<0.001). Exercise habits were more frequently observed among participants who were “early to rise” (88.6% for EE and 90.0% for LE) than among those who were “late to rise” (76.2% for EL and 84.5% for LL) (all P<0.001).

Table 2 presents the use of health care services by sleep–wake habits. In total, 60.8% of the participants reported to visit doctors more than once, and the corresponding proportions of the four groups were 59.9%, 61.7%, 61.3%, and 62.1%, respectively (P = 0.020). There were 36.4% of the participants have been hospitalized in the past year, and the hospitalization rates for the four groups were 38.4%, 40.9%, 34.3%, and 33.4%, respectively (P < 0.001). In the full sample, compared with EE participants, LL participants were more likely to report physician visits (P = 0.011); the results of the gender- and age-stratified analyses were highly similar (except for women, with P = 0.060). In the full sample, LL participants were less likely than EE participants to report hospitalization (P < 0.001); the results of the gender-stratified analysis were similar, but the results of the age-stratified analysis differed. For participants who were younger than 60 years, there were no significant differences among the four groups. For participants aged 60–69 years, differences in hospitalization by sleep–wake habits were similar to those in the total sample. For participants aged ≥ 70 years, the difference in hospitalization between EE and LL was not statistically significant (P = 0.103).

To explore the associations between sleep-wake habits with the number of physician visits and hospitalization, we conducted multivariable logistic regression analyses by adjusting potential confounders (i.e., age, sex, educational attainment, marital status, number of chronic diseases, smoking, passive smoking, drinking, and exercise) (Figure 1). In the unadjusted analyses, EL, LE, LL participants were 1.08, 1.06, 1.09 times more likely to have a higher number of physician visits than EE participants though the effect for the EL group was not statistically significant (Supplementary Table 1). The odd ratio of hospitalization for EL participants was 11% (95%CI: 1.01–1.22) higher than the EE group. However, we found no such association for the LE and LL groups (LE: OR = 0.84, 95% CI :0.80–0.88; LL: OR = 0.81, 95% CI: 0.75–0.87). After adjusted for all the covariates, LE and LL participants were more likely to report higher number of physician visits than EE participants (OR = 1.13, 95% CI: 1.08–1.19 and OR = 1.27, 95% CI: 1.18–1.38, respectively); the results of the gender- and age-stratified analyses were highly similar (Figure 1). Compared with EE participants, hospitalization did not differ significantly for the other three categories of sleep–wake habits (EL: OR = 1.11, 95% CI: 1.00–1.22; LE: OR = 0.96, 95% CI :0.91–1.01; LL: OR = 1.01, 95% CI: 0.94–1.10); the results of the gender- and age-stratified analyses were highly similar.

Discussion

To our knowledge, this investigation was the first large-scale empirical study to examine the association between sleep–wake habits and physician visits. The results showed that, compared with EE, those who were late to bed, whether LL or LE, reported higher numbers of physician visits but not more frequent report hospitalization.

The lower number of physician visits by EE participants may result from the health effects of circadian rhythms, endogenous rhythms with a periodicity of approximately 24 hours [13–15]. Recent evidence strongly points to the pervasive influence of circadian timing on nearly all physiologic functions, including brain function, hormonal changes, mood disturbances, the endocrine system, and the metabolic process [16–19]. EE is considered the ideal sleep timing, in line with physiological circadian rhythms [20–22]. However, this ideal sleep timing has encountered serious challenges in the past few decades. With the rapid development of the Internet, increasing numbers of people stay up late taking part in activities such as chatting online, playing games, and watching videos [23, 24]. By 2013 when this present study was conducted, 71% of the Chinese population owns a mobile phone and social media has become embedded in ordinary people’s life [25]. Social media platforms such as WeChat has become a daily communication tool for most urban Chinese people (93%) including older adults [26]. Although the use of information and communicative technology (ICT) was found to enhance the psychological well-being of older adults, excessive use of ICT and social media can challenge the ideal sleep timing [27].

The difference in the impact of EE on physician visits and hospitalization reveals the applicable conditions of the proverb in health care services, similar to the indications for a medication or surgical procedure. The vast majority of aphorisms, including “Early to bed and early to rise makes a man healthy,” were generated in the context of crude medical practice. These experiences were not derived from dealing with serious illnesses or advanced stages of disease in clinical practice, but rather from the practice of dealing with minor illnesses or the early stages of disease or maintaining a condition of well-being. Therefore, the indication of “Early to bed and early to rise makes a man healthy” would be for pre-clinical practice with the minor illnesses or in the early stages of disease, to make the aphorism more precise for health care services.

Although the effect of EE is weak, it may nip the evil in the bud, and there may also be potential to reduce individual and national health care spending. If the observed association between sleep–wake habits and reduced physician visits is causal, the promotion of EE could, at least in theory, help to contain the costs associated with physician visits. In a rough and conservative calculation, based on Chinese data for the approximate average health care spending on physician visits for adults of $23 and annual physician visits in 2016 of 7.7 billion [28], we estimate that the savings in health care spending associated with a rise in EE would be $17 billion per year, with a 30% reduction in physician visits for LL and a 15% reduction in physician visits for LE. Given the approximate average travel expenses ($2) and the cost of lost working time ($15/day) for patients and their accompanying family members, the savings in indirect spending would be $136.14 billion per year. The total cost savings is about $153.14 billion per year.

Several potential limitations of the present study should be noted. First, our study sample consisted of middle-aged and elderly people, which is not representative of the general population. However, the consistency of the results from the age-stratified analysis reveals that the effects of sleep–wake habits on physician visits may not vary by age. Second, although we account in our analyses for a variety of health-related characteristics, we cannot dismiss the possibility of residual confounding. EE participants were, in fact, measurably different from others and would be expected to differ on other, unmeasured factors. Specifically, EE individuals may be more health-conscious and hold more industrious values, which could entirely explain the associations we observed. Third, because of the cross-sectional study design, we are unable to assess cause and effect in the observed relationships between EE and the successful avoidance of physician visits. However, we anticipate that causality is likely, given the universality of circadian rhythms for both men and women and in the past, present, and future [14, 29, 30].

The practice of medicine has developed greatly in the 21^st century. However, the health problems that human beings encounter are still complex and severe, and societies remain overburdened with the disease. We believe that it is very necessary to examine what we can learn from the past. Evidence from the present study shows that compared with EE, LL would increase about 30% physician visits. Additionally, our investigation allows us to carry forward the wisdom of our predecessors involved in the aphorism through revealing that the indication of “Early to bed and early to rise makes a man healthy” may be mainly in pre-clinical practice. This makes the aphorism more precise for health care services. Faced with an increasing number of young people staying up late and increasing health care spending in modern society, follow-up research on these topics would be of particular value to young people as the workforce of today and the patients of tomorrow.

Conclusion

This study found that middle-aged and elderly people who stayed up late and got up late are more likely to visit the doctors than those who went to bed early and got up early. However, it is also possible that early to bed and early to rise individuals may be more health-conscious, which could explain the associations observed. Nevertheless, this is the first large-scale empirical study that provided rigorous evidence to support the association between sleep–wake habits and the use of health care services for middle-aged and older adults in China. Limited to the cross-sectional design, we cannot interpret such association as causality. Moreover, we cannot eliminate bias due to unmeasured confounders. Further prospective cohort studies are needed to validate our findings through the rigorous design and comprehensive consideration of potential confounders.

Materials and Methods

Supplementary Materials

Supplementary Table 1