In this large nationally representative sample of adults with hypertension, we find numerous sociodemographic and health characteristics are significantly associated with habitual sleep duration. Among US adults with hypertension, help-seeking for sleeping difficulty, defined in the present study as a history of telling a doctor or other health professional that one had trouble sleeping, was independently associated with short sleep duration. Previous research has also noted significant associations between self-reported difficulties initiating or maintaining sleep and short sleep duration . These findings have important implications because short sleep duration and other sleep problems, such as insomnia synergistically increase the risk for adverse cardiovascular outcomes [37, 38]. The significant correlation of help-seeking for sleeping difficulty with short sleep duration among those with hypertension provides a point of intervention for health care teams to comprehensively assess and support all aspects of sleep health, including habitual sleep duration, in patients who report sleep problems.
Findings from studies with other populations showed no differences by gender in habitual sleep duration [13, 14]. In contrast, we found that women with hypertension were less likely to be short sleepers and more likely to be long sleepers than men with hypertension. The inconsistency may be partly due to the different approaches used in measuring sleep duration. The two studies defined sleep duration based on the amount of sleep obtained in a day  or 24 hours , increasing the likelihood that napping time was captured. In contrast, our study measured the amount of sleep obtained in the main sleep period. Studies have reported a higher frequency of napping among men [39, 40]. The differences in napping may account for sex differences in sleep duration when the assessment is limited to the amount of sleep obtained during the main sleep period.
The relationship between chronic disease and sleep duration varies across studies. Consistent with other studies [41, 42], we found that CKD was associated with long sleep duration but not short sleep duration. Other previous research findings found that both short (< 6 hours/night) and long sleep duration (> 8 hours/night) were positively associated with incident CKD with the strongest association occurring in those who habitually slept < 4 hours per night . The increased risk of CKD in short sleepers may be due to the increased sympathetic nervous system activity and inflammation that occurs with sleep loss . On the other hand, there is minimal evidence of a pathway through which long sleep duration can cause disease. In those with chronic disease, the long sleep duration may be due to ill health or the underlying pathophysiological mechanisms of the specific condition, such as increased inflammation [1, 44]. The inflammatory processes can cause excessive fatigue and sleepiness [44, 45] and contribute to more time spent in bed.
Similar to our findings in adults with hypertension, many previous studies in the general population have found no association between habitual sleep duration and COPD, heart disease, diabetes [16, 46], stroke [16, 46], or arthritis [13, 16]. However, other studies found that short sleep duration was significantly associated with COPD , heart disease [13, 14], and arthritis , while stroke was associated with long sleep duration . The inconsistencies across studies may be partly related to differences in baseline characteristics of study participants, the varied ways in which sleep duration is measured and analyzed, differences in the proportion of those with comorbidities, and other factors controlled in the studies.
Consistent with prior findings across different adult populations, including older adults and those attending outpatient care settings [16, 36, 48, 49], we found that moderate to severe depressive symptoms were associated with long sleep duration. Other findings have also shown a correlation between sleeping less than 6 hours and risk for depression  which may point to a dose-response relationship between short sleep and depression. Overall, the literature points to a bidirectional relationship between sleep duration and mental health indicators, including depressive symptoms. Circadian rhythm disruptions and insomnia are common in depression and can contribute to short sleep . Sleep loss has also been linked to impairment in emotional regulation and decreased positive and increased negative emotions, which can trigger and worsen depressive symptoms [51, 52]. Depression has also been linked to elevated levels of pro-inflammatory markers [53, 54], whose effects include excessive fatigue and sleepiness [45, 54, 55]. Hypertension is also associated with increased inflammation [56, 57]. Therefore, having both hypertension and depression may worsen the effects of inflammation in the body.
The relationship between sleep duration and other health-related behaviors has been examined in many studies. Consistent with prior studies, we found no significant relationship between alcohol intake and habitual sleep duration [13, 16]. Other previous findings have shown heavy alcohol intake to be associated with sleep continuity disturbances but not sleep duration . Alcohol intake near bedtime can affect sleep quality because it disrupts the normal sleep architecture . Consistent with previous studies, we found no association between cigarette smoking and habitual sleep duration [13, 16].
Research findings on the relationship between physical activity and sleep duration are mixed. Similar to a prior study in older adults , we noted no associations between physical activity and sleep duration in adults with hypertension. Other findings suggest that being physically active is associated with reduced odds of having a long sleep duration [14, 60]. The negative association between physical activity and long sleep duration may be related to other confounding factors, such as health conditions, not controlled for in studies. Findings from experimental studies suggest that physical activity substantially improves other sleep parameters such as sleep regularity and sleep quality but only has modest effects on total sleep time [61, 62]. On the other hand, adequate sleep can enhance the ability to remain physically active because it improves mental and physical performance .
We noted that individuals working ≥ 45 hours per week were more likely to be short sleepers, while those working < 35 hours per week or not working were more likely to be long sleepers when compared to individuals working 35–44 hours per week. Previous studies also have found significant associations between longer work hours and short sleep duration [64, 65]. Findings from the American Time Use Survey showed that the main activity exchanged for sleep is paid work , and other results show that a reduction in work hours leads to a significant increase in sleep duration . In the present study, those in retirement also had a slight but significant increase in odds of being a short sleeper than those working 35–44 hours/week. These findings on retirement and sleep duration may be partly explained by how sleep duration was quantified in our study. The present study only considered sleep obtained during the main sleep period. Studies have found that individuals who have transitioned from work to retirement tend to nap more frequently or for longer durations when compared to those who are employed [67, 68].
Our results on the relationship between race/ethnicity and habitual sleep duration among adults with hypertension are partly consistent with previous findings in the general US adult population. Consistent with other studies [69–71], we found that non-Hispanic Black adults were more likely than non-Hispanic White adults to be short sleepers. However, prior studies also reported significantly higher odds for short sleep duration among Hispanic  and non-Hispanic Asian adults than non-Hispanic White adults [70, 71]. In contrast, our study showed that the odds of being a short or long sleeper in Hispanic or non-Hispanic Asian adults were not significantly different from that of non-Hispanic White adults. Of note, in these studies, the strongest association between race/ethnicity and short sleep duration has been observed in non-Hispanic Black adults [70, 71].
Previous studies have also found objectively measured total sleep time of non-Hispanic Black US adults to be significantly shorter than that of non-Hispanic White, Hispanic, and non-Hispanic Asian US adults [69, 71]. These findings on non-Hispanic Black race/ethnicity and short sleep duration are noteworthy because hypertension is more prevalent among non-Hispanic Black adults than in non-Hispanic White, Hispanic, or non-Hispanic Asian adults . Previous findings suggest that the racial/ethnic differences in sleep duration may be related to differences in stress and perceived racial discrimination [15, 72]. Further research is needed to explore further the mechanisms underlying these racial/ethnic differences in sleep duration.
Despite this study being nationally representative, there are limitations to be noted. The study was cross-sectional, and consequently, we cannot infer any temporality of the relationships between predictors such as chronic health conditions and habitual sleep duration. Second, NHANES did not have data on other potential predictors of sleep duration, such as inflammatory biomarkers, stress, sleep attitudes, napping, and caregiving [15, 16, 39, 73, 74]. Third, data for many variables used in this study, including habitual sleep duration, chronic health conditions, and health behaviors, were measured using self-reports. Even though the collection of self-reported data is more feasible than objective measurements in large-scale surveys such as NHANES, it increases the risk for measurement error related to factors such as poor recall and social desirability . To reduce measurement error, part of the NHANES protocols involved clarifying with participants any extreme high or low self-reports on various questionnaire items. For example, for sleep duration, NHANES interviewers verified with participants sleep duration ≤ 4 or ≥ 10 hours.
Despite the limitations noted, the information gathered on factors associated with short and long sleep duration in adults with hypertension provides important contributions to the literature on predictors of poor sleep outcomes in this population. Future studies in hypertensive adults should consider using objective measures of sleep duration and other health behaviors, such as physical activity, and include other important predictors of sleep duration in their analyses.