This study was presented at the American Aging Society meeting, Boulder, Colorado, June 1, 2008.
Circadian Activity Rhythms and Mortality: The Study of Osteoporotic Fractures
Article first published online: 26 JAN 2010
© 2010, Copyright the Authors. Journal compilation © 2010, The American Geriatrics Society
Journal of the American Geriatrics Society
Volume 58, Issue 2, pages 282–291, February 2010
How to Cite
Tranah, G. J., Blackwell, T., Ancoli-Israel, S., Paudel, M. L., Ensrud, K. E., Cauley, J. A., Redline, S., Hillier, T. A., Cummings, S. R., Stone, K. L. and for the Study of Osteoporotic Fractures Research Group (2010), Circadian Activity Rhythms and Mortality: The Study of Osteoporotic Fractures. Journal of the American Geriatrics Society, 58: 282–291. doi: 10.1111/j.1532-5415.2009.02674.x
- Issue published online: 27 JAN 2010
- Article first published online: 26 JAN 2010
- circadian rhythm;
OBJECTIVES: To determine whether circadian activity rhythms are associated with mortality in community-dwelling older women.
DESIGN: Prospective study of mortality.
SETTING: A cohort study of health and aging.
PARTICIPANTS: Three thousand twenty-seven community-dwelling women from the Study of Osteoporotic Fractures cohort (mean age 84).
MEASUREMENTS: Activity data were collected using wrist actigraphy for a minimum of three 24-hour periods, and circadian activity rhythms were computed. Parameters of interest included height of activity peak (amplitude), midline estimating statistic of rhythm (mesor), strength of activity rhythm (robustness), and time of peak activity (acrophase). Vital status, with cause of death adjudicated through death certificates, was prospectively ascertained.
RESULTS: Over an average of 4.1 years of follow-up, there were 444 (14.7%) deaths. There was an inverse association between peak activity height and all-cause mortality rates, with higher mortality rates observed in the lowest activity quartile (hazard ratio (HR)=2.18, 95% confidence interval (CI)=1.63–2.92) than in the highest quartile after adjusting for age, clinic site, race, body mass index, cognitive function, exercise, instrumental activity of daily living impairments, depression, medications, alcohol, smoking, self-reported health status, married status, and comorbidities. A greater risk of mortality from all causes was observed for those in the lowest quartiles of mesor (HR=1.71, 95% CI=1.29–2.27) and rhythm robustness (HR=1.97, 95% CI=1.50–2.60) than for those in the highest quartiles. Greater mortality from cancer (HR=2.09, 95% CI=1.04–4.22) and stroke (HR=2.64, 95% CI=1.11–6.30) was observed for later peak activity (after 4:33 p.m.; >1.5 SD from mean) than for the mean peak range (2:50–4:33 p.m.).
CONCLUSION: Older women with weak circadian activity rhythms have higher mortality risk. If confirmed in other cohorts, studies will be needed to test whether interventions (e.g., physical activity, bright light exposure) that regulate circadian activity rhythms will improve health outcomes in older adults.
Many biological functions are under circadian control, including release of certain hormones, temperature, blood pressure, heart rate, bone remodeling, sleep, and activity cycles. With age, circadian activity rhythms phase advance, resulting in an earlier onset of sleepiness in the evening and an earlier morning waking time.1 Some older adults also show a decrease in rhythm amplitude (peak activity),2 shorter circadian periods of less than 24 hours, and loss of robustness in the rhythm.1,3–7 Little is known about the causes of age-related changes in circadian patterns and the subsequent effects of these changes on health and well-being. A disrupted or less-robust circadian activity rhythm has been associated with medical illness, such as dementia and cancer. Disturbances of the sleep–wake cycle, which are reflected in poor activity rhythms, are particularly pronounced in Alzheimer's disease8 and are hypothesized to be one of the primary causes of institutionalization.9,10
Exposures that influence circadian activity rhythms also may contribute to disease. For example, shift work11 and chronic jet lag12 have been shown to reduce mental acuity and increase the risk of a number of medical problems. Night work has been linked to specific pathological disorders, including higher risks of breast cancer, cardiovascular disease, gastrointestinal disease, diabetes mellitus, and metabolic impairment.13,14 Several epidemiological studies have reported greater risk of mortality in people with self-reported short or long sleep duration15–19 and that sleep disturbances may also increase the risk for a variety of diseases and conditions,20–24 such as diabetes mellitus25 and cardiovascular disease.26
Although the association between circadian activity rhythms and illness is fairly strong, evidence for an association between disrupted activity rhythms and mortality is limited.27–29 Two-year survival in patients with metastatic colorectal cancer was five times as higher in those with stronger circadian activity rhythms as in those with rhythm abnormalities.29 Furthermore, those with more daytime than nighttime activity had better quality of life.28,29 Activity phase abnormalities in older adults with dementia have been shown to predict shorter survival.27 It is not clear whether activity rhythms directly influence mortality or represent biomarkers of advanced physiological aging that provide additional risk beyond that of traditional covariates.
The relationship between circadian activity rhythms and risk of mortality in community-dwelling elderly populations has not been studied. This study examined data gathered in the Study of Osteoporotic Fractures (SOF), a longitudinal study designed to examine the risk factors of osteoporotic fractures in women, to test the hypothesis that circadian activity rhythms measured objectively using actigraphy are prognostic indicators of mortality in a large sample of community-dwelling older women.
The SOF is a longitudinal epidemiological study of 10,366 women aged 65 and older recruited from four study centers located in Baltimore, Maryland; Minneapolis, Minnesota; Portland, Oregon; and the Monongahela Valley, near Pittsburgh, Pennsylvania. Women were excluded if they had had a bilateral hip replacement or were unable to walk without assistance. The baseline SOF examinations were conducted from 1986 to 1988, when 9,704 Caucasian women were recruited.30 The SOF was originally designed to investigate risk factors for osteoporosis and osteoporotic fractures, and African-American women were initially excluded from the study because of their low incidence of hip fractures, but from February 1997 to February 1998, 662 African-American women were enrolled.31 At all subsequent visits, no exclusion criteria were used. All participants were community dwelling at baseline. Since then, follow-up examinations have taken place approximately every 2 years.
The focus of this analysis was data gathered at SOF Examination 8, which took place between January 2002 and February 2004. Of the 4,727 women at this visit, 3,676 (77.8%) participants with clinic or home visits were eligible for collection of wrist actigraphy data. Eligibility was based on having a home visit, participant willingness, and overall ability to perform the study. Of the 3,676 participants with a home or clinic visit, 12.4% refused or had advanced frailty or cognitive problems and were deemed ineligible by the study staff. The success rate for those given actigraphs was 94.0%. The patient flow diagram gives more information on the reasons for missing actigraphy data (Figure 1). Of those who were eligible to participate in the actigraphy study, 0.8% had an actigraph malfunction, 0.4% had a software or initialization problem, 1.3% removed the actigraph and did not replace it, 2.4% did not have adequate proportional integration mode data (which computes movement as counts per minute based on an area under the receiver operating characteristic curve analysis that takes into account intensity and frequency of movement), and 0.8% did not have full 24-hour proportional integration mode data to calculate the rest–activity parameters. The institutional review boards on human research approved the study at each institution, and all participating women provided written informed consent.
Activity data were collected using the Sleep-Watch-O (Ambulatory Monitoring, Inc., Ardsley, NY), a small device worn on the wrist. A piezoelectric linear accelerometer (sensitive to ≥0.003 g), which generates a voltage each time the actigraph is moved, measures movement. These voltages are gathered continuously and summarized over 1-minute epochs. The actigraph was initialized in the clinic before the visit, and the examiner placed it on the participant's nondominant wrist during the visit. Women wore the actigraphs continuously for a minimum of three 24-hour periods (i.e., 72 hours).
An extension to the traditional cosine model was used to map the circadian activity rhythm to the activity data.32 This extended cosine model applies a nonlinear transformation to the cosine curve, the antilogistic function, and is sometimes referred to as a five-parameter extension of the 24-hour cosine curve. Activity data often assume a shape more similar to a squared wave than a cosine curve, and this extension to the traditional cosine curve fits each individual's data and allows for this shape. Circadian activity parameters of the extended cosine model were calculated using nonlinear least squares. The following activity rhythm parameters were calculated from the extended cosine curve: amplitude, an indicator of the strength of the rhythm, the peak to nadir difference in activity (measured in arbitrary units of activity (counts/min)); midline estimating statistic of rhythm (mesor), mean level of activity (measured in arbitrary units of activity (counts/min)); robustness of the circadian activity rhythm (pseudo-F statistic for goodness of extended cosine fit; higher pseudo-F values indicate stronger rhythms); and acrophase, timing of peak activity measured in portions of hours (time of day). Circadian amplitude, midline mesor, and robustness were examined based on quartile distributions. Acrophase was examined in terms of deviation from the population mean. Three categories were identified based on having a peak time of more than 1.5 standard deviations (SDs) above and below the population mean for the study population. Phase-advanced participants were defined as having an acrophase of earlier than 12:50 p.m. (−1.5 SD from the mean), and phase-delayed participants were defined as having an acrophase later than 4:33 p.m. (+1.5 SD from the mean).
Vital status after Visit 8, with cause of death verified through death certificates, was ascertained during an average of 4.1 ± 1.1 years of follow-up. Participants were contacted by postcard or telephone every 4 months to ascertain vital status. Information from designated proxy sources (e.g., family member or a close friend) was used if the participant has died. Follow-up since the baseline visit has remained greater than 95% complete. Deaths were confirmed according to death certificates. Causes of death were confirmed according to death certificates and, when available, hospital discharge summaries. The median follow-up period was 4.3 years (range 5 days to 5.2 years). International Classification of Diseases, Ninth Revision, codes were used to classify causes of death as coronary heart disease (codes 410–414), stroke (codes 430–438), atherosclerosis (codes 401–404, 410–414, 425, 427.5, 428, 429.2, 430–438, 440–444, and 798), cancer (codes 140–239), or all other causes (noncancer and nonatherosclerotic deaths). There were insufficient numbers of specific other causes of death to analyze these as separate outcomes. The most common causes of death in the “other” category were pulmonary (n=38; codes 415–417.9, 460–529.9, 786, 796, 799.1) and cognitive related (n=17; codes 290–290.9, 331–331.9, 332–332.1).
All participants completed questionnaires that included questions about medical history, self-reported health, smoking status, alcohol use, caffeine intake, marital status, and whether the participant walked for exercise. The Geriatric Depression Scale (GDS) was used to assess depressive symptoms, with the standard cutoff of six or more symptoms used to define depression.33 Medication use was ascertained by asking participants to bring all current prescription and nonprescription medications used in the past 30 days to their clinic visits. For women who completed a home visit, the interviewer gathered medication use information at the home. A computerized medication coding dictionary was used to categorize all medications.34 The Mini-Mental State Examination (MMSE) was administered to assess cognitive function, with higher scores on a scale of 0 to 30 representing better cognition.35 Functional status was assessed by collecting information on six instrumental activities of daily living (IADLs), which included walking two to three blocks on level ground, climbing up 10 steps, walking down 10 steps, preparing meals, doing heavy housework, and shopping for groceries or clothing. Participants were asked whether pain made sleeping difficult. Body weight and height were measured, and body mass index (BMI) was calculated as weight in kilograms divided by the square of height in meters. A history of cardiovascular disease was defined as a prior diagnosis of myocardial infarction, angina pectoris, congestive heart failure, or other heart disease. A history of medical conditions was defined as a prior diagnosis of stroke, diabetes mellitus, Parkinson's disease, Alzheimer's disease, chronic obstructive pulmonary disease, cancer, or cardiovascular disease. All measurements were collected at Visit 8.
Characteristics known to be related to activity rhythms or mortality36 were summarized using means and SDs for continuous data and percentages for categorical data. Characteristics were compared between categories of amplitude and acrophase using analysis of variance for continuous covariates that were normally distributed, Kruskal-Wallis tests for skewed continuous data, and chi-square tests for categorical data. To determine the relationship between circadian activity rhythms and mortality, Cox proportional hazards models were used to estimate hazard ratios (HRs) and 95% confidence intervals [CIs]. To identify potential confounders, a list of predictors thought to be associated with circadian activity rhythms and mortality, based on biological plausibility or previous studies, were considered, including age, BMI, cognitive function, self-reported health, number of IADL impairments, physical activity, depression, pain, race, benzodiazepine and antidepressant use, sleep medication use, alcohol use, smoking status, caffeine intake, marital status, and prior medical conditions. Variables that were significantly related (P<.10) to at least one activity rhythm predictor measure and all-cause mortality were included in the final multivariate analyses (age, clinic site, race, BMI, cognitive function, walking for exercise, IADL impairments, depression, current use of benzodiazepines or antidepressants, alcohol use, smoking status, self-reported health status, married status, and comorbidities). Multivariate-adjusted Kaplan-Meier curves were generated to assess the cumulative incidence of all-cause mortality. Additional analyses were performed to determine whether any association found between peak and mean activity levels and all-cause mortality were independent of physical activity level. Models were stratified according to whether walking for exercise was reported, and formal interactions between walking for exercise and mesor and amplitude were performed. Statistical analysis was performed using SAS version 9.1 (SAS Institute, Inc., Cary, NC).
Characteristics of the Study Population
The analysis cohort consisted of 3,027 women (mean age 84 ± 4, range 77–99). Characteristics of this analytical data set of women with actigraphy data had previously been compared with those of the remaining women who did not have actigraphy measured.21 The group without actigraphy was on average 1.4 years older and slightly less likely to be African American and had a slightly higher prevalence of several health conditions than women with actigraphy measurements and experienced almost twice the mortality rate. The 192 women who were given the actigraph but had unusable activity rhythm data were compared with those in the analysis subset. Those with unusable data had slightly lower MMSE scores (27.3 vs 27.9, P<.001) and shorter follow-up times (3.8 vs 4.0 years, P=.009).
Characteristics were compared according to quartiles of amplitude (Table 1). In general, women in the lowest quartile of amplitude were more likely to be older; have a higher BMI, more IADL impairments, more medical conditions, poorer health; and be more likely to be taking antidepressants and to smoke. Characteristics of the lower mesor and less-robust groups (Appendices SA1 and SB1; available on online version or communication to authors) were similar to those for amplitude. Characteristics of the delayed acrophase group were largely similar to characteristics of those in the lower quartile of amplitude, with younger age being the major exception (Table 2).
|Characteristic||Amplitude (Counts/Minute)||P -Value*|
|Overall (N=3,027)||<2,743 (n=756)||2,743–3,412 (n=757)||3,413–4,045 (n=757)||≥4,046 (n=757)|
|Age, mean ± SD||83.6 ± 3.8||84.6 ± 4.1||83.9 ± 3.8||83.1 ± 3.5||82.6 ± 3.4||<.001|
|Body mass index, kg/m2, mean ± SD||27.0 ± 5.0||28.5 ± 5.7||27.2 ± 4.9||26.5 ± 4.8||26.2 ± 4.3||<.001|
|African American, n (%)||323 (10.7)||81 (10.7)||64 (8.5)||79 (10.4)||99 (13.1)||.04|
|Any instrumental activity of daily living impairments, n (%)||1,594 (53.0)||554 (74.0)||436 (57.8)||333 (44.2)||271 (36.0)||<.001|
|History of any medical condition, n (%)||1,876 (62.0)||543 (72.0)||502 (66.4)||446 (59.0)||385 (50.9)||<.001|
|History of stroke, n (%)||398 (13.2)||145 (19.2)||106 (14.0)||83 (11.0)||64 (8.5)||<.001|
|History of diabetes mellitus, n (%)||335 (11.1)||126 (16.7)||94 (12.4)||55 (7.3)||60 (7.9)||<.001|
|History of cardiovascular disease, n (%)||1,000 (33.1)||300 (39.8)||275 (36.4)||234 (30.9)||191 (25.3)||<.001|
|History of cancer, n (%)||668 (22.1)||155 (20.6)||184 (24.3)||175 (23.2)||154 (20.4)||.17|
|Currently taking benzodiazepine, n (%)||219 (7.2)||61 (8.1)||64 (8.5)||56 (7.4)||38 (5.0)||.05|
|Currently taking antidepressant, n (%)||413 (13.7)||152 (20.2)||103 (13.6)||76 (10.1)||82 (10.8)||<.001|
|Current sleep medication user, n (%)||33 (1.1)||6 (0.8)||6 (0.8)||7 (0.9)||14 (1.9)||.14|
|Geriatric Depression Scale score≥6, n (%)||355 (11.8)||146 (19.4)||93 (12.3)||66 (8.7)||50 (6.6)||<.001|
|Mini-Mental State Examination score, mean ± SD||27.9 ± 2.0||27.5 ± 2.3||27.8 ± 1.9||27.9 ± 1.9||28.1 ± 1.8||<.001|
|Current smoker, n (%)||84 (2.8)||31 (4.1)||26 (3.4)||13 (1.7)||14 (1.9)||<.01|
|Average drinks per day in past 30 days, mean ± SD||0.5 ± 0.7||0.4 ± 0.7||0.46 ± 0.6||0.54 ± 0.7||0.6 ± 0.7||<.001|
|Average caffeine intake (mg/d), mean ± SD||151 ± 154||133 ± 137||135 ± 154||156 ± 150||180 ± 170||<.001|
|Walks for exercise, n (%)||1,114 (37.3)||189 (25.4)||275 (36.6)||315 (42.5)||335 (44.7)||<.001|
|Married, n (%)||794 (26.3)||130 (17.2)||196 (25.9)||223 (29.5)||245 (32.4)||<.001|
|Trouble sleeping in past month because of pain, times/week, n (%)|
|0||2,018 (66.7)||472 (62.6)||506 (66.9)||501 (66.2)||539 (71.3)||.02|
|<1||287 (9.5)||75 (9.9)||63 (8.3)||74 (9.8)||75 (9.9)|
|1–2||333 (11.0)||100 (13.3)||85 (11.2)||89 (11.8)||59 (7.8)|
|≥3||385 (12.7)||107 (14.2)||102 (13.5)||93 (12.3)||83 (11.0)|
|Self-reported health status, n (%)||<.001|
|Poor to very poor||66 (2.2)||40 (5.3)||14 (1.9)||5 (0.7)||7 (0.9)|
|Fair||680 (22.5)||241 (32.0)||179 (23.7)||150 (19.8)||110 (14.6)|
|Good to excellent||2,277 (75.3)||473 (62.7)||563 (74.5)||602 (79.5)||639 (84.5)|
|Characteristic||Acrophase (Time of Day)||P-Value*|
|Overall (N=3,027)||<12:50 p.m. (N=176)||12:50–4:33 p.m. (n=2,682)||>4:33 p.m. (n=169)|
|Age, mean ± SD||83.56 ± 3.8||83.73 ± 4.0||83.58 ± 3.8||83.09 ± 4.1||.23|
|Body mass index, kg/m2, mean ± SD||27.04 ± 5.0||26.72 ± 5.5||26.97 ± 4.9||28.46 ± 6.0||<.001|
|African American, n (%)||323 (10.7)||16 (9.1)||273 (10.2)||34 (20.1)||<.001|
|Any instrumental activity of daily living impairments, n (%)||1,594 (53.0)||85 (49.1)||1392 (52.1)||117 (70.1)||<.001|
|History of any medical condition, n (%)||1,876 (62.1)||104 (59.1)||1654 (61.8)||118 (70.2)||.06|
|History of stroke, n (%)||398 (13.2)||22 (12.5)||350 (13.1)||26 (15.5)||.65|
|History of diabetes mellitus, n (%)||335 (11.1)||12 (6.8)||297 (11.1)||26 (15.5)||.04|
|History of cardiovascular disease, n (%)||1,000 (33.1)||53 (30.1)||880 (32.9)||67 (39.9)||.12|
|History of cancer, n (%)||668 (22.1)||34 (19.3)||603 (22.5)||31 (18.6)||.32|
|Currently taking benzodiazepine, n (%)||219 (7.2)||8 (4.6)||193 (7.2)||18 (10.7)||.09|
|Currently taking antidepressants, n (%)||413 (13.7)||10 (5.7)||364 (13.6)||39 (23.2)||<.001|
|Current sleep medication user, n (%)||33 (1.1)||33 (1.2)||.12|
|Geriatric Depression Scale score≥6, n (%)||355 (11.8)||27 (15.4)||304 (11.4)||24 (14.3)||.16|
|Mini-Mental State Examination score, mean ± SD||27.86 ± 2.0||27.83 ± 1.8||27.87 ± 2.0||27.68 ± 2.3||.66|
|Current smoker, n (%)||84 (2.8)||3 (1.7)||72 (2.7)||9 (5.4)||.08|
|Average drinks per day in past 30 days, mean ± SD||0.5 ± 0.71||0.45 ± 0.69||0.51 ± 0.71||0.49 ± 0.77||.39|
|Average caffeine intake (mg/d), mean ± SD||151 ± 154||133 ± 137||152 ± 156||154 ± 148||.29|
|Walks for exercise, n (%)||1,114 (37.3)||57 (32.6)||1,019 (38.5)||38 (22.9)||<.001|
|Married, n (%)||794 (26.3)||42 (23.9)||712 (26.6)||40 (23.8)||.55|
|Trouble sleeping in past month because of pain, times/week, n (%)|
|0||2,018 (66.7)||120(68.2)||1,795 (67.0)||103 (61.3)||.15|
|<1||287 (9.5)||20 (11.4)||255 (9.5)||12 (7.1)|
|1–2||333 (11.0)||15 (8.5)||297 (11.1)||21 (12.5)|
|≥3||385 (12.7)||21 (11.9)||332 (12.4)||32 (19.0)|
|Self-reported health status, n (%)|
|Poor to very poor||66 (2.2)||2 (1.1)||55 (2.1)||9 (5.4)||.007|
|Fair||680 (22.5)||37 (21.0)||595 (22.2)||48 (28.6)|
|Good to excellent||2,277 (75.3)||137 (77.8)||2,029 (75.7)||111 (66.1)|
Circadian Activity Rhythms and Mortality
After a mean of 4.1 years of follow-up, 444 (14.7%) of the analytical sample had died. Subjects with lower peak (amplitude) and mean (mesor) levels of activity and less-robust rhythms (lower pseudo-F values indicate weaker rhythms) had the shortest overall survival. Amplitude was highly correlated with mesor (correlation coefficient (r)=0.78) and robustness (r=0.72), and the results for these measures were similar. Subjects in the lowest quartiles of amplitude, mesor, and robustness had approximately twice the risk of mortality after multivariate adjustment (Table 3). The risks of all-cause mortality increased from highest to lowest quartile of amplitude (P-trend<.001).
|Circadian Activity Rhythm Quartile||All-Cause (N=444)||Cancer (n=92)||Atherosclerotic (n=168)||Stroke (n=54)||Coronary Hear Disease (n=57)||Other Cause (n=184)|
|3,413–4,045||1.02 (0.74–1.40)||1.04 (0.56–1.94)||0.78 (0.45–1.33)||0.54 (0.21–1.37)||1.19 (0.48–2.93)||1.31 (0.78–2.21)|
|2,743–3,412||1.25 (0.92–1.70)||1.45 (0.80–2.63)||1.20 (0.74–1.95)||0.98 (0.44–2.19)||1.31 (0.55–3.17)||1.19 (0.70–2.01)|
|<2,743||2.18 (1.63–2.92)||1.39 (0.73–2.65)||1.81 (1.13–2.90)||1.57 (0.72–3.42)||2.23 (0.95–5.25)||3.11 (1.93–5.00)|
|Midline estimating statistic of rhythm, counts/min|
|2,092–2,386||1.14 (0.85–1.54)||1.44 (0.81–2.59)||0.92 (0.55–1.52)||0.74 (0.33–1.64)||1.72 (0.64–4.60)||1.2 (0.74–1.95)|
|1,796–2,091||1.08 (0.80–1.45)||1.09 (0.58–2.04)||1.01 (0.62–1.64)||0.58 (0.25–1.34)||2.20 (0.86–5.68)||1.14 (0.70–1.84)|
|<1,796||1.71 (1.29–2.27)||1.16 (0.61–2.21)||1.61 (1.02–2.54)||1.12 (0.54–2.35)||2.77 (1.08–7.08)||2.09 (1.33–3.28)|
|Robustness, pseudo-F statistic|
|774–1,096||0.96 (0.71–1.30)||0.98 (0.54–1.78)||1.11 (0.67–1.85)||0.86 (0.37–1.96)||1.30 (0.53–3.22)||0.85 (0.52–1.38)|
|523–773||1.11 (0.82–1.49)||1.17 (0.66–2.10)||1.19 (0.72–1.98)||0.47 (0.17–1.26)||2.18 (0.94–5.05)||1.02 (0.64–1.63)|
|<523||1.97 (1.50–2.60)||1.15 (0.62–2.14)||2.31 (1.45–3.68)||2.07 (0.99–4.31)||2.20 (0.92–5.24)||2.17 (1.43–3.31)|
|Earlier than 12:50 p.m.||0.94 (0.61–1.44)||1.03 (0.42–2.56)||1.08 (0.55–2.13)||2.26 (0.95–5.38)||0.34 (0.05–2.47)||0.76 (0.37–1.56)|
|Later than 4:33 p.m.||1.36 (0.95–1.96)||2.09 (1.04–4.22)||1.66 (0.95–2.90)||2.64 (1.11–6.30)||0.89 (0.28–2.90)||0.83 (0.42–1.64)|
A greater risk of atherosclerotic mortality was also observed for the lowest quartiles of amplitude and mesor activity levels and the lowest quartile of robustness (Table 3). Associations between circadian activity rhythms and mortality from coronary heart disease or stroke, which comprised 34% and 32% of atherosclerotic mortality cases, respectively, largely drove the relationships with atherosclerotic mortality. Greater risk of coronary heart disease mortality was observed in the lowest quartile of mesor. Robustness was associated with mortality from stroke, with greater risk for the lowest quartile than the highest.
Inverse associations between amplitude, mesor, robustness value, and higher “other cause” (nonatherosclerotic and noncancer) mortality rates were also observed. Greater risk of “other cause” mortality was observed in the lowest quartiles of amplitude, mesor, and robustness than in the highest quartiles. Further analysis excluding the two most common causes of “other cause” mortality (pulmonary and cognitive causes) produced similar results, suggesting that these two causes of death do not explain this association.
Acrophase deviation was not associated with all-cause or “other cause” mortality. Delayed acrophase (timing of peak activity later than 4:33 p.m.) was associated with greater mortality from cancer and stroke than the mean peak range of 12:50 to 4:33 p.m. A greater but nonsignificant association was observed between a delayed acrophase and atherosclerotic mortality than with the population mean. An advanced acrophase (timing of peak activity earlier than 12:50 p.m.) was associated with a higher but nonsignificant association risk of stroke mortality.
Additional analyses were performed to determine whether the associations between amplitude, mesor, and all-cause mortality were independent of physical activity level. The associations between amplitude, mesor, and all-cause mortality were consistent between women who did and did not walk for exercise, and the interactions (amplitude × exercise, P=.51; mesor × exercise, P=.67) were not significant. Interactions between mesor and three ambulatory IADLs (difficulty walking 2–3 blocks, climbing up 10 steps, and climbing down 10 steps) were not significant for all-cause mortality (mesor × walking, P=.15; mesor × up 10 steps, P=.14; mesor × down 10 steps, P=.16). The only statistically significant interaction detected was between difficulty walking two to three blocks and amplitude for all-cause mortality (P=.04). Participants in the lowest quartile of amplitude were still at greater risk of all-cause mortality whether they reported having difficulty (HR=3.07, 95% CI=1.80–5.22) or not having difficulty (HR=1.62, 95% CI=1.09–2.42) walking two to three blocks. The interactions between difficulty climbing steps and amplitude were not significant for all-cause mortality (amplitude × up 10 steps, P=.09; amplitude × down 10 steps, P=.25). These results suggest that these circadian parameters are not simply markers of low physical activity that increases the risk for death.
Circadian and homeostatic processes influence circadian rhythms and sleep. This study measured activity levels over several 24-hour periods, which allowed circadian activity rhythms and timing of activity to be specifically examined independent of sleep. Sleep duration was not significantly correlated (P<.05) with amplitude (r=0.05), mesor (r=−0.23), robustness (r=0.26), or acrophase timing (r=−0.03). Adding total sleep time or sleep efficiency to the models did not change the associations between circadian activity parameters and mortality (data not shown).
Kaplan-Meier survival curves show the cumulative incidence of all-cause mortality (Figure 2). There was evidence of a time trend of mortality in the lowest quartiles of amplitude, mesor, and robustness (Figure 2). The Kaplan-Meier curves diverged at approximately the first year and continued through 4 years, with more events in the lowest quartile than the top three quartiles (Figure 2). For acrophase, the Kaplan-Meier curves diverged at approximately the first year and continued through 4 years, with more events in the delayed acrophase group than in the mean or advanced acrophase groups (Figure 2).
This prospective study of actigraphy and mortality in 3,027 older, community-dwelling women showed that circadian activity rhythms were associated with greater risk for all-cause, atherosclerotic, stroke, and “other” mortality, independent of multiple confounders. Lower amplitude and mesor and a less-robust rhythm were consistently related to all-cause, atherosclerotic, stroke, and “other” mortality. The results suggest that the associations between peak and mean activity levels and all-cause mortality were independent of self-reported physical activity. A timing of peak activity later than 4:33 p.m. (>+1.5 SD of the mean) was associated with greater cancer and stroke mortality. To the authors' knowledge, this is the first study in community-dwelling older women to evaluate the relationship between mortality and circadian activity rhythms. Little is known about the causes of age-related changes in circadian patterns, including phase advance and dampened amplitude. These results suggest that the associations between circadian activity rhythms and mortality were independent of actigraphic measures of sleep performance. Evidence of an age-related phase advance is clear from body temperature, sleep–wake cycle, melatonin, and cortisol,37 with a phase difference of approximately 1 hour typically found between young and old individuals. Age-related phase advances are also found in the circadian rhythms of blood pressure, iron, magnesium, neutrophils, and lymphocytes.38 Acrophase deviations from the mean may represent an altered phase relationship between the circadian activity rhythm and the light–dark cycle.
Two possibilities are considered when interpreting these results. First, it is possible that activity rhythm qualities that directly influence mortality in older women independent of other features of aging have been identified. In support of this is emerging animal and human data showing the existence of central and peripheral (e.g., in the liver, pancreas, and other organs) circadian rhythms, with evidence that misalignment of internal rhythms may predispose people to impaired glucose tolerance and alterations in immunological and inflammatory processes. It is also possible that circadian activity rhythms are biomarkers of advanced physiological aging that provide additional risk beyond that of traditional covariates but that may have no direct causal association with mortality. In this instance, the data from the current study may provide evidence that circadian activity rhythms are markers for individuals with greater risk of atherosclerotic, stroke, or cancer death not measured by conventional markers.
A mechanistic connection between disrupted circadian activity rhythm and mortality is an area of active investigation. The circadian system of aged animals is altered in significant ways, which may have an adverse effect on health, especially during phase advances.39 Activity rhythm abnormalities may be a physiological attribute that independently contributes to greater mortality in older women regardless of other aging phenotypes. Specifically, the results of the current study suggest that circadian activity rhythms are associated with cerebrovascular, cardiovascular, and cancer mortality risk independent of other covariates. Pulmonary function, circulating levels of immunological and inflammatory mediators, and autonomic nervous system activity demonstrate marked circadian dependencies.40 Circadian influence of cardiovascular disease arises from a complex interaction of local oscillators in the heart, endothelium, and vascular smooth muscle. This network of local oscillators controls circadian cycles in vasodilatation,41 autonomic tone, blood pressure, and heart rate.42 The tendency of platelets to aggregate is greater after waking,43 and the efficacy of thrombolytic agents in breaking down clots is lowest in the morning.44–47 Shift work has been shown to induce marked alterations in the cardiac autonomic profile.48 The current study found that a delay in timing of peak activity is associated with subsequent risk for cancer and stroke death in elderly women. These findings suggest that acrophase could be used to identify elderly individuals at risk of cancer and stroke mortality. Future study may determine whether elderly individuals with advanced acrophase benefit from interventions (e.g., physical activity, bright light exposure) to regulate circadian activity rhythms and possibly improve health.
Previous findings have suggested that impaired circadian function is associated with a poor prognosis in patients with cancer,49 and the results of the current study suggest that delayed acrophase was associated with a greater risk of cancer mortality. The disruptive effect of circadian desynchronization was highlighted in a study examining patients with metastatic colorectal cancer.28,29 Patients with strong activity rhythms had better quality of life, and 2-year survival was five times as high as in those with rhythm abnormalities.28,29 The relationships between circadian timing, cell cycle, and tumor progression are being elucidated.50 Circadian regulation of the cell cycle has been shown to be important for tumor progression.51 For example, exposure to irregular light–dark cycles51 or destruction of the suprachiasmatic nucleus (SCN)52 in the host and tumor accelerates tumor growth. Survival bias did not influence the results for cancer mortality because the 1,700 participants not included in the analysis and those with wrist actigraphy had the same cancer mortality rates since the eighth visit.
Previous studies suggest that some but not all peripheral circadian oscillators exhibited age-related changes in rhythmicity39 and that some of these tissues retained the capacity to oscillate but were not being appropriately driven in vivo (e.g., by physical activity or feeding).53 The presence of arrhythmic peripheral tissues may be due to weakened behavioral and physiological rhythms that provide less-effective signals to the peripheral oscillators.54 Therefore the change in phase relationships of behavioral and physiological rhythms may not be due to age-related changes in the entrained phase of the SCN itself but rather because of age-related alterations in other rhythmic components of the circadian system.39 If it is the case that circadian desynchronization affects independent organ systems differently, as experimental systems have shown and the data from the current study suggest, and some peripheral oscillators retain their capacity to oscillate, then interventions to regulate circadian activity rhythm abnormalities including physical activity may be warranted in older adults.
This analysis had a number of strengths. This was a large cohort of community-dwelling older women with no inclusion requirements regarding circadian activity rhythms or sleep disorders. Multiple possible confounders were also adjusted for. This analysis also had several limitations. Findings are for older, community-dwelling women and may not be generalizable to other populations such as men, institutionalized older women, or younger women. The individuals who agreed to wear the actigraph were somewhat healthier than those who did not, although it is likely that this would have biased the findings toward the null.
These findings suggest that activity rhythm abnormalities are prognostic of greater risk of mortality in older community-dwelling women. The mechanism behind this association between circadian activity rhythms and mortality is not clear based on these observational data. If confirmed in other cohorts, studies will be needed to examine whether interventions (e.g., physical activity, bright light exposure) that regulate circadian activity rhythms will improve health outcomes in older adults.
This work was supported by grants from the National Institutes of Health: AG05407, AR35582, AG05394, AR35584, AR35583, AR46238, AG005407, AG08415, AG027576-22, AG005394-22A1, AG027574-22A1, AG030474.
Investigators in the Study of Osteoporotic Fractures Research Group: San Francisco Coordinating Center (California Pacific Medical Center Research Institute and University of California San Francisco): S. R. Cummings (principal investigator), M. C. Nevitt (co-investigator), D. C. Bauer (co-investigator), D. M. Black (co-investigator), K. L. Stone (co-investigator), W. Browner (co-investigator), R. Benard, T. Blackwell, P. M. Cawthon, L. Concepcion, M. Dockrell, S. Ewing, M. Farrell, C. Fox, R. Fullman, S. L. Harrison, M. Jaime-Chavez, W. Liu, L. Lui, L. Palermo, N. Parimi, M. Rahorst, D. Kriesel, C. Schambach, R. Scott, J. Ziarno. University of Maryland: M. C. Hochberg (principal investigator), R. Nichols (clinic coordinator), S. Link. University of Minnesota: K. E. Ensrud (principal investigator), S. Diem (co-investigator), M. Homan (co-investigator), P. Van Coevering (program coordinator), S. Fillhouer (clinic director), N. Nelson (clinic coordinator), K. Moen (assistant program coordinator), F. Imker-Witte, K. Jacobson, M. Slindee, R. Gran, M. Forseth, R. Andrews, C. Bowie, N. Muehlbauer, S. Luthi, K. Atchison. University of Pittsburgh: J. A. Cauley (principal investigator), L. H. Kuller (co-principal investigator), J. M. Zmuda (co-investigator), L. Harper (project director), L. Buck (clinic coordinator), M. Danielson (project administrator), C. Bashada, D. Cusick, A. Flaugh, M. Gorecki, M. Nasim, C. Newman, N. Watson. The Kaiser Permanente Center for Health Research, Portland, Oregon: T. Hillier (principal investigator), K. Vesco (co-investigator), K. Pedula (co-investigator), J. Van Marter (project director), M. Summer (clinic coordinator), A. MacFarlane, J. Rizzo, K. Snider, J. Wallace.
Conflict of Interest: J. A. Cauley receives funding from Merck & Company, Eli Lily & Company, Pfizer Pharmaceuticals, and Novartis Pharmaceuticals. K. E. Ensrud has received research support from California Pacific Medical Center, which receives funding from Roche Molecular Systems. S. Cummings, K. Stone, T. Blackwell, and G. Tranah receive research support from Roche Molecular Systems. S. Ancoli-Israel has received grants for educational activities from Cephalon, Sepracor, and Takeda Pharmaceuticals North America. She has served as an advisor or consultant to Acadia, Cephalon, Ferring, Pfizer, Respironics, Sanofi-Aventis, Sepracor, Somaxon, and Takeda Pharmaceuticals North America.
Author Contributions: G. Tranah participated in the conception and design of the project and drafted and revised the manuscript. T. Blackwell conducted all statistical analyses and participated in interpretation of data analyses and critical revision of the manuscript. K. Stone obtained funding and participated in the acquisition of data, interpretation of data analyses, and critical revision of the manuscript. S. Redline and K. E. Ensrud participated in the acquisition of data, interpretation of data analyses, and critical revision of the manuscript. M. Paudel, J. A. Cauley, T. Hillier, and S. Cummings participated in interpretation of data analyses and critical revision of the manuscript. S. Ancoli-Israel participated in the conception and design of the project, interpretation of data, and critical revision of the manuscript.
Sponsor's Role: None.
- 21Poor sleep is associated with impaired cognitive function in older women: The Study of Osteoporotic Fractures. J Gerontol A Biol Sci Med Sci 2006;61A:405–410., , et al.
- 27The timing of activity rhythms in patients with dementia is related to survival. J Gerontol A Biol Sci Med Sci 2004;59A:1050–1055., , et al.
- 33Geriatric Depression Scale (GDS): Recent evidence and development of a shorter version. In: Brink TL, editors. Clinical Gerontology: A Guide to Assessment and Intervention. Haworth, NY: The Haworth Press, 1986, pp 165–173.,
Appendix SA1. Comparison of Characteristics of the Study of Osteoporotic Fractures Cohort According to Quartiles of Mean Activity Level.
Appendix SB1. Comparing Characteristics of the Study of Osteoporotic Fractures (SOF) Cohort Among Quartiles of Robustness (Pseudo-F Statistic for Goodness of Extended Cosine Fit).
Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.
|JGS_2674_sm_appendix.doc||115K||Supporting info item|
Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.