The Relationship Among Step Count, Step Intensity, and Mortality in Adults
August 1, 2020
Reprints
By Ellen Feldman, MD
Altru Health System, Grand Forks, ND
Dr. Feldman reports no financial relationships relevant to this field of study.
SUMMARY POINTS
• In this prospective investigation based on data collected between 2003 and 2006, 4,840 U.S. adults used accelerometers to track seven days of step count and step intensity (measured by steps/minute).
• Mortality records for the group were followed until 2015.
• Participants taking fewer than 4,000 steps/day had an unadjusted incident density rate for all-cause mortality rate of 76.7/1,000 person-years, compared to 4.8/1,000 person-years in the group taking at least 12,000 steps/day.
• When comparing 12,000 steps daily to 4,000 steps daily, the hazard ratio (HR) for adjusted all-cause mortality was 0.35 (confidence interval [CI], 0.28-0.45) or a 65% risk reduction associated with longer walking.
• After adjusting step intensity for number of steps/day, there was no association noted between step intensity and mortality; HR was 0.90 (95% CI, 0.65-1.27; P = 0.34) when comparing the highest and lowest quartiles of step intensity and measuring association with mortality.
SYNOPSIS: The authors of this observational study of 4,840 adults show an association between higher number of steps taken daily and lower all-cause mortality, lower mortality from cardiovascular disease, and lower mortality from cancer, but no association between intensity of steps and mortality in any of those areas.
SOURCE: Saint-Maurice PF, Troiano RP, Bassett DR, et al. Association of daily step count and step intensity with mortality among U.S. adults. JAMA 2020;323:1151-1160.
In a 1786 letter to his son-in-law, Thomas Jefferson wrote, “I have known some great walkers and had particular accounts of many more; and I never knew or heard of one who was not healthy and long lived.”1 More than 200 years later, Saint-Maurice et al have used research methods from 2020 to understand and delineate the relationship among walking, health, and longevity long ago noted by one of the United States’ “founding fathers.”
Saint-Maurice et al noted the lack of evidence-based studies supporting the popular goal of achieving 10,000 steps each day. Prior investigations involving walking and health have affirmed the health benefits of walking, but typically involved limited and specific populations.2,3 Therefore, findings are not necessarily applicable to other groups or to a more general population.
In addition, there have been conflicting results in studies looking at step intensity and association with health benefits.4,5 Citing these reasons, this team set out to look at any association between step count, step intensity, and mortality in a representative sample of the U.S. population older than 40 years of age.
The National Health and Nutrition Examination Survey (NHANES), a Centers for Disease Control and Prevention-sponsored program, annually collects data from a nationally representative sample of the U.S. population.6
From 2003 to 2006, 4,840 respondents aged 40 years or older wore standardized hip accelerometers during a seven-day period. Saint-Maurice et al used data from the accelerometers to determine participant step count and step intensity. Given the breadth of information collected by NHANES, the team was able to adjust results for a variety of factors, including age, education, substance use, specific medical diagnosis, and self-reported general health. There were 6,355 respondents initially eligible for inclusion in the study, but only 4,840 had sufficient data for inclusion; 1,515 individuals either elected not to wear the accelerometer or did not submit a record for at least one 12-hour period.
The National Death Registry was accessed periodically until Dec. 31, 2015, to determine death of any participant from any cause. Indications that a participant had died casued by either cancer or cardiovascular disorder were noted as well.
The mean number of steps/day for the entire group of 4,840 participants was 9,124. The mean length of time wearing the accelerometer was 5.7 days, while the mean time each day was 14.4 hours.
Respondent data was stratified according to number of daily steps:
- < 4,000 steps/day (n = 655)
- 4,000-7,999 steps/day (n = 1,727)
- 8,000-11,999 steps/day (n = 1,539)
- > 12,000 steps/day (n = 919)
Table 1 shows the selected characteristics of these groups and highlights statistically significant differences.
Table 1. Characteristics of Respondent Groups by Step Count |
|||||
|
< 4,000 steps per day (n = 655) |
4,000-7,999 steps per day (n = 1,727) |
8,000-11,999 steps per day (n = 1,539) |
> 12,000 steps per day (n = 919) |
P value |
|
|
Mean age (years) |
69.9 |
59.9 |
54.0 |
51.1 |
< 0.001* |
|
Body mass index |
31.4 |
29.9 |
28.6 |
27.1 |
< 0.001* |
|
Current smoker |
109 (18.7%) |
344 (22%) |
302 (19.9%) |
218 (22.8%) |
0.62 |
|
Diabetes |
204 (29.9%) |
284 (13.9%) |
168 (7.8%) |
69 (4.5%) |
< 0.001* |
|
Heart failure |
102 (16.2%) |
92 (4.5%) |
38 (1.7%) |
12 (1.0%) |
< 0.001* |
|
*Statistically significant data |
|||||
Table 2 depicts the unadjusted results, including incident density rate, defined as the mortality rate within a specific timeframe, for all-cause mortality, as well as number and percentage of deaths/quartile. Given the differences between the groups, and an analysis showing significant attenuation of results with adjustment for some of the variables, Saint-Maurice et al presented several interpretive models for review and discussion.
Table 2. Mortality Data for Respondent Groups by Step Count |
||||
|
< 4,000 steps (n = 655) |
4,000-7,999 steps per day (n = 1,727) |
8,000-11,999 steps per day (n = 1,539) |
> 12,000 steps (n = 919) |
|
|
Number of deaths |
419 |
488 |
176 |
82 |
|
Percent of deaths |
56.5% |
21.3% |
7.3% |
5.1% |
|
Unadjusted incident density rate all-cause mortality per 1,000 person years |
76.7 |
21.4 |
6.9 |
4.8 |
Table 3 shows the results adjusted for multiple variables, including age; diet quality; sex; race/ethnicity; body mass index; education; alcohol consumption; smoking status; diagnoses of diabetes, stroke, coronary heart disease, heart failure, cancer, chronic bronchitis, and emphysema; mobility limitation; and self-reported general health. In addition to all-cause mortality, the authors examined death as the result of cardiovascular disease and cancer. Each revealed a similar pattern of a significant decrease in mortality rate with increasing step count. In addition, results indicated consistent findings of decreased mortality rate with increased step count when reported separately for men, women, age groups, and ethnicity.
Table 3. Adjusted Mortality Rate and Hazard Ratios |
||||||
|
4,000 steps per day |
6,000 steps per day |
8,000 steps per day |
10,000 steps per day |
12,000 steps per day |
P trend |
|
|
Adjusted mortality rate per 1,000 adults per year |
14.4 (95% CI, 13.0, 15.7) |
9.7 (95% CI, 8.6-10.8) |
7.1(95% CI, 5.7-8.4) |
5.7 (95% CI, 4.4-7.0) |
5.1 (95% CI, 3.8-6.5) |
< 0.001* |
|
Hazard ratio |
REFERENT |
0.68 (95% CI, 0.64, 0.72) |
0.49 (95% CI, 0.44, 0.55) |
0.40 (95% CI, 0.34, 0.46) |
0.35 (95% CI, 0.28, 0.45) |
|
|
CI: Confidence interval |
||||||
Table 4 shows the findings for cardiovascular and cancer deaths for each step quartile.
Table 4. Unadjusted MR, Number, and Percentage of Deaths as a result of CVD and Cancer |
||||
|
< 4,000 steps |
4,000-7,999 steps |
8,000-11,999 steps |
> 12,000 steps |
|
|
Cancer deaths |
62 |
133 |
56 |
32 |
|
Cancer death percentage |
7.8% |
6.4% |
2.5% |
2.0% |
|
MR*/1,000 adults/year because of cancer |
10.5 |
6.4 |
2.3 |
1.8 |
|
CVD deaths |
169 |
162 |
52 |
32 |
|
CVD death percentage |
23.2% |
6.5% |
2.3% |
1.1% |
|
MR/1,000 adults/year because of CVD |
31.6 |
6.6 |
2.1 |
1.0 |
|
MR: Mortality rate CVD: Cardiovascular disease |
||||
The second major part of this study looked at step intensity and mortality rate (MR). Step intensity was calculated in several manners (based on time and number of steps). Higher step intensity was associated with significantly lower MR — until adjusting these figures for number of steps daily.
For example, the unadjusted MR was 5.2 (95% confidence interval [CI], 3.2-7.3) per 1,000 adults/year for those in the highest quadrant of step intensity and 10.0 (95% CI, 7.1-12.9) for the respondents on the other end of the scale. However, when adjusting for the number of steps in both quadrants, this difference virtually vanishes, with an adjusted MR of 8.4 (95% CI, 4.0-12.9) per 1,000 adults/year and 9.2 (95% CI, 6.9-11.8) per 1,000 adults/year. The P value for the trend is 0.34.
Commentary
This study gives backbone to the principle of viewing physical activity as essential “medicine” for health. The association between step count and longevity found by Saint-Maurice et al is consistent with studies of step count and health both in investigations centered on older adults and in studies of persons with chronic illnesses.2,3 This work adds to the literature by generalizing the findings from these narrower studies and pointing toward a dose relationship between step count and longevity.It is important to note that this is an observational study; there is no evidence for causation. There are some significant limitations in this investigation, including that the participant group was self-selected, and that wearing of the accelerometer itself may have had an effect (such as motivating more walking or motivating a more active lifestyle). Future investigations with randomization and controls will be helpful in further understanding and quantifying the relationship between step count and longevity.
It is notable that, even when adjusted for a substantial number of variables, results still indicate a dose-response relationship between step count and longevity, with hazard ratio indicating a 65% reduced chance of death among the group taking 12,000 or more steps daily when compared to the referent group (taking 4,000 steps or fewer daily).
Future investigations may want to look more closely at the lifestyle of the most active group to see if there are other variables influencing longevity, such as participation in activities that may not register on an accelerometer (e.g., biking, swimming, etc.).
Another point to consider is that the only accelerometer data came from a seven-day period between 2003 and 2006. There is little evidence to suggest that this period is representative of a respondent’s lifestyle in general and/or over time. More information and data from subsequent years will be helpful in evaluating whether accelerometer patterns stay stable over time and ultimately may help in attempts to address the question of causation.
The investigators expected to see an association of step intensity with mortality and were surprised at not finding this relationship. They noted that there have been only a smattering of studies looking at the impact of step intensity on mortality and suggested further studies to understand this in full.
In this paper, it appears that step intensity is associated with greater number of steps in general.
Even with the limitations noted, the Saint-Maurice et al study represents a true “step” forward in the field. While it is prudent to wait for evidence of causation, there is no need to wait for recommending an active lifestyle to patients. Today, many of our patients have wearable devices that measure not only step count, but also many other health parameters associated with activity, including heart rate, recovery time, etc. Use of these devices and active monitoring may become an integral component of a modern wellness plan.
We know that activity is helpful for a variety of health measures and conditions, including reducing the risk of type 2 diabetes, obesity, and heart disease. This study strengthens the argument for continuing to keep physical activity as the foundation of a healthy lifestyle and adds evidence to a role for physical activity in longevity.
REFERENCES
- Thomas Jefferson's Monticello. Exercise. Nov. 6, 1990. https://www.monticello.org/site/research-and-collections/exercise
- O'Donnell J, Smith-Byrne K, Velardo C, et al. Self-reported and objectively measured physical activity in people with and without chronic heart failure: U.K. Biobank analysis. Open Heart 2020;7:e001099.
- Dohrn IM, Welmer AK, Hagströmer M. Accelerometry-assessed physical activity and sedentary time and associations with chronic disease and hospital visits – a prospective cohort study with 15 years follow-up. Int J Behav Nutr Phys Act 2019;16:125.
- Lee IM, Shiroma EJ, Kamada M, et al. Association of step volume and intensity with all-cause mortality in older women. JAMA Intern Med 2019;179:1105-1112.
- Ekelund U, Steene-Johannessen J, Brown WJ, et al. Does physical activity attenuate, or even eliminate, the detrimental association of sitting time with mortality? A harmonised meta-analysis of data from more than 1 million men and women. Lancet 2016;388:1302-1310.
- Centers for Disease Control and Prevention. About the National Health and Nutrition Examination Survey. Updated Sep. 15, 2017. https://www.cdc.gov/nchs/nhanes/about_nhanes.htm
The authors of this observational study of 4,840 adults show an association between higher number of steps taken daily and lower all-cause mortality, lower mortality from cardiovascular disease, and lower mortality from cancer, but no association between intensity of steps and mortality in any of those areas.
Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.