Six Longevity Markers and the Week That Moves Them

Resting heart rate carries 18% of Titan Age, VO2 max carries 16%, and nightly sleep carries 12%. Those three weights, along with eight smaller ones, turn a week of Apple Health data into an age you can set beside the date on your birth certificate. Each weight stands in for a body of mortality research, some of it built on a million participants and some on a few dozen. Knowing which is which tells you how much to trust the number, and which habits move it for reasons that would still hold if the app did not exist.

This article walks through six markers a phone and a watch can measure, with the named study and the numbers behind each. It then lays out one training week that moves all six together, explains why WHOOP Age, Oura Cardiovascular Age, Garmin Fitness Age and Titan Age can disagree by years on the same week, and gives a cadence for reading the trend. Everything here stays inside what a wrist sensor and a smart scale can record. Blood panels, clinical treadmill tests and epigenetic clocks are out of scope.

01Healthspan, lifespan and pace of aging

Lifespan is the count of years you live. Healthspan is the count of those years you spend without disease or disability that limits what you do. For someone who already trains, the practical target is narrower still. It is the number of years you can keep training at a level you care about, and the markers below predict that span better than they predict a date of death.

Biological age is an estimate of how old your body behaves, set on the same scale as your actual age. Pace of aging is its rate of change. A 45 year old with a biological age of 40 has a good starting position. A 45 year old whose biological age drops by half a year every calendar year is improving, whatever the starting position. Pace is the more useful number for an athlete, because training changes the rate long before it changes the level.

Laboratory research measures pace with epigenetic tools such as DunedinPACE, which reads DNA methylation from a blood sample. Kawamura and colleagues (2025, GeroScience) tested 144 men aged 65 to 72 with a cycle ergometer and breath-by-breath gas analysis, then compared their VO2 peak with DunedinPACE. Higher fitness went with slower biological aging, with a correlation of r = −0.17 after adjustment. That is a weak correlation. Fitness explains a small share of the variation in one pace measure among older men, and the study is cross-sectional, so it cannot say which way the effect runs. Keep that result in mind when a wearable reports your pace of aging to one decimal place.

Titan reports pace as a percent. With fewer than 12 weeks of history it is the gap between Titan Age and your actual age as a share of your actual age. From 12 weeks on, it is the yearly rate of change in Titan Age, drawn from up to 52 weeks, as a percent of your current Titan Age. A negative pace is the favorable direction in both cases. The Titan Age help article has the full definition.

02Six markers with mortality data behind them

Each marker below has at least one large cohort or meta-analysis linking it to all-cause mortality, and each one responds to training within weeks to months. The table gives the headline finding. The sections that follow give the detail and the caveats.

Every row is an observational association. Fit people differ from unfit people in income, diet, smoking and chronic disease, and adjustment removes only the confounders the authors measured. The consistency of the direction across cohorts, sexes and countries is what gives the list its weight.

VO2 max and cardiorespiratory fitness

Kodama and colleagues (2009, JAMA) pooled 33 cohort studies covering 102,980 healthy adults and 6,910 deaths. Each 1 MET of higher maximal aerobic capacity went with a pooled risk ratio of 0.87 for all-cause mortality and 0.85 for coronary and cardiovascular events. One MET equals 3.5 ml/kg/min of oxygen uptake, roughly a 1 km/h faster running speed. People below 7.9 METs, about 28 ml/kg/min, had 1.70 times the mortality risk of people at 10.9 METs or above.

Mandsager and colleagues (2018, JAMA Network Open) followed 122,007 patients for a median of 8.4 years after a treadmill test at Cleveland Clinic. The lowest fitness group for age and sex had 5.04 times the adjusted mortality of the elite group, and elite performers still had lower risk than the high group, with a hazard ratio of 0.77. The authors found no upper limit to the benefit. The American Heart Association statement by Ross and colleagues (2016, Circulation) drew the clinical conclusion. It argued that cardiorespiratory fitness predicts mortality at least as well as smoking, hypertension, high cholesterol and type 2 diabetes, and that clinicians should record it like a vital sign.

The cohorts above measured fitness on a treadmill. A watch estimate is a model output with its own error. Apple Watch estimates VO2 max, labeled Cardio Fitness in the Health app, from heart rate against pace on outdoor walks, runs and hikes, and validation studies put its error at several ml/kg/min. Titan Age takes the latest VO2 max reading in Apple Health from the past 30 days. Age and sex norms, testing methods and wearable accuracy are covered in VO2 max and aerobic capacity, and the short definition is in the glossary entry on VO2 max.

Resting heart rate and heart rate recovery

Zhang, Shen and Qi (2016, CMAJ) combined 46 prospective studies with 1,246,203 participants and 78,349 deaths. Each 10 bpm increase in resting heart rate carried a relative risk of 1.09 for all-cause mortality and 1.08 for cardiovascular mortality. People above 80 bpm had a relative risk of 1.45 against the lowest category. The rise in all-cause risk was linear from 45 bpm upward, and adjustment for traditional cardiovascular risk factors did not change it.

A falling resting heart rate over months reflects a larger stroke volume and more vagal tone at rest, two of the most reliable adaptations to aerobic training. Titan Age gives resting heart rate its largest single weight, 18%, and needs readings on at least 4 days of the week to use it. The reference value is 80 bpm and the step is 5 bpm. A week averaging 60 bpm sits four steps better than the reference, which lowers Titan Age by about 1.7 years at full coverage.

Heart rate recovery measures how fast your heart rate falls after hard effort, which depends on how quickly the vagus nerve reengages. Cole and colleagues (1999, New England Journal of Medicine) followed 2,428 adults for six years after a symptom-limited exercise test. They defined an abnormal result as a drop of 12 bpm or less in the first minute after peak exercise. The 26% of patients below that line died at 4.0 times the rate of the rest, and 2.0 times after adjustment for age, workload, resting heart rate and perfusion defects.

Apple Watch records heart rate recovery at one minute after a workout. Titan multiplies that value by 1.65 to estimate the two-minute drop, keeps the result between 10 and 70 bpm, and needs at least two workout days in the week. Cole's patients stopped at a true peak, and a watch reading after an easy run starts from a lower heart rate and shows a smaller drop. Compare recovery values after sessions that end at a similar intensity, such as the last interval of the same workout each week.

The Vitals screen shows resting heart rate as a latest day value, a baseline median for the period you pick, and a shaded band covering the 20th to 80th percentile of your days. Set the period to 6M and the band describes half a year, so one hard week barely moves it. That is the view to use for the slow trend this research describes.

Heart rate variability

HRV has the thinnest longevity evidence of the six markers and the most marketing. Hernández-Vicente and colleagues (2020, Frontiers in Physiology) recorded 15 minutes of seated RR intervals in 20 young adults, 18 octogenarians and 17 centenarians. Parasympathetic HRV indices fell with age, with little difference between the octogenarians and the centenarians. Among the 14 centenarians followed until death, only SDNN correlated with survival time, and SDNN below 19 ms carried a hazard ratio of 5.72 for death within a year. Fourteen people is a very small sample, and the finding says more about frailty at 100 than about training at 40.

Wiersema and colleagues (2022, Aging) looked for a longevity signal earlier in life. They fitted wearable ECG monitors for 24 to 90 hours on 37 middle-aged offspring of long-lived parents and 36 of their partners, plus 35 young adults. Mean heart rate, HRV and 24-hour heart rate rhythm did not differ between the offspring and the partners. Age did separate the groups, with young participants showing larger daily rhythms and better long-range HRV structure than the middle-aged ones.

The two studies together place HRV where it works best, as a measure of current autonomic state that falls with age and illness. Neither study shows that a higher HRV in midlife predicts a longer life once fitness and resting heart rate are accounted for, and Wiersema's null result argues for caution. Titan's design matches that reading. HRV is not one of the 11 Titan Age signals. It drives the daily Recovery score instead, where its day-to-day movement is the point.

Titan does not judge HRV against a fixed millisecond cutoff. It compares last night's HRV and sleeping heart rate with the median of your measured nights in the Recovery window, 60 nights by default, in units of your own median absolute deviation, and reports a 0 to 100 score. Low is 0 to 39, Moderate is 40 to 74 and High is 75 to 100. Outside the app, use this rule of thumb. A morning HRV more than one median absolute deviation below your baseline median for two or more consecutive days, with resting heart rate above baseline, is a signal to reduce training intensity. Personal baselines explains the window, and the glossary entry on HRV covers the measurement itself.

One practical warning follows from the centenarian study. Its 19 ms SDNN figure came from a 15-minute seated recording on a chest strap. Apple Watch reports SDNN by default, and Titan can switch every HRV value to RMSSD with Use RMSSD for HRV under Settings > Recovery & Sleep. Neither is comparable to a laboratory cutoff. Compare your HRV with your own baseline and nobody else's.

Sleep regularity

Windred and colleagues (2024, Sleep) gave 60,977 UK Biobank participants a wrist accelerometer for seven days and computed a Sleep Regularity Index, the probability of being in the same sleep or wake state at any two time points 24 hours apart. Over an average follow-up of 6.3 years there were 1,859 deaths. Against the least regular fifth, the four more regular fifths had fully adjusted hazard ratios of 0.80, 0.75, 0.72 and 0.70 for all-cause mortality. Most of the benefit arrived by the second fifth, which means the largest gain comes from leaving the most irregular group. In model comparisons, regularity predicted mortality better than sleep duration, and adding duration to a regularity model did not improve it.

Titan measures two sides of this. Sleep Consistency scores each night's hours asleep against the median of the prior 7 days, losing 25 percentage points for every half hour of difference. Target Wake Time is the median wake time over the last 14 days. The regularity index in Windred's study rewards stable timing as well as stable duration, so the wake time anchor is the closer match to the research. Titan Age uses both sleep hours, at 12%, and Sleep Consistency, at 7%, and each needs at least 4 days of data in the week. The sleep consistency help article shows the scoring table.

Strength and grip strength

Shailendra and colleagues (2022, American Journal of Preventive Medicine) pooled 10 studies of resistance training and mortality. Any resistance training, compared with none, went with a 15% lower risk of all-cause mortality. The dose-response analysis found the largest reduction, 27%, at about 60 minutes per week, with smaller reductions at higher volumes.

Zhang and colleagues (2026, British Journal of Sports Medicine) followed 147,374 nurses and other health professionals for up to 30 years, with resistance training reassessed every two years. After adjusting for aerobic activity, 90 to 119 minutes per week went with a hazard ratio of 0.87 for all-cause mortality, and above 120 minutes added nothing. The joint analysis is the more useful result for planning. People who did 30 to less than 45 MET-hours per week of aerobic activity plus 60 to 119 minutes of resistance training had a hazard ratio of 0.55 against people with little aerobic activity and no lifting. At 45 MET-hours of aerobic activity or more, roughly four and a half hours of running at 10 km/h each week, risk bottomed out at hazard ratios of 0.53 to 0.58 whatever the lifting volume. For most people, two lifting sessions on top of a moderate aerobic week reach that floor at a lower cost in hours than more running.

Grip strength is a marker of whole-body strength, and squeezing a dynamometer is not the intervention. Xiong and colleagues (2023, Frontiers in Nutrition) followed 10,470 NHANES participants for an average of 81 months. Against the strongest fifth, the weakest fifth had hazard ratios of 3.45 for all-cause mortality in men and 3.3 in women after full adjustment. The authors set low grip strength at under 37.4 kg for men and under 24 kg for women. Titan Age counts weekly strength training minutes at 5%, and a single session in the week counts. Titan has no grip input, so the practical step is a squeeze test every few months if you own a dynamometer.

Body composition and lean mass

Body mass index cannot separate muscle from fat, and the two carry opposite risk. Srikanthan and Karlamangla (2014, American Journal of Medicine) followed 3,659 older NHANES III participants and measured skeletal muscle with bioelectrical impedance. They divided muscle mass by height squared to get a muscle mass index. The highest quarter had a hazard ratio of 0.80 for all-cause mortality against the lowest, after adjustment for waist-to-hip ratio.

Lee and colleagues (2018, BMJ) predicted lean and fat mass for 38,006 men in the Health Professionals Follow-up Study and followed them for an average of 21.4 years. Fat mass rose with mortality in a straight line, with a hazard ratio of 1.35 for the highest fifth against the lowest. Mortality risk followed a U-shaped curve across lean mass levels, with risk in the middle three fifths 8% to 10% below the lowest fifth. Their conclusion was that the familiar J-shaped curve for BMI comes from adding two opposite curves together. See the glossary entries on body composition and body mass index.

Titan's Lean Mass Index scales lean mass to height, the same idea as Srikanthan's muscle mass index. It needs body mass, height and body fat percentage in Apple Health, uses the latest values from the past 30 days, and carries 3% of Titan Age. A smart scale that writes body fat to Apple Health is enough to populate it. Consumer scales estimate body fat from impedance, which shifts with hydration, so read the index over months.

03What a week that moves every marker looks like

The Zone 2 against HIIT argument has a clearer answer than its volume suggests. Milanović, Sporiš and Weston (2015, Sports Medicine) pooled 28 controlled trials with 723 adults aged 18 to 45. Against no-exercise controls, continuous endurance training raised VO2 max by 4.9 ml/kg/min and high-intensity interval training by 5.5 ml/kg/min. Intervals gave slightly more, and the gap grew in people with lower starting fitness and in longer programs. Both work. The mortality cohorts add a second reason to keep the easy volume. Zhang's lowest risk groups reached 30 to 45 MET-hours of aerobic activity a week, and that much work has to be mostly easy to be sustainable.

The week below combines two to three strength sessions, three easy aerobic sessions and one interval session, with a fixed wake time every day. Titan's default Zone 2 is 74% to 80% of max heart rate, and Zone 4 starts at 87%. Check or edit your zones in heart rate zones, and see the glossary entry on heart rate zones for other zone models.

The totals are about 120 minutes of lifting, 195 minutes of Zone 2 and 16 minutes of hard interval work plus warm-up. The lifting total sits at the plateau Zhang found and above the 60-minute peak in Shailendra's pooled data. Two sessions of 45 minutes cover most of the benefit if the third does not fit. The interval format is a common one for VO2 max, and VO2 max and aerobic capacity covers the alternatives.

Three adjustments keep the week honest. First, the Wednesday session is the one to move when Recovery reads Low, since the app describes that band as a day to favor easy work, sleep and nutrition. Second, watch the load ratio in training load. Titan treats a ratio of 0.8 up to 1.4 as the optimal zone, so add volume in small steps if you are starting from less. Third, the easy days are where steps and stand hours come from. Titan Age gives weekly steps 9% and stand hours 5%, together nearly three times the weight it gives lifting.

For a longer treatment of the aerobic side, see the endurance training guide. For the strength side, see the muscle building guide.

04Why every biological age number disagrees

Four apps can look at the same week and report four ages several years apart. None of them is broken. They measure different things over different windows and scale the result to age in different ways.

The first source of disagreement is the input. Oura's number describes your arteries, and nothing in a hard training block changes arterial stiffness in a week. Garmin's number starts from a VO2 max estimate, and Garmin, Apple and WHOOP each estimate VO2 max with a different algorithm, so the same runner can carry three VO2 max values. WHOOP and Titan both use broad sets of recovery and activity signals, with different weights. Titan also leaves sex out of the model, while Garmin compares you with age and sex norms.

The second source is time. A six-month window, as described for WHOOP Age, means one week is a small fraction of the data. Titan Age runs on a single completed week, then moves between 8% and 62% of the way from last week's age toward the new weekly result. Better coverage and a history of large weekly swings move it faster. Sparse data and a steady history move it slower. After three quiet months, a strong training block should show up in Titan Age within a few weeks and in a six-month model much later.

The third source is the reference scale. Each app turns its inputs into years using its own reference values and step sizes. Titan's resting heart rate reference is 80 bpm with a 5 bpm step, and each signal changes the age estimate by progressively fewer years for each additional step away from its reference. A competitor with a different reference population will translate the same 55 bpm into a different number of years.

Each score is a model built from markers with mortality data, and none of the vendor pages cited here reports a mortality validation of the score itself. That leads to one practical rule. Pick one score, learn how it moves, and trust its direction over months rather than its level on any single week. The companion article on Titan Age and biological age goes deeper into how these models and the laboratory clocks compare.

05How to read a trend instead of chasing a number

Each marker changes on its own schedule, so each deserves its own check-in.

Resting heart rate and HRV change daily, and Recovery already handles that scale. Look at them each morning to decide the day's session, and look at the 6M view in Vitals once a month for the slow trend. VO2 max estimates move over weeks, and a single reading after a hot or hilly run can swing several points, so read the monthly direction. Lean mass moves over months. Sleep regularity responds within a week, which makes the wake time anchor the fastest lever you have.

Titan Age updates once a week, after Sunday closes. Coverage decides how much each week counts. Resting heart rate, sleep, sleep consistency, steps and stand hours each need 4 days of data. When a signal is missing, Titan fills it with the median of your recorded weeks from the past 26, once you have at least 6, and missing signals reduce how much every other signal changes the age estimate. A week without your watch at night pulls the estimate toward your history instead of toward a worse age. Pace becomes a yearly rate once you have 12 weeks of history, and a Titan Age that falls 0.02 years each week at a Titan Age of 35 reads as a pace of −3.0%.

A reasonable cadence is to open Titan Age on Monday, read the top three movers, and act only on a pattern. One week of fewer steps during travel is noise. Three weeks with Zone 2 minutes contributing the largest increase in Titan Age means the easy volume has slipped, and the fix is on Tuesday and Saturday of the plan above. If Titan Age shows no estimate at all, Titan Age missing lists the causes, starting with a missing date of birth in Apple Health.

06The Titan Age screen

Titan Age is the one screen in Titan that combines these markers into a single number. Tap the Titan Age card on Today to open it. The Age history chart shows Titan Age and pace for any week you select. Below it, each signal row shows its age contribution in years, and Calculation details lists each signal's value, target, weight and days of coverage for the week. The top three weekly movers rank the signals by how much their contribution changed from the week before. A negative mover pulled Titan Age down, and a positive one pushed it up. The Titan Age help article documents every signal, weight and rule used above.

07References

• Cole CR, et al. (1999). Heart-rate recovery immediately after exercise as a predictor of mortality. New England Journal of Medicine 341(18):1351-1357. doi:10.1056/NEJM199910283411804
• Garmin (2025). What is your Garmin fitness age? How can you lower it? Garmin blog. garmin.com
• Hernández-Vicente A, et al. (2020). Heart rate variability and exceptional longevity. Frontiers in Physiology 11:566399. doi:10.3389/fphys.2020.566399
• Kawamura T, et al. (2025). Revisiting the relationship between cardiorespiratory fitness and biological aging: insights from DunedinPACE analysis. GeroScience. doi:10.1007/s11357-025-01930-0
• Kodama S, et al. (2009). Cardiorespiratory fitness as a quantitative predictor of all-cause mortality and cardiovascular events in healthy men and women: a meta-analysis. JAMA 301(19):2024-2035. doi:10.1001/jama.2009.681
• Lee DH, et al. (2018). Predicted lean body mass, fat mass, and all cause and cause specific mortality in men: prospective US cohort study. BMJ 362:k2575. doi:10.1136/bmj.k2575
• Mandsager K, et al. (2018). Association of cardiorespiratory fitness with long-term mortality among adults undergoing exercise treadmill testing. JAMA Network Open 1(6):e183605. doi:10.1001/jamanetworkopen.2018.3605
• Milanović Z, et al. (2015). Effectiveness of high-intensity interval training (HIT) and continuous endurance training for VO2max improvements: a systematic review and meta-analysis of controlled trials. Sports Medicine 45(10):1469-1481. doi:10.1007/s40279-015-0365-0
• Oura Team (2025). All about Cardiovascular Age. Oura blog. ouraring.com
• Ross R, et al. (2016). Importance of assessing cardiorespiratory fitness in clinical practice: a case for fitness as a clinical vital sign. A scientific statement from the American Heart Association. Circulation 134(24):e653-e699. doi:10.1161/CIR.0000000000000461
• Shailendra P, et al. (2022). Resistance training and mortality risk: a systematic review and meta-analysis. American Journal of Preventive Medicine 63(2):277-285. doi:10.1016/j.amepre.2022.03.020
• Srikanthan P, Karlamangla AS (2014). Muscle mass index as a predictor of longevity in older adults. American Journal of Medicine 127(6):547-553. doi:10.1016/j.amjmed.2014.02.007
• Vagish E (2026). How apps calculate your biological age (WHOOP, Oura, Garmin and Dalia). Dalia Health blog. daliahealth.com
• Wiersema JM, et al. (2022). The association between continuous ambulatory heart rate, heart rate variability, and 24-h rhythms of heart rate with familial longevity and aging. Aging 14(18). doi:10.18632/aging.204219
• Windred DP, et al. (2024). Sleep regularity is a stronger predictor of mortality risk than sleep duration: a prospective cohort study. Sleep 47(1):zsad253. doi:10.1093/sleep/zsad253
• Xiong L, et al. (2023). The association of handgrip strength with all-cause and cardiovascular mortality: results from the National Health and Nutrition Examination Survey database prospective cohort study with propensity score matching. Frontiers in Nutrition 10:1183973. doi:10.3389/fnut.2023.1183973
• Zhang D, Shen X, Qi X (2016). Resting heart rate and all-cause and cardiovascular mortality in the general population: a meta-analysis. CMAJ 188(3):E53-E63. doi:10.1503/cmaj.150535
• Zhang Y, et al. (2026). Long-term resistance training with all-cause and cause-specific mortality: assessing dose-response and joint associations with aerobic physical activity. British Journal of Sports Medicine 60(12):874-883. doi:10.1136/bjsports-2025-110503