Lifestyle Factor

Magnitude of Impact

(Approximate All-Cause Mortality Hazard Ratio)

Notes on the Science

Aerobic Capacity

HR   Runner Group

5.04 Low vs Elite

3.90 Low vs High

2.75 Low vs Above Average

1.95 Low vs Below Average

2.59 Below Average vs Elite

2.00 Below Average vs High

1.41 Below Average vs Above Average

1.84 Above Average vs Elite

1.42 Above Average vs High

1.29 High vs Elite

Very High Confidence. Results consistently replicated in both observational studies and randomized trials (demonstrating cause-and-effect).

HRs from JAMA Cohort Study of 122,007 patients at tertiary care academic medical center from January 1, 1991, to December 31, 2014, with a median follow-up of 8.4 years.

Study Notes:

- Term Elite is misleading. Elite runners are considerably faster than the elite runners in this study.

- Study measured heart rates for running at different speeds on a treadmill to assign each participant to a runner group.

BMI

HR        BMI

1.48 15

1.31 16

1.15 17.5

NSS 20

NSS 22

1 (Ref) 23

1.01 24

1.03 25

1.11 27.5

1.24 30

1.42 32.5

1.66 35

1.98 37.5

2.37 40

2.88 42.5

3.54 45

Very High Confidence. Results consistently replicated.

HRs from a meta analysis of 230 cohort studies of 30,233,329 participants.

Study Notes:

- The HRs presented are for healthy never smokers. The study includes many other scenarios in its tables. In general, the higher your starting health risks, the larger the HR for increased BMI.

- NSS means there was no statistically significant HR found for that BMI category (BMIs 20-22).

- All of the Hazard Ratios are showing your all-cause mortality risk in comparison to having a BMI of 23 (the reference point).

- ~9% of the US has a BMI of 40+.

Smoking

2.29 Female Current Smoker

1.35 Female Former Smoker

2.24 Male Current Smoker

1.30 Male Former Smoker

1.6   Male Occasional Smoker

Very High Confidence. Results consistently replicated.

HRs from two studies:

- A study analyzing US Adults aged 35+ using 1990–2011 National Health Interview Survey Linked Mortality Files.

- A study analyzing Finnish Adult Health Behavior Surveys from 1978 to 1995.

Study Notes: No statistically significant results were found linking occasional female smokers to all-cause mortality.

Sugar

HR    Quintile  % calories from added sugar

1 (Ref)  Q1       7.4%

1.07      Q2       11.4%

1.18      Q3       14.8%

1.38      Q4       18.7%

2.03      Q5       25.2%

HR         % of calories from added sugar

1 (Ref)    0-10%

1.30        10-25%

2.75        25%+ (average of 28.7%)

High Confidence. Results consistently replicated but with smaller participation sizes than some other health studies.

HRs from a cohort study of a nationally representative sample of 42,880 US adults. The study shows HR broken out by 20% intervals (quintiles) of the population as well as by % of calories coming from sugar.

Study Notes:

 - This study looks at cardiovascular mortality alone (not all-cause mortality).

 - This study controls for any impact on body weight and a “healthy eating index score.”

Strength Training

1.41 to go from high grip strength to low grip strength

1.16 for each 5kg decrease in grip strength

Very High Confidence. Results consistently replicated.

HRs from a meta analysis of 42 cohort studies containing 3,002,203 participants.

Study Notes: Study tested grip strength, which is a good proxy for overall strength.

Relationships

1.26 for loneliness

1.29 for social isolation

1.32 for individuals living alone

High Confidence. Results consistently replicated but with smaller participation sizes than some other health studies.

HRs from a meta analysis of 70 studies with 48,673 global participants with an average age of 66.

Notes on the Science:

- There are conflicting results on whether actually being alone or simply feeling lonely matters more. Both are bad.

- The impact to health follows physiological stress pathways: increased blood pressure, faster “aging”, etc.

Mental Health

Cardiovascular Mortality only (not all-cause):

1.41 for anxiety

1.27 for high perceived stress

All-cause Mortality (being under 1 means a reduction in risk)

0.86 for being an optimist

Very High Confidence. Results consistently replicated. Typically a linear correlation found (i.e. the more stressed you are, the bigger the health risk).

HRs from:

- Anxiety: A meta analysis of 46 cohort studies containing 2,017,276 participants.

- Stress: A meta analysis of 6 cohort studies with 118,696 participants tracked for an average of 13.8 years.

- Optimism: A meta analysis of 15 studies with 229,391 participants tracked for an average of 13.8 years.

Therapy is proven to reduce your risks.

Religion / Spirituality

All-cause Mortality (being under 1 means a reduction in risk)

0.84 for being religious / spiritual

0.73 if you go to church weekly

High Confidence. Results consistently replicated. Although it’s unclear if the impact of religion and spirituality is separate from the benefits of strong relationships and improved mental health. 

HRs from a meta analysis of 91 cohort studies with 3224 participants. Similar HR for being religious found in a meta analysis of 126,000 people.

Sleep

1.12 for getting less than 6 hours of sleep

1.3 for getting more than 9 hours of sleep

High Confidence. Results consistently replicated but it’s likely that confounding variables contribute to the health risks of oversleeping (correlation and not causation).

HRs from a meta analysis of 27 cohort studies with 1,382,999 participants and a follow up range of 4 to 25 years.

Alcohol

All-Cause Mortality

Men:

NSS: < 3.2 drinks / day

1.15: 3.2 - 4.6 drinks / day

1.34: 4.6+ drinks / day

Women:

NSS: < 1.8 drinks / day

1.21: 1.8 - 3.2 drinks / day

1.34: 3.2 - 4.6 drinks / day

1.61: 4.6+ drinks / day

Heart Disease Risk Factors

Numbers below come from study’s supplemental material

- Systolic Blood Pressure (approximate increase in mmHg)

     - 4: 1 drink/day

     - 6: 2 drinks/day

     - 12: 3 drinks/day

- LDL (approximate increase in mg/dL)

     - 4: 1 drink/day

     - 9: 2 drinks/day

     - 16: 3 drinks/day

High Confidence. The science on alcohol consumption is nuanced but robust.

All-Cause Mortality HRs come from a meta analysis of 104 cohort studies with 4,838,825 participants. Data supported by another highly regarded cohort study of 371,463 individuals that used genes to randomize the population (a highly regarded practice to eliminate confounding variables). This study provided the Heart Disease Risk Factor metrics.

Notes on the science:

 - There are many studies that show a reduction in all-cause mortality for “moderate” amounts of drinking, but they typically suffer from confounding variables (a lot of people that don’t drink gave it up because they were unhealthy).

- These two highly regarded studies both found no statistically significant correlation to all-cause mortality for “moderate” amounts of drinking (see the HRs for the limits of “moderate”).

- Both of these studies found a steep increase in all-cause mortality and heart disease risk for each drink past “moderate” drinking levels, showing a nonlinear trend. I.e. Going from 4 to 5 drinks per day is far more risky than going from 3 to 4.

- One of these studies also looked for heart disease risk factors and found a correlation even at low to moderate levels of drinking.

- However, these studies didn’t control for alcohol type and thus don’t contradict the commonly found correlation between heart health benefits and small amounts (1 daily glass for women, 2 for men) of red wine.

- Any amount of alcohol consumption later in the day negatively impacts sleep quality, which has near term impacts on health and likely long term impacts on longevity.

Processed Foods

Umbrella Study

1.21 to go from the top 25% of ultra processed food consumers to the bottom 25%.

1.02 if you swap 10% of your calories to ultra processed foods.

Healthcare Professionals Study

1.04 if you go beyond 3 “servings” of ultra processed food / day.

High Confidence. Results consistently replicated but it's difficult to control for variables like BMI and overall diet quality since processed foods contribute to these variables.

HRs from two studies:

- An umbrella study that looked at 42 meta analyses of 9,888,373 participants. Some studies analyzed had controls for BMI and diet quality but not all.

- A cohort study with 114,064 healthcare professionals as participants. This study did control for BMI and diet quality.

Study Notes:

- The umbrella study suggests a nonlinear trend as you increase the amount of ultra processed food consumed. This is likely because it doesn’t control for overall diet quality, and at high levels of processed food consumption, there isn’t room left to eat more nutritious foods.

- The healthcare professional study found that overall diet quality had a bigger impact than the amount of processed foods consumed. Being in the top 25% of diet quality reduced your all-cause mortality risk by 14%, while being in the top 25% of processed food consumption only increased your all-cause mortality risk by 4%.

- The healthcare professionals study showed a nonlinear impact of increasing processed foods. Once you went above 3 servings / day, it didn’t seem to matter how much further you increased.

“Junk” food causes inflammation and impacts your metabolic health, regardless of whether or not you gain weight.

Diet Quality /

Whole Foods

0.70 if your diet goes from the bottom 20% of US adults to top 20%.

0.86 the HR of going from the bottom 25% of healthcare professionals’ diets to the top 25%.

Going from the “low-intake” group to the “high-intake” group

0.80 Nuts

0.81 - 0.89 Unsaturated vegetable oils

0.88 Whole grains

0.91 Fruits

0.93 Vegetables

0.95 Fish

0.96 Legumes

High Confidence. Results consistently replicated but a large variety of findings also exist. Nutrition studies are notoriously difficult.

HRs presented come from multiple studies:

- A study of 5525 US participants aged 40+, tracked for an average of 9.8 years.

- A cohort study with 114,064 healthcare professionals as participants.

- A meta analysis of 103 studies with hundreds of thousands of participants – broken out by the impact of various food groups on all-cause mortality.

Notes on Studies:

In the general US population study, there was no statistically significant impact of improving your diet until you reached the top 20%. The top 20% had about double the “healthy eating index” score as the bottom 20%.

The diet difference across the healthcare professionals in servings / day was:

- Whole Fruits: 2.1 v 1.3 (men) and 2.0 v 2.3 (women)

- Whole Vegetables: 4.4 v 3.2 (men) and 3.5 vs. 2.6 (women)

- Whole grains: 0.8 v 0.4 (men) and 0.5 v 0.3 (women)

- Nuts: 1.0 v 0.8 (men) and 0.6 v 0.5 (women)

The healthcare professionals study controlled for BMI.

Red Meat

General Population

1.3 to go from bottom 20% of red meat consumers to top 20%.

1.03 for adding 2 red meat servings per week.

Healthcare Professionals

1.13 for 1 unprocessed red meat serving per day.

1.2 for 1 processed red meat serving per day.

Very High Confidence. Results consistently replicated.

HRs are from:

- A cohort study that pooled the data of 29,682 participants across 6 cohort studies.

- A prospective study that followed 121,342 healthcare professionals (30% men age 40-75 and 70% women age 30-55).

Study Notes:

- The results are fairly consistent across study populations.

- The healthcare professional study had consistent diets across groups except for the red meat intake.

The source of meat is an incredibly important variable. It’s likely that grass-fed, organic red meat doesn’t carry the same health risks.

Mobility / Flexibility

1.87 for highly flexible vs. inflexible men

4.78 for highly flexible vs. inflexible women

Low Confidence. There is not a robust body of evidence to pull from.

HRs from a study that followed 3139 participants (66% men) aged 46-65 for an average of 12.9 years.

Study Notes:

- The results for females suffer from a small sample size with high variance. I don’t really trust those numbers.

Loss of functional movement has a major impact on quality of life and falls are a serious longevity concern for older individuals. While we don’t have a great all-cause mortality hazard ratio to reference, maintaining mobility while aging is an established health concern.

Air Pollution

1.12 (per 10 µg/m³ increase in PM2.5 or PM10)

High Confidence. Results consistently replicated.

HRs from a meta analysis of 71 articles studying locations across the Asia-Pacific region.

Air pollution exposure is dose dependent and highly variable based on where you live. For very poor air quality locations, the impact to health can be as large as chronic smoking.

Sun Exposure

HRs for Melanoma, not All-cause Mortality

0.86 the risk reduction of getting melanoma if you work outdoors instead of indoors.

1.71 the risk of getting melanoma if you go outdoors infrequently

2.0 the increase in risk of getting melanoma if you sunburn more than 5 times

High Confidence in the connection between sun exposure and skin cancer but no studies connecting sun exposure to all-cause mortality.

Takeaways from several related studies:

33% of Americans are Vitamin D deficient and would benefit from more exposure to sunlight. However, if this exposure results in a sunburn, it doubles your risk of getting skin cancer. Getting consistent sun exposure lowers your chances of getting skin cancer vs. infrequent exposure. Having a job where you work out in the sun actually decreases your risk of getting skin cancer.

While some studies suggest sun screen increases your risks of skin cancer, this meta-analysis of 21 studies and 23,434 participants finds no statistically significant connection between sunscreen use and Melanoma.

Sun exposure also carries mental and emotional health benefits and improves sleep quality.

Water

0.77

Medium Confidence. Some studies show a connection to all-cause mortality, but the body of evidence isn’t robust.

HR from a prospective study of 35,463 US adults tracked for an average of 88 months.

Environmental Carcinogens, Heavy Metals, Microplastics, etc.

N/A (No direct meta-analysis data for all-cause mortality)

Low Confidence. While specific high-dose exposures are harmful (e.g., lead, arsenic), there isn’t research that quantifies the effect of general low-level exposure on all-cause mortality. It’s likely beneficial to limit exposure, but that’s speculation at this point.

Supplements

(Creatine, etc.)

N/A (No direct mortality data)

Low Confidence. Creatine may indirectly support longevity by enhancing lean mass gains from resistance training. Other supplements lack enough evidence to make definitive claims.

"Life Hacks"

(Cold/Heat Therapy, Breath work, etc.)

N/A (No direct mortality data)

Low Confidence. Sauna usage has observational data supporting heart health benefits but no impact on all-cause mortality. Others lack direct all-cause mortality evidence.

Organic Produce

N/A (No direct morality data)

Medium Confidence. There are some studies showing health benefits of eating organic food but there are no connections to all-cause mortality.

Takeaways from several studies:

- Organic food may be a bit more nutrient dense and expose the consumer to less pesticides, but there isn’t much evidence showing this makes a substantial impact on mortality.

- A cohort study of 68,946 French adults (75% women) found a 25% reduction in cancer risk for those on a high organic diet vs. nonorganic.

- Here is a summary of all of the other documented health benefits.

Weed

N/A (No direct mortality data)

Low Confidence. We don’t have any data connecting THC consumption to mortality. This is not a well-studied subject.

What we do know:

- The near-term impacts aren’t great.

- Smoking weed has negative impacts on our lungs.

- We also know that chronic weed consumption worsens sleep quality.

- There are also complex connections between weed and mental health, including potential connections to schizophrenia and bipolar disorder.

Salt

No clear consensus

Low Confidence. There is no scientific consensus on the impact of salt consumption on all-cause mortality. However, increased sodium consumption (if not matched with an increase in water or other electrolyte consumption) is linked to an increase in blood pressure.

- This cohort study of 7154 US adults found a reduction in all-cause mortality for participants that consume more than the American guidelines of 2300mg/day (HR of 1.2).

- A randomized controlled trial with 6100 participants found trends between increased sodium intake and all-cause mortality, but no statistically significant correlation between the groups. A statistically significant correlation was found for increasing sodium intake by 1g per day (HR of 1.12).

- A randomized controlled trial of 416 participants found reducing the sodium intake of white individuals without hypertension from 3.5g to 2.3g reduced systolic blood pressure (SBP) by 2.1 mmHg and reducing from 3.5g to 1.1g reduced SBP by 4.6 mmHg.

- A cohort study of 13,855 participants suggests the important metric is actually the ratio of sodium to potassium intake.

These studies don’t do a great job of controlling for physical activity / sweating, which further complicates salt guidelines.

Lifestyle Factor

My Assessment

Physical Activity

I’m already in the “elite” runner category of this group. There are likely some small benefits from improving my aerobic capacity, but what’s more important is that I don’t lose capacity. By continuing to train, I assume that I’m preventing myself from sliding back to the “Below Average” category.

HR 2.59

BMI

I have a healthy BMI and won’t let this change. HR 1.

Smoking

I don’t smoke and won’t smoke. HR 1.

Sugar

I have tracked my food consumption in the past, and added sugar was around 10%. I think I’ve let this creep up in recent years. I should track my diet again, but in the absence of doing that, I’ll assume my intake is somewhere just below the average American’s. This gives me an HR of 1.07-1.18. I’ll call it 1.12 for now.

Strength Training

I’ve still got plenty of room to improve here and I’m avoiding a drop in performance by continuing to lift. HR of ~1.3.

Relationships

I feel I’ve got very strong relationships. HR 1.

Mental Health

I consider myself to be on the higher end of the population regarding perceived stress and anxiety. HR ~1.3

Religion / Spirituality

This is a tough one to manipulate. HR 1.

Sleep

I regularly get 7-9 hours of sleep. HR 1.

Alcohol

This is an interesting one. I drink less than 3 drinks / day, so I won’t see an improvement in my all-cause mortality HR by reducing any further. But I am pre-hypertensive, and a reduction in blood pressure from giving up drinking could potentially eliminate this. Plus I can definitely feel the impact alcohol has on my sleep.

On the other hand, alcohol has been an important role in solidifying many of my relationships. I think the key for my life could be more intentionality with each drink.

I’ll assign an HR of 1.1.

Processed Foods

I think my diet is pretty close to the diet used in the healthcare professionals study. I do eat too much processed foods (mainly on the weekends), but I think my diet quality blunts most of the impact.

HR 1.04

Whole Foods

I do pretty well here, but it takes conscious daily effort. If I don’t keep this focus, I’ll increase my risk. Currently eat lots of nuts, fruits, vegetables. Will consciously increase fish intake and whole grains based on this data.

HR 1.1

Red Meat

I already made a conscious decision to substitute poultry for beef in the majority of my diet. I could further improve by only eating red meat when it’s organic / grass-fed.

HR 1.03

Mobility / Flexibility

I do well here. I need to continue to “use it” so I don’t “lose it”, but this is not a major risk for me right now. HR 1

Air Pollution

My location has annual air quality averages under the WHO guidelines, and I pay attention to the Air Quality Index. HR 1

Sun Exposure

I live in an area where sun exposure is a major risk, and I don’t work outdoors. I get pretty consistent sunlight exposure through running year-round and avoid sunburns pretty effectively.

HR 1

Water

I would benefit from paying more attention to this. Could help my blood pressure. HR 1.1

Environmental Carcinogens, Heavy Metals, Microplastics, etc.

I’ll take steps to reduce exposure when it’s convenient but not worry when it’s difficult. HR 1.

Supplements (Creatine, etc.)

No HR.

"Life Hacks" (Cold/Heat Therapy, Breathwork, etc.)

No HR.

Eat Organic Produce

I’ll try to eat organic when given the option but won’t think much about it when it isn’t an option. No HR.

Weed

No HR.

Salt

I think increasing my potassium intake would benefit me here. While there’s no clear HR, my prehypertension means I’ll assign some risk to getting this right. Since most of my salt intake comes from processed foods, HR 1.03.