How to choose an at-home longevity blood test

A panel comes back with forty markers, thirty-eight of them inside the reference range, and the cover note says “no action needed.” Buried in it: an hs-CRP of 2.8 mg/L and a fasting insulin nobody ordered, because the panel ran a complete blood count and a metabolic screen and called that a longevity test. It wasn’t. It was thirty-eight numbers that didn’t move and two that mattered, one of them missing. If you’re comparing at-home longevity tests right now, this is the trap to shop your way out of: a long list of in-range noise dressed up as thoroughness.

A longevity panel has a different job from a regular blood test. A regular test asks whether you’re sick today. A longevity panel asks where you’re heading, tracking the trajectory of your cardiometabolic markers years before they drift out of range. That changes how you judge it. You don’t judge it on how many markers it counts. You judge it on whether it carries the few that move first, and whether it tracks them long enough to show a direction.

What a longevity panel is actually for

Marker count is the wrong thing to shop on, and it’s exactly what the vanity panels sell. Forty markers sounds rigorous. Most of that forty is a complete blood count and an electrolyte screen, and for a healthy person those sit flat inside their bands year after year, telling you nothing about your trajectory.

Six markers do most of the work: ApoB, Lp(a), fasting insulin, hs-CRP, HbA1c, ferritin. These are the ones that drift early, predict where your cardiometabolic risk is going, and respond to what you actually change. A panel that carries those six and reads them well beats a forty-marker printout that buries them.

The other failure isn’t the blood at all. It’s what happens after. Most longevity tests hand you a PDF of green checkmarks against a population range and stop there, which is the part you’re really buying and the part most providers skip.

Six longevity markers to consider

ApoB is the cardiovascular marker to anchor on, and the first thing to check a panel carries. Every atherogenic particle, whether LDL, VLDL, IDL, or Lp(a), carries exactly one ApoB protein, so ApoB is a direct count of the particles that lodge in an artery wall, not the cholesterol cargo inside them. The mechanism, and why your LDL-C can read perfectly normal while your ApoB climbs, is the full story of how the particle count beats the cargo count. What matters for buying: the lab’s “normal” band runs wide, into the 90s and past 100 mg/dL, while the optimal target for someone playing a long game sits lower, nearer 60 to 80. A panel that flags only against the wide band is reading the wrong line.

Lp(a) is the buy-once number. It’s largely genetic, set near birth, and barely moves across your life, so you measure it a single time and file it. It’s also a risk multiplier a large fraction of people carry and are almost never told about, because it’s not on a standard order sheet. A longevity panel that omits Lp(a) is missing the one marker you only ever have to pay for once. Both Lp(a) and the next marker are the kind a fifteen-minute visit skips, for reasons that have nothing to do with whether they’re worth knowing.

Fasting insulin flags insulin resistance years before glucose drifts. It’s cheap, it’s almost never ordered, and it can sit elevated for a decade while your fasting glucose reads flawless, because your pancreas is quietly working overtime to hold that glucose steady. By the time the glucose finally climbs, the compensation that kept it down went unmeasured the whole way. A panel without fasting insulin can’t see that decade.

hs-CRP is the cheapest window on low-grade inflammation, and the noisiest. A cold, a hard leg day, a bad night’s sleep all bump it, so a single reading is close to meaningless. It’s useful only as a trend across draws taken away from acute illness: the 3.0 mg/L line marks elevated risk, with optimal under roughly 1.0. This is the clearest argument for buying a panel that repeats rather than a one-off kit. One hs-CRP tells you almost nothing about the body underneath the cold.

HbA1c averages roughly three months of glucose into one number, which is its strength and its blind spot. Two people at the same 5.4% can have different daily patterns underneath, and the single number can’t separate them. Other measurements may add context, but a clinician should interpret the result alongside your history and decide whether more testing is appropriate.

Ferritin is iron status, and it’s propped up as readily by inflammation as by iron stores. A reassuring 180 ng/mL can sit on top of a smoldering hs-CRP while transferrin saturation runs low near 16%, which means you can read “plenty of iron” off a number that’s actually masking a shortfall. Ferritin is the cleanest example of why you’re buying interpretation, not just the figure: read next to hs-CRP it means one thing, read alone it can mean the opposite. This is the whole case for why a marker that reads “normal” can be hiding what props it up.

What separates a good at-home panel from a basic one

Four things move a panel from a basic screen to a real longevity test.

First, advanced markers. Does it actually carry ApoB, Lp(a), fasting insulin, and hs-CRP, or just the standard lipid, CBC, and metabolic screen with a longevity label on the box? Read the marker list before anything else.

Second, a real venous draw at home. A phlebotomist comes to your door and draws from a vein, which is what the advanced assays need to run reliably. A fingerprick kit can’t reliably deliver the same panel, and fasting timing matters: fasting insulin and glucose are only interpretable on a clean overnight fast, which a scheduled at-home draw makes easy to control.

Third, trends over time, not a single PDF. The panel should re-draw and show you slopes, because one hs-CRP is noise and one HbA1c is a three-month snapshot with no direction. The whole longevity case rests on the slope, and a one-off kit can’t draw one.

Fourth, interpretation against your own baseline, not just flags against a population range. A reference range is the band roughly 95% of a reference population falls inside, which describes the average human, not a thriving one. Your own line over time is the only reference that tells you whether you’re improving.

A short checklist to take to any provider:

• Does it carry ApoB, Lp(a), fasting insulin, and hs-CRP, not just the standard screen?
• Is it a venous draw at home, with controlled fasting, not a fingerprick?
• Does it re-draw on a schedule so you see trends?
• Does it read each marker against your baseline, not only a population band?
• Does it read the markers against each other, ferritin next to hs-CRP, HbA1c next to glucose?

The part most at-home tests skip: reading blood against your wearables

The marker is half a sentence. The meaning is in the second signal, and this is the part no fingerprick kit or one-off lab can do.

Take an HbA1c of 5.5%. Two people, identical number. Read the first against a continuous glucose trace and the line is flat all day, post-meal bumps that settle in an hour. The second is the same 5.5%, but the CGM shows lunch driving glucose to 160 mg/dL and a crash by mid-afternoon, the spikes and dips averaging out to the same tidy figure. Same blood, opposite bodies, different action. One needs nothing. The other needs to look hard at lunch. The HbA1c alone can’t tell you which person you are.

Now fasting insulin, creeping up two draws running. The blood says “up.” Read against a sleep log showing late, carb-heavy snacks three or four nights a week, and the two signals together say why, and what to change: the number is the symptom, the snack timing is the cause. The panel flags the drift. The blood plus the wearable names the lever you can actually pull tonight.

Apple Health is Depth’s only current health-data connection. Depth reads user-authorized fitness and wellness data from Apple Health and can place those records beside results you add to the app. Records from another app or device may be available when that source writes them to Apple Health; Depth does not directly communicate with, control, or certify the device. A slope on two or three draws is still a thin line. Discuss a surprising result with a qualified clinician, who can decide whether and when to repeat it.

How to choose, in one checklist

The decision rule, short enough to carry: pick the panel that carries the six markers, especially ApoB, Lp(a), and fasting insulin; runs a real venous draw at home on a clean fast; re-draws so you see trends; and interprets each marker against your own baseline and your wearables, not just a population band.

The honest caveat: if you only ever want one snapshot and a clean standard panel, a basic kit is a reasonable cheap bet, and you don’t need any of this. The longevity case is for people who want the slope, the direction the body is heading while every individual reading still reads normal. If you want that, the one-off kit will never give it to you, no matter how many markers it lists.

Where the service is available, you can use Depth to request an at-home collection from an independent partner. The partner confirms location, availability, and appointment details, performs the collection and testing, and issues the clinical result. Depth is not a laboratory, phlebotomy provider, or diagnostic provider. In the app, Depth reads user-authorized fitness and wellness data from Apple Health; it does not directly integrate with health hardware.