Self vs Dependent Health Load

Loop tests employees far more often than it tests the people on their policies. Of the annual-checkup reports with a usable HbA1c result, roughly 63% are the employee's own and 37% belong to a spouse, child, or parent — even though the claims book shows dependents are two-thirds of who the policy actually pays for. The dependent layer is under-screened and over-claimed at the same time. That asymmetry is the spine of this page: the biomarker file is a window into a population the policy mostly does not look at until it ends up in a hospital.

The headline numbers invite a simple reading. Dependents who do get a blood panel look much worse than employees on the marquee metabolic marker — 51.4% in the glucose-dysfunction band against 31.3%. The simple reading is that dependents are sicker. The data supports a sharper, more useful version of that claim, and corrects part of it.

The dependent biomarker pool is older, and on glucose that is almost the whole story#

Of dependents Loop ran an HbA1c on, 51.4% sit in the glucose-dysfunction band against 31.3% of employees — a 20-point gap. The instinct is to read that as a metabolic difference between the two groups. It is mostly an age difference between them.

The dependent pool that shows up for a blood test is not a cross-section of dependents. It is heavily weighted toward parents — the layer most likely to be sent for a checkup precisely because someone is already worried about them.

Hold age constant and the glucose gap collapses.

Through the working-age bands the two lines run together. A dependent in their 30s is no more likely to show glucose dysfunction than an employee in their 30s. The headline 20-point gap exists because the dependent group tested skews two decades older. Only at 60-plus does a real same-age dependent excess open up, and that band is small for employees by construction. The Blood Sugar page reads the dependent layer's 51.4% from the metabolic side, where it anchors the report's dependent-risk thesis.

Anemia is the one dependent disadvantage that age cannot explain#

Glucose equalises within age band. Anemia does not. Dependents run more anemic than employees at every working-age band, by a margin that holds whether the dependent is 25 or 55.

The gender split rules out the obvious confound. Maternal-age iron loss would predict a female-driven gap; instead the dependent excess is sharpest among men.

A male employee in this dataset is anemic 5.5% of the time. A male dependent is anemic 21.7% of the time — close to four times as often. Age does not explain it; the table above holds age flat and the gap persists. Gender does not explain it; the gap is largest within men. What is left is the difference between a population that gets an annual checkup and one that mostly does not. The Anemia page carries the full marker; this is the one dependent disadvantage that survives every control the biomarker file allows.

On three markers, the employee is the sicker one#

The thesis that dependents are uniformly worse off does not survive contact with the full marker set. On three of the seven markers tested, the working-age employee is the one in trouble.

Low HDL, low B12, and raised liver enzymes all peak in the working-age employee. Age-controlled, the pattern sharpens: in the 30s, employees show low HDL at 52.1% against dependents' 41.0%, and raised liver enzymes at 23.7% against 12.9%. These are the markers of the desk-bound urban professional — the sedentary day, the late dinners, the drink after work, the vegetarian diet thin on B12. They are conditions the employee's own age band produces, not conditions the dependent layer is spared by youth.

Dependents are two-thirds of who the policy pays — and more than twice the cost#

The biomarker file looks at employees. The claims file is paid for by everyone else. Where Loop tests employees roughly two to one over dependents, the claims book runs the other way: two-thirds of what the policy pays out is for people other than the employee.

The dependent layer costs the policy ₹217.4 Cr against the employee's ₹103.1 Cr — more than two to one. The parent layer alone, at ₹118.2 Cr, exceeds the entire employee block. And the parent claim is the most expensive on average, at ₹53,981, because it concentrates in the cardiac admissions, fracture surgeries, and oncology that arrive in the 60s and 70s — the same age band where the glucose curve finally separates self from dependent.

The two datasets meet at the same conclusion from opposite directions. The biomarker file shows a dependent population that is under-tested and, where tested, carries an age-independent anemia burden and an age-driven metabolic one. The claims file shows that same population absorbing two-thirds of the bill, concentrated in conditions that build quietly for years before they are admitted. The screening gap and the cost concentration are the same fact seen at two ends of the disease timeline.

For the deeper claims-side breakdown of the dependent layer — spouse, child, and parent as clinically distinct populations, with the gender and account-tier cuts — see Who the Policy Pays For. The Parent Claims page goes further into the costliest sub-layer, the ₹118.2 Cr the policy spends on the workforce's parents. This page's contribution is the biomarker half: the evidence that the dependent cost is foreshadowed in screening data the policy already could be collecting.

The dependent layer is the cleanest preventive-investment case in the dataset#

Three facts the 2026 data establishes together: