What Actuaries Think About Contactless Health Data

Actuaries evaluating contactless health data perspectives in 2026 confront a fundamental modeling question: can physiological signals captured without physical contact--through smartphone cameras, wearable sensors, and remote photoplethysmography--provide the actuarial credibility needed to replace or augment the fluid-verified data that mortality tables have relied upon for decades? The answer emerging from peer-reviewed research and early industry adoption is nuanced but directionally clear. Contactless biometric data introduces new predictive variables that improve mortality and morbidity risk differentiation, but it also introduces new sources of measurement uncertainty that actuarial frameworks must accommodate. For actuarial teams at carriers, reinsurers, and consulting firms, this is not a technology adoption question--it is a credibility and calibration question with direct implications for pricing, reserving, and experience analysis.

A study of 407,569 UK Biobank participants found that five basic physical measures--including resting heart rate and walking pace--could enhance or replace traditional clinical risk predictors in estimating premature death likelihood, according to RGA and University of Leicester research published in Mayo Clinic Proceedings.

The Actuarial Case for Contactless Health Data

The actuarial profession's relationship with data is utilitarian: a data source is valuable to the extent that it improves the discrimination between risk classes in a manner that is statistically credible, actuarially sound, and regulatorily defensible. Contactless health data must meet all three criteria.

Statistical credibility. The evidence base is substantial and growing. The RGA-University of Leicester UK Biobank study demonstrated that non-traditional biometric factors--resting heart rate, walking pace, grip strength--"dramatically improved the ability to differentiate mortality and morbidity risks" when added to traditional underwriting variables. A separate PMC study (2014) found that "a summary measure of biomarkers improves mortality prediction compared with self-reports alone," and that individual biomarkers outperformed composite scores. These are not marginal improvements; they represent meaningful increases in the predictive power of mortality models.

Actuarial soundness. The challenge lies in integration. Mortality tables and pricing models are calibrated on populations that were assessed through specific underwriting protocols--typically including fluid collection, paramedical exams, and APS review. Introducing contactless biometric variables requires actuaries to either recalibrate existing models or develop parallel pricing frameworks for digitally underwritten cohorts. Munich Re's Biometric Portfolio Analysis, built on 15+ years of data from more than 30 participating insurers, provides one of the few industry-scale datasets for this calibration work.

Regulatory defensibility. The NAIC's Model Bulletin on AI use in insurance, adopted by 23 states and Washington, D.C. by late 2025, requires that algorithmic underwriting decisions be explainable and non-discriminatory. For actuaries, this means that any contactless health variable incorporated into a pricing model must be demonstrably relevant to mortality or morbidity risk and free from unfair discrimination--a standard that objective physiological measurements are well-positioned to meet relative to behavioral or socioeconomic proxies.

How Contactless Health Data Compares to Traditional Actuarial Inputs

Actuarial Input	Source	Objectivity	Longitudinal Value	Mortality Prediction Strength	Actuarial Experience Base
Blood panel (lipids, glucose, A1c)	Lab draw	Very high	Snapshot only	Very high (decades of data)	Deep--embedded in mortality tables
Cotinine / nicotine test	Lab draw or saliva	Very high	Snapshot only	High (tobacco is #1 modifiable risk)	Deep--standard rating factor
Resting heart rate (rPPG)	Smartphone camera	High	Repeatable; periodic capture	High (UK Biobank: significant mortality predictor)	Emerging--limited carrier experience
Heart rate variability (rPPG/wearable)	Smartphone or wearable	High	Continuous via wearable	High (autonomic function marker)	Early--research-stage for pricing
Blood pressure estimate (rPPG)	Smartphone camera	Moderate	Repeatable	High (established cardiovascular risk factor)	Early--algorithm-dependent measurement
Respiratory rate (rPPG)	Smartphone camera	High	Repeatable	Moderate (pulmonary health signal)	Minimal--not yet priced
Physical activity (wearable)	Wearable device	High	Continuous	Moderate-high (UK Biobank: walking pace significant)	Early--vitality programs generating data
Self-reported health status	Questionnaire	Low (7%+ nondisclosure)	None	Low-moderate (subject to bias)	Deep--but accuracy is deteriorating

Applications in Actuarial Practice

Mortality Model Enhancement

The most immediate actuarial application is incorporating contactless biometric variables into mortality models as supplementary predictors. The traditional actuarial approach segments populations by age, sex, tobacco status, build, and medical history. Contactless data introduces physiological variables that cut across these segments.

Consider resting heart rate. The UK Biobank data demonstrated that this single variable--capturable in 30 seconds via a smartphone camera--provides significant mortality risk differentiation independent of age, sex, and known medical conditions. A PLOS One study using the same dataset further documented age-, sex-, and disease-specific associations between resting heart rate and cardiovascular mortality. For actuaries, this represents a variable that is inexpensive to collect, objective, repeatable, and empirically validated at population scale.

The practical challenge is credibility weighting. How much actuarial weight should a contactless biometric measurement carry relative to decades of fluid-verified experience? The answer depends on the quality of the measurement, the size of the reference population, and the length of the observation period. As carriers accumulate experience data on digitally underwritten cohorts, the credibility assigned to contactless variables will increase--but actuaries rightly demand demonstrated mortality experience before adjusting pricing.

Pricing for Accelerated and Instant-Issue Products

Gen Re's 2024 survey found that 57% of individual life applications are now eligible for accelerated underwriting processing, and 82% of carriers have implemented accelerated workflows. For actuaries pricing these products, a core question is whether the absence of fluid verification introduces systematic adverse selection that must be loaded into premiums.

The industry data suggests it does. Tobacco nondisclosure rates rose from 2.0% to 3.5% of applicants between 2015 and 2022, and CRL Corp estimates the mortality cost at $4 billion annually. Munich Re documented that medical misrepresentation ranks as insurers' top fraud concern at 4.0 on a 5-point severity scale.

Contactless health data offers actuaries a tool to reduce these loads. If rPPG-derived heart rate and HRV can identify a portion of misrepresenting applicants--or more precisely, can improve the separation between standard and substandard risks within the accelerated cohort--then the adverse selection load can be reduced, making accelerated products more price-competitive without increasing mortality risk.

Experience Analysis and Assumption Setting

Actuarial assumption setting depends on credible experience analysis. The challenge with contactless health data is that most carriers have limited mortality experience on cohorts underwritten with these tools. The data that does exist is encouraging:

Munich Re's Biometric Portfolio Analysis provides multivariate and machine learning analyses that enable actuaries to "gain deep insights into their own datasets and derive new actuarial findings." This platform--built on data from 30+ insurers over 15+ years--represents the most comprehensive industry resource for benchmarking contactless data against traditional mortality experience.

LIMRA's market data shows that individual life insurance premiums reached record levels in 2025, with accelerated underwriting cited as a key driver. As the volume of digitally underwritten business grows, the experience base for actuarial analysis will expand correspondingly. The profession's standard credibility frameworks--limited fluctuation, Buhlmann, and Bayesian approaches--provide established methodologies for blending emerging contactless data experience with traditional mortality assumptions.

Research Informing the Actuarial Perspective

The peer-reviewed research that actuaries weigh most heavily when evaluating contactless health data includes:

• RGA and University of Leicester (UK Biobank, published in Mayo Clinic Proceedings). The foundational study for actuarial evaluation of contactless biometric variables. Analysis of 407,569 participants demonstrated that resting heart rate, walking pace, and grip strength significantly enhanced mortality prediction when added to traditional risk variables. The population scale and longitudinal design give this study substantial actuarial credibility.

• PMC (2014) -- Biomarkers vs. self-reports. Found that objective biomarker measurement improves mortality prediction compared with self-reports alone. Critically for actuaries, the study demonstrated that individual biomarkers outperform composite scores--suggesting that even a single contactless measurement (e.g., resting heart rate) can add meaningful predictive value.

• Journal of Insurance Medicine (2013) -- Laboratory predictors of cancer mortality. Demonstrated that biometric measurements including liver function tests, cholesterol, and blood pressure were independently predictive of cancer mortality in insured populations. While some of these measurements require contact, the study establishes the actuarial principle that objective physiological data outperforms self-report for cause-specific mortality prediction.

• Frontiers in Digital Health (2025) -- rPPG systematic review. Documented the current state of remote photoplethysmography technology, including well-established outputs (heart rate, respiratory rate, HRV) and emerging capabilities (blood pressure estimation, stress detection). For actuaries, this review maps the technological frontier of what contactless measurement can reliably capture.

• PLOS One -- Resting heart rate and cardiovascular mortality (UK Biobank). Provided detailed age-, sex-, and disease-stratified analysis of the relationship between resting heart rate and cardiovascular mortality--exactly the kind of granular segmentation that actuaries need for pricing and reserving.

The Future of Contactless Data in Actuarial Work

Three developments will determine how quickly contactless health data moves from experimental to standard actuarial practice:

Experience data accumulation. The actuarial profession is empirical by nature. Until carriers have 5-10 years of mortality experience on cohorts underwritten with contactless biometric tools, pricing actuaries will apply conservatism--either through explicit adverse selection loads or through blending techniques that weight traditional experience more heavily. The pace of adoption--82% of carriers now using accelerated underwriting--suggests that meaningful experience data will emerge within the next 3-5 years.

Standardization of measurement protocols. Actuarial credibility requires consistency. If different carriers capture resting heart rate using different rPPG algorithms under different conditions, the resulting data cannot be pooled for industry-level experience analysis. The development of standardized measurement protocols--analogous to the standardized lab panels used in traditional underwriting--is a prerequisite for actuarial adoption at scale.

Regulatory alignment on acceptable variables. The NAIC's AI Model Bulletin provides a framework, but actuaries need specific regulatory guidance on which contactless biometric variables are permissible rating factors. Objective physiological measurements have a strong case for regulatory acceptance--they measure health status directly, are not proxies for protected characteristics, and are applied uniformly across applicants. However, state-level variations in biometric privacy law and insurance regulation create a fragmented compliance landscape that actuaries must navigate.

The Society of Actuaries and the American Academy of Actuaries have both begun publishing research and practice notes addressing the integration of non-traditional data sources into actuarial models. These professional resources, combined with the growing empirical base, suggest that the actuarial profession is moving toward structured engagement with contactless health data rather than resistance to it.

Frequently Asked Questions

How do actuaries assess the credibility of contactless health data?

Actuaries apply standard credibility theory--evaluating the volume, consistency, and relevance of data before incorporating it into models. For contactless health data, credibility currently depends on the specific variable: resting heart rate has strong population-scale evidence (UK Biobank, 407,569 participants), while blood pressure estimation via rPPG has less established actuarial credibility. Actuaries typically assign partial credibility to emerging data sources and blend them with traditional experience using Buhlmann or Bayesian frameworks.

Will contactless health data change how life insurance is priced?

Over time, yes. As experience data accumulates on digitally underwritten cohorts, actuaries will be able to differentiate pricing between policies underwritten with objective biometric data and those relying solely on self-report. The direction is toward rewarding data quality: carriers and reinsurers that can demonstrate objective verification of health status at the point of underwriting will be able to price more competitively than those relying on questionnaires alone.

What is the biggest actuarial concern about contactless biometric measurement?

Measurement uncertainty. Unlike a lab panel that produces precise, standardized values, contactless measurements are influenced by environmental factors (lighting for rPPG, device quality for wearables), algorithmic variation, and physiological confounders. Actuaries need to understand the error distributions of these measurements and incorporate appropriate margins into their models. The Frontiers in Digital Health (2025) review acknowledged remaining challenges with motion artifacts and blood pressure estimation accuracy--exactly the kind of limitations actuaries must quantify.

How are reinsurance actuaries evaluating contactless health data?

Reinsurance actuaries are evaluating contactless data through two lenses: treaty pricing and portfolio monitoring. For treaty pricing, the question is whether business underwritten with contactless biometric data warrants different terms than questionnaire-only business. Swiss Re's Assessment Engine (30,000+ rules, 2,322 risk factors) and Munich Re's Biometric Portfolio Analysis provide the analytical infrastructure for this differentiation. For portfolio monitoring, reinsurance actuaries track mortality experience on digitally underwritten cohorts to detect emerging adverse selection trends.

---

The actuarial profession has always followed the data--and the data on contactless health measurement is now reaching the scale and rigor that actuarial standards demand. The transition from theoretical interest to practical integration is underway. For actuarial teams evaluating how contactless biometric signals fit into their mortality models and pricing frameworks, explore how health-scan technology is generating actuarially relevant physiological data.