An assessment of intrinsic capacity from midlife to early old age in the 1958 British birth cohort

Using data from the 1958 British birth cohort, this study validates a five-domain intrinsic capacity construct that can be reliably measured and tracked from midlife to early old age, revealing a measurable decline in functional health between ages 50 and 62 before clinical symptoms typically appear.

The Gist

Imagine your body and mind as a high-performance car. For a long time, doctors and scientists have mostly checked this car's engine only when it starts making loud noises or breaking down in old age. They looked for specific broken parts (diseases) rather than checking the overall health of the engine itself.

This paper introduces a new way of looking at the car: Intrinsic Capacity (IC). Think of IC as the car's total "fuel reserve" and "engine power" combined. It's not just about whether the car has a flat tire (a disease); it's about how well the engine, brakes, sensors, and battery are working together to keep the car moving smoothly.

Here is what the researchers did, explained simply:

1. The Big Question: Can we check the engine before it breaks?

Most studies on this "total engine power" (Intrinsic Capacity) only look at people who are already 60 or older. The researchers asked: Can we measure this power earlier, say in your 50s, before you even feel old?

They wanted to know if this "power meter" works consistently from age 50 to age 62. If the meter changes its rules between age 50 and 62, you can't compare the scores. It would be like using a ruler made of rubber for a 50-year-old and a steel ruler for a 62-year-old; the numbers wouldn't tell you if the car actually shrank or if the ruler just changed.

2. The Experiment: The "1958 Class"

The researchers used data from a massive group of people born in the UK in 1958 (the "1958 British birth cohort"). They looked at these people twice:

• Time A: When they were 50 years old (midlife).
• Time B: When they were 62 years old (early old age).

They gathered 30 different "checklist items" to measure five specific areas of the car:

• Sensors: Hearing and vision.
• Brain: Memory and thinking speed.
• Chassis: Physical strength and ability to move (walking, climbing stairs).
• Mood: Emotional stability and mental health.
• Battery: Energy levels and sleep quality.

3. The Results: The Meter Works!

The researchers built a complex mathematical model (a "super-ruler") to combine all 30 items into one single score.

• The Ruler is Solid: They proved that this ruler measures the same thing at age 50 as it does at age 62. You can compare the scores directly.
• The Score Drops: As expected, the average "power score" went down between age 50 and 62. On a scale of 0 to 100, the average person lost about 6.3 points.
• Different Parts Wear Differently: Not all parts of the car wore out at the same speed.
  • The Physical part (muscles, walking) showed the biggest drop.
  • The Mood and Brain parts stayed relatively stable.
  • The Sensors (hearing/vision) showed the most variability—some people's sensors stayed sharp, while others declined quickly.

4. Does the Score Make Sense? (Validation)

To make sure their "power score" wasn't just a made-up number, they checked it against real-life health:

• Self-Rating: People who said, "I feel great," had high power scores. People who said, "I feel terrible," had low scores.
• The "Hidden" Decline: Interestingly, some people who said they felt "Excellent" still had low power scores. This suggests the new meter can spot subtle declines that a person might not even feel yet.
• Disease: People with more chronic conditions (like diabetes or back pain) had lower power scores, but the score captured more than just the diseases; it captured the function of the person.

5. The Main Takeaway

This study proves that we don't have to wait until people are 65 or 70 to start measuring their "aging engine." We can start checking the Intrinsic Capacity in midlife (age 50s).

The study shows that:

1. We can measure this capacity reliably in people in their 50s.
2. We can track how it changes over time without the measurement tool breaking.
3. By age 62, a measurable decline has already happened, especially in physical strength.

In short: The "Intrinsic Capacity" meter is a valid tool that works for people in their 50s and 60s, allowing us to see the early, subtle signs of aging before they turn into major health problems. It's like checking your car's oil and tire pressure every year, rather than waiting for the engine to smoke.

Technical Summary

1. Problem Statement

• Context: The World Health Organization (WHO) defines "Intrinsic Capacity" (IC) as the composite of all physical and mental capacities an individual can draw upon, comprising five domains: sensory, cognitive, locomotor, psychological, and vitality.
• Gap: Existing evidence on IC is predominantly cross-sectional and focused on populations aged 60 and older. There is limited empirical evidence regarding whether IC can be validly measured as a coherent, multidomain construct in midlife (ages 50–65), a critical period where functional decline often begins subclinically.
• Challenge: To track functional change over time, it is essential to establish longitudinal measurement invariance. Without this, observed score differences between time points could reflect changes in measurement properties rather than true functional decline. Previous studies have rarely tested this in midlife populations.

2. Methodology

• Data Source: The study utilized data from the 1958 British Birth Cohort (National Child Development Study; NCDS), specifically comparing two sweeps:
  • Sweep 8 (S8): Participants aged 50–51 (2008–2009).
  • Sweep 10 (S10): Participants aged 60–65 (2020–2024).
• Sample: The final analytic sample consisted of 7,804 participants at S10, with 7,132 having data at both sweeps for longitudinal analysis.
• Indicator Selection: Researchers selected 30 indicators across the five WHO domains based on availability and comparability across sweeps:
  • Sensory: Hearing and vision difficulties (2 indicators).
  • Cognitive: Word recall, letter cancellation, semantic fluency (4 indicators).
  • Physical: SF-36 Physical Functioning subscale (10 indicators).
  • Psychological: Malaise Inventory (9 indicators).
  • Vitality: Energy levels and sleep-related items (5 indicators).
• Statistical Approach:
  • Measurement Modeling: Three competing models were tested using Confirmatory Factor Analysis (CFA) with the Weighted Least Squares Mean- and Variance-adjusted (WLSMV) estimator:
    1. Correlated five-factor model.
    2. Second-order CFA (a general IC factor influencing five domain factors).
    3. Bifactor model.
  • Measurement Invariance: A three-step sequence (configural, metric, scalar) was applied to test invariance across sex and time (S8 vs. S10). Scalar invariance is required to compare latent means over time.
  • Validity Testing: Construct validity was assessed via associations with self-rated health, multimorbidity, long-standing illness, and living arrangements. Known-groups validity was tested via latent mean comparisons between sexes.
  • Longitudinal Change: Mean IC scores were compared between S8 and S10 using paired t-tests on factor scores.

3. Key Contributions

• Extension to Midlife: This is one of the first studies to operationalize and validate the WHO IC framework in a midlife-to-early-old-age transition (ages 50–65), moving beyond the traditional focus on older adults.
• Psychometric Rigor: The study establishes longitudinal scalar invariance, proving that IC scores can be meaningfully compared across a 12-year period in this age group. This allows researchers to distinguish genuine functional decline from measurement artifacts.
• Model Selection: The study provides empirical evidence favoring a second-order hierarchical structure over bifactor or correlated five-factor models for IC in this demographic, aligning with the theoretical WHO framework.

4. Key Results

• Measurement Structure:
  • The second-order CFA demonstrated superior fit (CFI = 0.989, RMSEA = 0.042) compared to correlated five-factor (CFI = 0.955) and bifactor models (which produced inadmissible solutions).
  • Scalar invariance was established across both sex and time, validating the comparison of scores between men/women and between age 50 and 62.
• Validity:
  • Concurrent Validity: IC scores showed strong, graded associations with health indicators. Self-rated health explained 43% of the variance in IC scores. There was a clear dose-response relationship: IC scores increased with better self-rated health and decreased with higher multimorbidity.
  • Known-Groups Validity: Women had significantly lower latent IC means than men ($\Delta = -0.144$), consistent with prior literature.
• Longitudinal Change:
  • Mean IC scores declined by 6.3 points (on a 0–100 scale) from age 50 to 62 ($p < 0.001$).
  • Domain-Specific Decline: The decline was not uniform.
    • Physical capacity: Largest decline ($\Delta = -6.8$).
    • Vitality: $\Delta = -5.6$.
    • Sensory: $\Delta = -5.3$ (showed the highest individual variability).
    • Cognitive: $\Delta = -3.7$.
    • Psychological: Smallest decline ($\Delta = -2.3$).
  • 99.97% of participants showed some degree of decline, though the magnitude varied.

5. Significance and Implications

• Early Intervention Window: The findings suggest that functional decline begins in midlife, well before the onset of clinical frailty or disability. This supports the WHO life-course perspective and highlights midlife as a critical window for monitoring and intervention.
• Differentiated Trajectories: The heterogeneous rates of decline across domains (e.g., physical vs. psychological) indicate that functional change in midlife is a process of differentiation rather than a uniform global decline.
• Methodological Foundation: By establishing a valid, invariant measurement tool for midlife, this study enables future longitudinal research to track the "formation, change, and early differentiation" of intrinsic capacity. This is crucial for developing policies and interventions aimed at preserving functional ability before irreversible decline occurs.
• Limitations: The study noted limitations including the use of self-reported data (few objective measures), potential mode effects due to the pandemic (video interviews in S10), and the inability to estimate a weighted second-order model due to convergence issues (though unweighted and weighted downstream results were consistent).

In conclusion, the paper successfully validates the Intrinsic Capacity construct for midlife populations, demonstrating that it is a measurable, stable, and sensitive indicator of early functional decline, thereby extending the utility of the WHO healthy aging framework to earlier stages of the life course.