COGNITIVE FUNCTION
The NPI™ method and managing cognitive capacity in knowledge work
Knowledge work productivity is still mostly measured through outputs: projects delivered, hours billed, deals closed. These are late indicators. They report what capacity achieved, not how much capacity remained. When capacity falls, output holds for a while and then drops all at once.
The Neuro Performance Index (NPI™) is a measure developed by NeuroAudit® Ltd. that makes cognitive function a trackable variable. Its purpose is not to replace business metrics but to supply the leading indicator they lack. This article covers what the method rests on, why twelve pillars rather than one number, and how the result is used in management.
Why cognitive function needs its own measurement
Cognitive function is a factor of production in knowledge work in the same way that machine utilisation is in manufacturing. It is, however, not directly observable. Processing speed is a central biomarker of cognitive performance (Salthouse 2010) and, together with executive functioning, predicts cognitive change (PMC 2025). Neither is visible in a calendar.
The limits of working memory are the key structural factor here. Working memory maintenance relies on sustained prefrontal neuronal activity (Wang et al. 2011), and this capacity is relatively stable. The amount of work can grow, working memory does not.
Subjective judgement cannot replace measurement, but it is not meaningless either. Subjective cognitive decline predicts later measurable change (Rabin et al. 2023), and self-report instruments are considered useful for screening (Crook & Feher 2018). For the method this means self-report is a useful signal when it is structured and repeatable.
Why twelve pillars rather than one number
A single total score reports the state but not the cause. Two people can reach the same value for entirely different reasons: for one the bottleneck lies in sleep and circadian rhythm, for the other in load and recovery. The appropriate action differs completely, so a total alone would drive wrong decisions.
The pillar structure is built on identified modifiable factors. Lancet Commission analyses show that a substantial share of the risk of cognitive decline relates to modifiable factors (Livingston et al. 2020; Livingston et al. 2024), and population attributable estimates support the same conclusion (Norton et al. 2014). WHO guidance on risk reduction covers the same factor groups (WHO 2023).
Finnish evidence supports a multidomain structure. In the FINGER trial the effect arose from a combination of several simultaneous components (Ngandu et al. 2015), and persistence of benefits has been examined in long extended follow-up (Lehtisalo et al. 2023). For the method this is the decisive argument: because the effect arises from a combination, the measurement must be multicomponent as well.
What NPI™ contains and what it does not
NPI™ describes the state of cognitive function and the factors loading it on a scale of 0 to 100. It is assembled from the sub-results of twelve pillars, which include cognitive function, sleep and circadian rhythm, stress and workload, recovery, hormonal balance, and work environment and motor function.
NPI™ is not a medical diagnosis or a screening test for disease. It is a description of function and load whose purpose is to direct action and make change visible. If clinical warning signs appear in the answers, the correct route is healthcare, not interpretation of the index.
The method is also deliberately conservative in interpretation. Scoring is deterministic and repeatable, so that the difference between two measurements reflects change rather than instrument noise. That is a precondition for using the result in management.
Cognitive reserve makes the result strategic
Cognitive reserve is the ability to maintain performance under load and change. Reserve is a modifiable property (Stern et al. 2020), and its accumulation across the life course has been examined in large data (Livingston et al. 2024; Valenzuela & Sachdev 2006). Pathways from intellectual stimulation to reserve have been described in a separate review (PMC 2021).
For an organisation, reserve is a competitiveness factor. It is invisible in good times, because in good times almost everyone performs. It becomes visible when the situation changes rapidly and decisions must be made under uncertainty. Those are exactly the moments when the value of decisions is highest.
The reverse side is equally concrete. High job strain is associated with cognitive decline (Singh-Manoux et al. 2016) and burnout with measurable decline in executive function (Gavelin et al. 2021). Reserve is therefore both built and consumed at work, and both directions are manageable.
How the result is used in management
NPI™ creates value only when it is connected to decisions. The following pattern has proven workable in expert organisations.
- Measure the baseline once and interpret it at unit level, not at person level. The employer sees an aggregated picture, the individual sees their own result.
- Choose one bottleneck per unit. Running three simultaneous changes fails, because change itself creates load.
- Aim the change at the structure of work: meeting practice, protected time, communication channels or work redesign.
- Run the follow-up measurement after 3 or 6 months with the same method. Comparability matters more than refining the instrument.
- Report the result to the leadership team in the same format as other key figures. Brain health stays on the agenda only if it is a number among other numbers.
The boundary between individual and organisation stays clear
Measuring cognitive function in working life is ethically sensitive, and the boundary must be explicit. In the NeuroAudit® Ltd. model the individual receives their own result and personal recommendations, while the employer receives only an anonymous, aggregated picture at unit level.
This is not only a data protection requirement but a condition for the method to work. If a respondent suspects that an answer affects their position, the answer is distorted and the value of the instrument is lost. Trust is therefore a measurement requirement, not merely a principle.
In practice this also means organisational measures are aimed at structure rather than at individuals. Once a bottleneck is identified at unit level, the correction is made to how work is organised, not to who is selected for it.
How NPI™ differs from an ordinary wellbeing survey
A wellbeing survey measures experience. It asks how things feel, and the answer is valuable but sensitive to mood, season and whatever happened that morning. NPI™ measures function and the factors loading it as a structure, and structure changes more slowly than mood.
The second difference is deterministic scoring. The same combination of answers always produces the same result, which means the difference between two measurements reflects change rather than instrument noise. Without that property a follow-up measurement tells you nothing.
The third difference is action guidance. A wellbeing survey usually produces a total score and a comparison with last year. NPI™ produces a pillar level profile showing where in the structure to intervene first. In practice the difference is the same as between a thermometer and a diagnosis.
Same score, two different bottlenecks
Consider two experts with the same NPI™ value. For the first, the lowest sub-results are in sleep and circadian rhythm, while load is moderate. For the second, sleep is fine but load and recovery are low.
For the first, the most effective move targets rhythm: a fixed wake time and morning light exposure, because circadian disruption affects everything else (AGS 2025). For the second, the same measure would achieve little, because the bottleneck is exposure rather than rhythm. Their most effective move is structural reduction of load, which is directly linked to cognitive performance (Singh-Manoux et al. 2016).
This is the argument for the whole pillar structure in miniature. One number says that something must be done. Twelve numbers say what. Without the latter an organisation buys the right measure for the wrong problem, and the result looks as if prevention does not work.
Key findings
Cognitive function is a factor of production in knowledge work. A summary of what managing it requires.
- Output metrics are lagging indicators. They report what capacity achieved, not how much remains.
- One total number reports the state, twelve pillars report the cause and therefore the action.
- Deterministic scoring is the precondition for a follow-up measurement reflecting change rather than instrument noise.
- Cognitive reserve is modifiable and is both built and consumed at work (Stern et al. 2020).
- Individual results and organisational reporting are kept separate. Trust is a measurement requirement, not merely a principle.
Where to start
Managing cognitive function starts with one measurement and one structural change. It requires no new organisation and no large programme, but it does require a commitment to look at the result again.
The NeuroAudit™ Method produces an NPI™ value, a twelve pillar profile and personal recommendations. The organisation receives an anonymous overall picture from which a bottleneck can be selected and change followed. The most important factor of production in knowledge work then moves from guesswork into measurement.
