Understanding Life Table Columns: From Age Groups to Life Expectancy

A life table looks intimidating at first glance – rows of numbers, columns with cryptic notations like lx, ndx, and ex, and a structure that seems to belong in an actuary’s office rather than a population studies classroom. But once you understand what each column actually represents, the life table becomes one of the most elegant tools in demography. It is essentially a story of a hypothetical group of people moving through life together, with every column answering a specific question: How many are still alive? How many died this year? How many years of life do they have left? Let’s break down each column and see how demographers, public health planners, and even insurance companies use them to make decisions that affect millions of lives.

The basic structure of a life table 🔗

A life table tracks a hypothetical cohort – usually 1,00,000 newborns – and follows them through every age until the last person dies. The numbers in each column are not arbitrary. They are derived from observed age-specific mortality rates in a real population, then applied to this imaginary cohort to reveal what survival and death look like at every stage of life.

There are two main types: the complete life table, which shows data for every single year of age, and the abridged life table, which groups ages into intervals (usually 5 years). India’s Sample Registration System publishes abridged life tables annually, and these form the backbone of mortality analysis in the country. According to a recent study published in BMC Public Health, abridged life tables built from SRS and National Family Health Survey data now allow researchers to estimate life expectancy for over 700 districts across the country.

The age column: Setting the stage 🔗

The age column, often denoted as x or (x, x+n), defines the time intervals over which the life table operates. In a complete life table, each interval covers one year – age 0 to 1, age 1 to 2, and so on. In an abridged life table, you typically see intervals like 0-1, 1-4, 5-9, 10-14, all the way up to 85+.

Why does this matter? Because mortality risk is not evenly distributed across the lifespan. Infants face higher death rates than older children, and risk rises sharply again in old age. By dividing the lifespan into intervals, demographers can capture these patterns precisely. The age column also determines the population structure the table represents – it tells you which segments of life you are analyzing and at what resolution.

Why a separate 1-4 interval matters 🔗

Most abridged life tables split the first five years of life into two unequal intervals: 0-1 and 1-4. This is deliberate. Infant mortality, especially in the neonatal period (the first 28 days), is dramatically higher than mortality between ages 1 and 4. Combining them would mask critical information about early childhood survival, which is central to public health planning in countries where infant mortality is still a concern.

The survival column (lx): Who is still alive? 🔗

The lx column is arguably the heart of the life table. It shows the number of people from the original cohort who are still alive at the beginning of each age interval. The starting value, called the radix, is typically set at 1,00,000 – representing 1,00,000 simultaneous births.

As age increases, lx decreases. If l₀ = 1,00,000 and l₁ = 97,500, that means 2,500 babies died before reaching their first birthday. The UK Office for National Statistics describes lx as the number of survivors out of 100,000 live births who would experience the mortality rates of the period being studied. The column has a powerful probabilistic interpretation: dividing lx by the radix gives the probability that a newborn will survive to age x.

What lx tells us about a population 🔗

Plotting lx against age produces a survivorship curve, which is one of the most visually striking outputs of demographic analysis. In countries with high infant mortality, the curve drops sharply in the first year and then declines gradually. In countries with strong health systems, the curve stays nearly flat until middle age before falling steeply in old age – a pattern called the “rectangularisation of survival.”

India’s survivorship curve has steadily shifted toward this rectangular shape over the decades. According to analysis from Data For India, a child born here in 1950 could expect to live only until age 41, while today life expectancy has crossed 70 years – a transformation driven largely by reductions in infant and child mortality reflected directly in the lx column.

The death column (ndx): Counting losses 🔗

The ndx column shows the number of people from the cohort who die during each age interval. It is calculated simply as the difference between lx values: ndx = lx − l(x+n). If 97,500 children are alive at age 1 and 96,800 at age 5, then 700 died during that interval.

This column is incredibly useful for identifying which age groups contribute most to overall mortality. In high-mortality settings, the ndx column shows large numbers in the youngest age groups. In low-mortality settings, deaths cluster heavily in the 70+ age groups. A useful property: the sum of all ndx values across all intervals equals the radix, because everyone in the cohort eventually dies.

Identifying mortality bumps 🔗

The ndx column also reveals unusual patterns. A spike in young adult deaths might point to road accidents, occupational hazards, or maternal mortality. A bulge in midlife could indicate the burden of non-communicable diseases like cardiovascular illness or diabetes – which the World Health Organization in India has flagged as the country’s leading cause of death today. Public health planners use these patterns to design age-targeted interventions.

The probability of dying (nqx) and surviving (npx) 🔗

The nqx column represents the probability that a person who has reached exact age x will die before reaching age x+n. It is calculated as ndx divided by lx. Its companion, npx, is simply 1 − nqx, showing the probability of surviving the interval.

These columns are critical for actuaries, who use them to price life insurance policies and design pension products. If the probability of a 60-year-old dying within five years is, say, 0.05, an insurer can use that figure – adjusted for risk factors and loading – to determine premiums for a term policy. The Insurance Regulatory and Development Authority of India requires insurers to use mortality tables derived from life table principles when filing product designs.

The person-years column (nLx): Time lived, not just lives counted 🔗

This is where the life table starts doing something subtler. The nLx column represents the total number of person-years lived by the cohort within each age interval. If everyone in an interval survived the full n years, nLx would simply equal n × l(x+n). But some people die mid-interval, so we need to account for the partial years they contributed.

The standard approximation is: nLx = (n/2) × (lx + l(x+n)), assuming deaths are spread evenly across the interval. For infants, this assumption fails badly – most infant deaths cluster in the first weeks of life – so a separate adjustment factor (called a₀) is used for the 0-1 interval.

Why does nLx matter? Because it is the bridge between counting people and measuring life experience. A population that loses 1,000 people at age 80 has lost far fewer “life-years” than one that loses 1,000 people at age 20. The nLx column captures this difference.

The total person-years column (Tx) 🔗

The Tx column is the cumulative sum of all nLx values from a given age onward. It represents the total number of years that will be lived by the entire cohort from age x until everyone has died. T₀ is therefore the sum of all life-years lived by the original 1,00,000 newborns.

You will not see Tx talked about often outside textbooks, but it is the engine that makes the final and most famous column possible.

Life expectancy (ex): The headline number 🔗

The ex column shows the average number of additional years a person who has reached exact age x can expect to live, assuming current mortality rates continue. It is calculated as: ex = Tx / lx.

The most famous value in any life table is e₀, life expectancy at birth. According to India’s Ministry of Health and Family Welfare and SRS data, life expectancy at birth in India has risen from around 32 years at independence to over 70 years today, with women now outliving men by more than three years on average.

Why life expectancy changes with age 🔗

A common point of confusion: if life expectancy at birth is 70 years, why does a 60-year-old’s life expectancy not equal 10 years? The answer is that ex is conditional. By surviving to age 60, you have already escaped all the mortality risks of infancy, childhood, and young adulthood. So your expected remaining lifespan is calculated from a different starting point. In most populations, a 60-year-old can expect to live well beyond age 70 – often into their late 70s or 80s.

How these columns work together in real applications 🔗

Life tables are not just academic exercises. They feed directly into public health planning, insurance pricing, pension design, and even ecological research. When the government plans how many old-age homes to build, when LIC sets premiums for a 30-year endowment policy, or when researchers compare the impact of a new vaccination programme, life table columns are doing the quiet work behind the scenes.

Subnational life tables are also becoming more granular. Recent research using NFHS-4 and NFHS-5 data has constructed life tables for hundreds of Indian districts, revealing stark inequalities – life expectancy can vary by more than 15 years between districts in the same country. This kind of insight, drawn directly from lx, ndx, and ex columns, is what makes life tables indispensable for evidence-based policy.

What do you think? If you compared the life table of your home state with that of Kerala or Bihar, which column would tell the most revealing story about health inequalities? And how would you expect India’s life table to look 30 years from now – more rectangular, or shaped by entirely new mortality patterns?

References

1. https://link.springer.com/article/10.1186/s12889-024-18278-3
2. https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/healthandlifeexpectancies/methodologies/guidetocalculatingnationallifetables
3. https://www.dataforindia.com/life-expectancy/
4. https://www.who.int/india/health-topics/noncommunicable-diseases
5. https://www.irdai.gov.in/
6. https://main.mohfw.gov.in/
7. https://www.healthknowledge.org.uk/public-health-textbook/health-information/3a-populations/life-tables-demographic-applications
8. https://pmc.ncbi.nlm.nih.gov/articles/PMC11021017/