Start with who answered and what was asked
Table 10.25 reports women and men aged 15–49 separately. Each food has four response categories: daily, weekly, occasionally and never. The table is an estimate for these age and sex groups, not a tally of the entire population. Children and older adults are outside this table’s age range.
For the combined fish, chicken or meat category, 29.4% of women and 16.6% of men reported never. For eggs, the corresponding figures were 28.0% and 15.4%. These should be quoted with the food category, age range, sex and survey year. In this question, never describes a reported habitual frequency; it is not proof that the respondent has never eaten the food at any point in life.
Evidence: National Family Health Survey (NFHS-5), 2019–21: India, Volume I
Why these four percentages do not produce one vegetarian rate
Vegetarianism can be defined in different ways, especially when eggs are involved. The survey’s combined fish, chicken or meat category excludes eggs, which have their own row. A person’s response to the egg question may overlap with their response to the meat question. The two separate percentages do not reveal the size of that overlap.
Adding never-eggs to never-meat percentages would double-count some respondents. Subtracting one from the other would not identify a consistent diet group either. A defined vegetarian measure would require the relevant responses for the same individuals, appropriate survey weights, an explicit treatment of eggs, and a stated population. The published marginal table alone cannot supply that calculation.
Nor should the women’s and men’s percentages simply be averaged and labeled India. That would ignore their population weights and would still omit people outside ages 15–49. A more honest headline answers the question the source actually measured: how often these respondents reported eating the specified food.
Evidence: National Family Health Survey (NFHS-5), 2019–21: India, Volume I
Frequency is different from quantity and nutrition
The table also covers milk or curd, pulses or beans, dark green leafy vegetables, fruit, fried foods and aerated drinks. This makes it possible to compare patterns across a broader food basket. For example, it distinguishes a daily response from a weekly one without pretending that both represent the same frequency.
However, a weekly response does not tell us portion size, the number of meals, household purchasing volume or nutrient intake. A food-frequency percentage cannot by itself establish calorie adequacy, protein adequacy or diet quality. Those questions need the relevant nutrition measurements and careful interpretation of their own age groups and definitions.
Evidence: National Family Health Survey (NFHS-5), 2019–21: India, Volume I
How to use this evidence in a food-market study
A restaurant or food retailer can use the table to form better research questions, such as whether to investigate egg demand separately from meat demand. It cannot infer how many local customers will buy a product, how much they will spend, or which menu they prefer from the national percentages alone.
For spending, the Household Consumption Expenditure Survey provides a different lens: consumption values for household food categories. For local demand, relevant geography, price points, purchase frequency and current primary customer research are still necessary. Do not infer a household’s diet from its religion or from the average for its state.
These detailed food-frequency values remain labeled NFHS-5, 2019–21. The newer NFHS-6 national factsheet does not contain a replacement for this detailed table. A newer survey’s release date is not a reason to relabel older food responses as current.
Evidence: National Family Health Survey (NFHS-5), 2019–21: India, Volume I · National Family Health Survey (NFHS-6), 2023–24: India and State/UT Fact Sheets · NSS Report 592: Survey on Household Consumption Expenditure, 2023–24