lasi-spirometry-reference 0.1.0

National LASI GAMLSS/LMS spirometry reference equations for Indian adults aged 45 years and older.

LASI spirometry reference equations

This folder makes the national LASI GAMLSS/LMS spirometry reference equations usable outside the manuscript.

It contains:

• pyproject.toml and src/lasi_spirometry_reference/: a small Python package.
• r/lasiSpirometryReference/: a small R package.
• Bundled LMS data from gamlss/gamlss_LMS_table.csv.
• Tests against the published lookup examples.

The equations apply to Indian adults aged 45-90 years. They should not be extrapolated to younger adults, and they should be externally validated before being treated as a definitive clinical standard.

Formula

For each sex, age, and parameter, the package reads the tabulated GAMLSS/BCCG values:

• L: Box-Cox skewness.
• M: median at the sex-specific reference height.
• S: coefficient of variation.
• b: height exponent for FEV1 and FVC.

For FEV1 and FVC:

M(age, height) = M_ref(age) * (height_cm / reference_height_cm)^b
z = ((observed / M)^L - 1) / (L * S)
LLN = M * (1 + L * S * -1.6448536269514722)^(1 / L)

For FEV1/FVC, M is in percent and no height scaling is applied.

Integer ages 45-90 are tabulated. Non-integer ages are linearly interpolated between adjacent LMS rows. Ages outside 45-90 raise an error by default.

Python

Install locally from this folder:

python -m pip install -e .

Example:

from lasi_spirometry_reference import predict, score_spirometry

print(predict("M", 60, 165, "fvc"))

result = score_spirometry(sex="M", age=60, height_cm=165, fev1_l=2.10, fvc_l=2.60)
print(result["classification"])

Run tests:

python -m pytest tests

R

Install from the R package subfolder:

install.packages("r/lasiSpirometryReference", repos = NULL, type = "source")

Example:

library(lasiSpirometryReference)

lasi_predict("M", 60, 165, "fvc")
lasi_score_spirometry("M", 60, 165, fev1_l = 2.10, fvc_l = 2.60)

Citation

Please cite the manuscript:

Siddalingaiah HS. Nationally Representative Spirometry Reference Equations for Middle-Aged and Older Indians: A Cross-sectional Derivation and Validation Study from the Longitudinal Ageing Study in India, and Re-estimated Burden of Restrictive and Preserved-Ratio Impairment.

Release checklist

Before public package release, confirm:

• The manuscript citation and DOI/preprint URL, once available.
• The intended license for code and reference tables.
• Whether package publication should be GitHub-only, PyPI, CRAN, or r-universe.