Hypothesis

The annual rate of change in satellite-measured artificial light at night (ALAN) — urbanization velocity — is a stronger predictor of county-level ER+ breast cancer incidence than absolute nighttime light intensity, because rapid circadian disruption prevents melatonin-pathway adaptation.

What Exists

Multiple epidemiological studies link satellite-measured ALAN to breast cancer, with strongest evidence for ER+ tumors (Front Environ Health 2023). The NTP concluded with sufficient evidence that circadian disruption from light exposure causes cancer via melatonin suppression and clock gene dysregulation (BMAL1, PER, CRY) (NTP 2021). A SEER-based study successfully linked VIIRS nighttime light data to thyroid cancer at the county level, demonstrating technical feasibility.

The Gap

Every published study uses static, cross-sectional ALAN measurements — a single snapshot of how bright a county is. No one has tested whether the rate of brightening matters more than the absolute level. This is biologically critical: organisms can partially adapt melatonin secretion to stable light environments, but rapid changes in nocturnal light exposure may overwhelm circadian adaptation mechanisms.

Testable Predictions

1. VIIRS annual radiance slope (2012-2022, nW/cm²/sr per year) at the county level should be a stronger predictor of age-adjusted ER+ breast cancer incidence (from SEER) than mean VIIRS radiance, as measured by AIC comparison in Poisson regression models adjusting for county demographics
2. The effect should be specific to ER+ breast cancer (hormone-dependent, melatonin-pathway-mediated) and absent or weak for ER- breast cancer and non-hormone-dependent cancers (lung, colorectal) — providing a mechanistic control
3. Counties with rapid brightening (top decile of VIIRS slope) but moderate absolute light (middle tertile of mean radiance) should show higher breast cancer incidence than counties with high absolute light but stable illumination (bottom decile of slope, top tertile of mean) — directly testing velocity vs. magnitude
4. Latency: The VIIRS slope from 2012-2017 should predict SEER breast cancer incidence in 2019-2022 better than concurrent slope, consistent with a 2-5 year cancer development lag

Falsification

If VIIRS radiance slope adds no significant predictive power (likelihood ratio test p > 0.05) beyond absolute mean radiance in county-level Poisson models of ER+ breast cancer incidence across 500+ SEER counties, the velocity hypothesis is falsified.

Data Sources (All Free)

• NASA VIIRS DNB: Monthly nighttime light composites 2012-present (eogdata.mines.edu) — radiometrically calibrated, higher resolution than legacy DMSP-OLS
• NCI SEER: Cancer incidence by county, hormone receptor status (seer.cancer.gov)
• Census ACS: County demographics for confounder adjustment
• NOAA: Cloud-free VIIRS composites for temporal trend analysis

Significance

If urbanization velocity matters more than absolute brightness, current risk assessments based on static light maps systematically underestimate cancer risk in rapidly developing areas (suburbs, exurbs, developing countries) while overestimating risk in stably bright cities. This reframes light pollution policy from 'how bright is too bright' to 'how fast is too fast.'