Artificial Light at Night (ALAN) and breast cancer incidence worldwide: A revisit of earlier findings with analysis of current trends

Abstract

In a study published in Cancer Causes & Control in 2010, Kloog with co-authors tested, apparently for the first time, the association between population-level ambient exposure to artificial light at night (ALAN) and incidence of several cancers in women from 164 countries worldwide. The study was based on 1996–2002 data and concluded that breast cancer (BC) incidence was significantly and positively associated with ALAN, while no such association was revealed for other cancer types. An open question, however, remains whether the trends revealed by Kloog and co-authors were time specific or also hold true for more recent data. Using information obtained from the GLOBOCAN, US-DMSP and World Bank’s 2002 and 2012 databases, we reanalyzed the strength of association between BC incidence rates in 180 countries worldwide and ALAN, controlling for several country-level predictors, including birth rates, percent of urban population, per capita GDP and electricity consumption. We also compared BC age-standardized rates (ASRs) with multi-annual ALAN measurements, considering potentially different latency periods. Compared with the results of Kloog et al.’s analysis of the year-2002 BC-data, the association between BC and ALAN appears to have weakened overall, becoming statistically insignificant in the year 2012 after being controlled for potential confounders (t < 0.3; p > 0.5). However, when the entire sample of countries was disaggregated into geographic clusters of similarly developed countries, a positive BC-ALAN association re-emerged as statistically significant (t > 2.2; p < 0.01), helping to explain, along with other factors covered by the analysis, about 65–85% of BC ASR variability worldwide, depending on the model type. Although the present analysis reconfirms a positive BC-ALAN association, this association appeared to diverge regionally in recent years, with countries in Western Europe showing the highest levels of such association, while countries in Southeast Asia and Gulf States exhibiting relatively low BC rates against the backdrop of relatively high ALAN levels. This regional stratification may be due to additional protective mechanisms, diminishing BC risks and potentially attributed to the local diet and lifestyles.

• Age standardized rates
• artificial light at night
• breast cancer
• regional differences
• world countries

DECLARATION OF INTEREST

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

Notes

¹Urban population refers to people living in urban areas, and was obtained from the United Nations World Urbanization Prospects database (WB, 2014b).

²In addition to night lights (directly captured by satellite sensors), there is a widespread use (especially in developed countries) of various hand-held electronic devices, such as ipads, personal computers and smart phones, kept close to the eyes and often used during night-time hours. Although such devices are a non-quantified source of ALAN by satellites, their use is reflected, at least indirectly, by per capita GDP and similar development measures.

³Due to the fact that the original values of country-specific BC ASRs deviated from normality (KS=0.096, p < 0.01), data on BC incidence ASR were transformed using the Box–Cox transformation procedure (Li, 2005) to comply with the normality assumption, stipulated by the multivariate regression analysis.

⁴Different functional forms of the models (both linear and logarithmically transformed) were tested, and logarithmic transformations were subsequently selected as the best performing ones for most input variables.

⁵Hereon we use birth rates a measure of actual births, not fertility rates, a potential birth measure, used by Kloog et al. (2010).

⁶ The Gulf States are outstanding in their per capita electricity consumption, due to extensive use of air-conditioning (Krane, 2012).