909Testing ‘hardening hypotheses’ in 30 Global Adult Tobacco Survey countries: an ecological study

Abstract

\n \n \n Evidence from high-income countries suggests “softening” but not “hardening” occurs when smoking prevalence decreases; however, evidence is lacking from developing countries.\n \n \n \n We constructed heaviness of smoking index (HIS) based on cigarettes smoked per day (CPD) and time-to-first-cigarette (TTFC) and a score of 4-6 was high nicotine dependence; Hardcore smokers (HCS) are daily smokers who smoked within 30\u2009minutes after waking up, smoke ≥10 cigarettes per day, have not made any quit attempts in 12 months, and have no intention to quit at all or during next 12 months; light smokers (<10 CPD or occasional smokers). We plotted two-way linear prediction graphs using country-level data and estimated Spearmańs rank-correlation coefficient. On pooled data multi-level binary logistic regression analyses for HCS, high dependence and light smoker.\n \n \n \n Among men and women range (lowest vs. highest) of rates varied between countries and were lower among women. Daily smoking prevalence was moderately and positively correlated with hard-core smoking (Rsp=0.473), and high dependence (Rsp=0.384) but negatively correlated with light smoking (Rsp=-0.677). HCS and high dependence were associated with socio-economic factors but effect sizes were higher for women. Among GDP and tobacco control score showed protective effect for HCS and high dependence but increased risk for light smoking.\n \n \n \n Positive relationship of daily smoking with HCS and high dependence and its negative relationship with light smoking is contrary to “hardening hypothesis”.\n \n \n \n Hardening may not happen in developing countries as well.\n