Table 4.

Characteristics of studies included in the systematic review, Temperature-change-effect Group.

Source	Location	Time period	Definition of cold spell (n of episodes)	Main outcomes and stratification by sex and age	Potential confounders taken into account	Main findings and effect modification
Chen et al. 2010	349 townships, Taiwan	1997–2003	Cold surge: fast drop in temperature (over 8°C temperature drop in 24 hr, or Tmin < 10°C); (n = 13)	Mortality: CVD	Temporality	Statistically nonsignificant positive association. Social determinants had spatial nonstationary effects
Yang et al. 2009	Taiwan	2000–2003	Cold surge: fast drop in temperature (> 8°C temperature drop in 24 hr, or Tmin < 10°C); (n = 4)	Mortality: CVD	Temporality	Statistically nonsignificant positive associations. A spatially varying pattern of tolerance to cold surges
Ha et al. 2009	City of Seoul, South Korea	1994–2006	Cold wave index (CWI) = difference between Tmin of consecutive days; threshold derived from model fitting; (n = not available)	Mortality: total, CVD, Cardioresp; all ages, 0–64, ≥ 65 years	RH, Temporality	Associations between CWI and mortality. Elderly > other
Temperature-change-effect Group: overall, main, or added effect of a cold spell defined as a sudden change in temperature on mortality or morbidity. Abbreviations: Cardioresp, cardiorespiratory (cardiovascular diseases and respiratory diseases combined); CVD, cardiovascular diseases; RH, relative humidity; Tmin, daily minimum temperature; Temporality, long- or short-term temporal trends and/or seasonal variation and/or day of the week.