Objective   To investigate the relationship between temperature extremes and daily number of non-accidental and cause-specific deaths and to identify populations vulnerable to adverse effect of temperature extremes in Ji′nan city, Shandong province.

  Methods  We collected data on daily number of non-accidental and cause-specific deaths from death registry of Ji′nan city over the period of 2011 – 2014 and meteorological variables were also collected simultaneously. We applied time-series adjusted Poisson regression to assess the effects of extreme temperature on mortality and death risk of non-accidental and cause-specific deaths by controlling for lagged effect, autocorrelation, age, gender, and other confounders.

  Results   Extreme low-temperature events (cold spells) significantly increased daily number of non-accidental and cause-specific deaths (P < 0.01). Extreme high-temperature events (heat waves) increased non-accidental deaths but not cause-specific deaths. Cold spells significantly increased the risk of deaths due to non-accidental injury (relative risk RR = 1.27), respiratory disease (RR = 1.19), stroke (RR = 1.11), and chronic obstructive pulmonary disease (COPD) (RR = 1.08) (P < 0.01 for all). Heat waves were most pronounced for increased deaths of non-accidental injury (RR = 1.02) and stroke (RR = 1.06）(both P < 0.05). The vulnerability to cold spells was for the whole population regardless of age and gender. The elderly were more vulnerable to adverse effect of heat waves (RR = 1.03, P < 0.01).

  Conclusion   Both extreme high and low temperature could increase daily number and risk of non-accidental and cause-specific deaths among residents in Ji′nan city; the whole population is vulnerable to cold spells and the elderly aged more than 65 years are more vulnerable to heat waves. The findings suggest that interventions should be developed for vulnerable populations during extreme weather events.