Modeling absolute differences in life expectancy with a censored skew-normal regression approach.

Parameter estimates from commonly used multivariable parametric survival regression models do not directly quantify differences in years of life expectancy. Gaussian linear regression models give results in terms of absolute mean differences, but are not appropriate in modeling life expectancy, because in many situations time to death has a negative skewed distribution. A regression approach using a skew-normal distribution would be an alternative to parametric survival models in the modeling of life expectancy, because parameter estimates can be interpreted in terms of survival time differences while allowing for skewness of the distribution. In this paper we show how to use the skew-normal regression so that censored and left-truncated observations are accounted for. With this we model differences in life expectancy using data from the Swiss National Cohort Study and from official life expectancy estimates and compare the results with those derived from commonly used survival regression models. We conclude that a censored skew-normal survival regression approach for left-truncated observations can be used to model differences in life expectancy across covariates of interest.

What does your neighbourhood say about you? A study of life expectancy in 1.3 million Swiss neighbourhoods

BACKGROUND Switzerland had the highest life expectancy at 82.8 years among the Organisation for Economic Co-operation and Development (OECD) countries in 2011. Geographical variation of life expectancy and its relation to the socioeconomic position of neighbourhoods are, however, not well understood. METHODS We analysed the Swiss National Cohort, which linked the 2000 census with mortality records 2000-2008 to estimate life expectancy across neighbourhoods. A neighbourhood index of socioeconomic position (SEP) based on the median rent, education and occupation of household heads and crowding was calculated for 1.3 million overlapping neighbourhoods of 50 households. We used skew-normal regression models, including the index and additionally marital status, education, nationality, religion and occupation to calculate crude and adjusted estimates of life expectancy at age 30 years. RESULTS Based on over 4.5 million individuals and over 400 000 deaths, estimates of life expectancy at age 30 in neighbourhoods ranged from 46.9 to 54.2 years in men and from 53.5 to 57.2 years in women. The correlation between life expectancy and neighbourhood SEP was strong (r=0.95 in men and r=0.94 women, both p values <0.0001). In a comparison of the lowest with the highest percentile of neighbourhood SEP, the crude difference in life expectancy from skew-normal regression was 4.5 years in men and 2.5 years in women. The corresponding adjusted differences were 2.8 and 1.9 years, respectively (all p values <0.0001). CONCLUSIONS Although life expectancy is high in Switzerland, there is substantial geographical variation and life expectancy is strongly associated with the social standing of neighbourhoods.