Socioeconomic groups and cancer risk at death in the Swiss Canton of Vaud

F. Levi, E. Negri, C. La Vecchia, V. C. Te

Research output: Contribution to journalArticlepeer-review


Data collected by the Cancer Registry of the Canton of Vaud, Switzerland, were used to estimate proportional mortality ratios (PMR) and mortality odds ratios (MOR) for various neoplasms according to social class and sector occupation (agriculture versus others). Mortality ratios were elevated in lower social classes for cancers of the lung (MOR = 1.18 for social class IV or V vs I or II) and other sites strictly to tobacco (mouth or pharynx, oesophagus and larynx; MOR = 1.70), and (though not significantly) for cancers of the stomach (MOR = 1.16) and uterus (MOR = 1.30 for cervix and 1.47 for corpus uteri). Furthermore, there was a strong negative social class gradient for thyroid cancer (a neoplasm with particularly elevated incidence and mortality in Switzerland), probably attributable to higher prevalence of iodine deficiency in lower social classes (MOR = 3.17). Positive social class gradients emerged for cancers of the intestines (MOR = 0.77 for social class IV or V), skin (MOR = 0.74) and prostate (MOR = 0.87). Agricultural workers showed decreased ratios for cancers of the lung (MOR = 0.75), cervix uteri (MOR = 0.72) and prostate (MOR = 0.80), and excess mortality from cancers of the upper digestive and respiratory sites (MOR = 1.22), stomach (MOR = 1.18), testis (MOR = 2.05) and lympho-haematopoietic neoplasms, particularly myeloma (MOR = 2.14).

Original languageEnglish
Pages (from-to)711-717
Number of pages7
JournalInternational Journal of Epidemiology
Issue number4
Publication statusPublished - 1988

ASJC Scopus subject areas

  • Epidemiology
  • Statistics, Probability and Uncertainty
  • Applied Mathematics
  • Physiology (medical)
  • Reproductive Medicine
  • Obstetrics and Gynaecology


Dive into the research topics of 'Socioeconomic groups and cancer risk at death in the Swiss Canton of Vaud'. Together they form a unique fingerprint.

Cite this