Screening student drinking behaviors: examining AUDIT criterion validity using CIDI-based alcohol use disorder as the 'gold standard'.

Introduction: High levels of alcohol consumption among college students have been observed across countries. Heavy drinking episodes are particularly prevalent in this population, making early identification of potentially harmful drinking critical from a public health perspective. Short screening instruments such as the Alcohol Use Disorders Identification Test (AUDIT) are serviceable in this regard. However, there is a need for studies investigating the criterion validity of AUDIT in the student population. The aim was to examine the criterion validity of the full AUDIT and AUDIT-C (the first three items directly gauging consumption patterns) in a sample of college and university students using 12-month prevalence of alcohol use disorder derived from an electronic, self-administered version of the World Health Organization (WHO) Composite International Diagnostic Interview, fifth version (CIDI 5.0), which serves as the 'gold standard'. Methods: The study population of the current study is derived from the SHoT study (Students' Health and Wellbeing Study), which is a large national survey of students enrolled in higher education in Norway. In a follow-up study of mental disorders among participants of the SHoT2022 study, students were invited to complete a self-administered electronic version of the CIDI. A random sample of 4,642 participants in the nested CIDI-sample was asked to fill out a set of screening instruments, including AUDIT, before starting CIDI. Based on Youden Index maximization, we estimated the sex-specific optimal cut-offs for AUDIT and AUDIT-C in relation to alcohol use disorder, as determined by CIDI. Results: For the full AUDIT, the optimal cut-offs were 9 for males and 10 for females. The corresponding cut-offs for AUDIT-C were 6 for males and 5 for females. The same optimal cut-offs for both the full AUDIT and AUDIT-C were replicated in bootstrapped analyses with 1,000 runs. Conclusion: The full AUDIT demonstrated acceptable criterion validity with a balance between sensitivity and specificity. However, for AUDIT-C, caution should be exercised when interpreting screening results among college and university students. In conclusion, the full AUDIT is a reliable screening instrument for college and university students, while further modification may be needed for AUDIT-C in this setting.

Changes in life expectancy and disease burden in Norway, 1990-2019: an analysis of the Global Burden of Disease Study 2019.

BACKGROUND: Geographical differences in health outcomes are reported in many countries. Norway has led an active policy aiming for regional balance since the 1970s. Using data from the Global Burden of Disease Study (GBD) 2019, we examined regional differences in development and current state of health across Norwegian counties. METHODS: Data for life expectancy, healthy life expectancy (HALE), years of life lost (YLLs), years lived with disability (YLDs), and disability-adjusted life-years (DALYs) in Norway and its 11 counties from 1990 to 2019 were extracted from GBD 2019. County-specific contributors to changes in life expectancy were compared. Inequality in disease burden was examined by use of the Gini coefficient. FINDINGS: Life expectancy and HALE improved in all Norwegian counties from 1990 to 2019. Improvements in life expectancy and HALE were greatest in the two counties with the lowest values in 1990: Oslo, in which life expectancy and HALE increased from 71·9 years (95% uncertainty interval 71·4-72·4) and 63·0 years (60·5-65·4) in 1990 to 81·3 years (80·0-82·7) and 70·6 years (67·4-73·6) in 2019, respectively; and Troms og Finnmark, in which life expectancy and HALE increased from 71·9 years (71·5-72·4) and 63·5 years (60·9-65·6) in 1990 to 80·3 years (79·4-81·2) and 70·0 years (66·8-72·2) in 2019, respectively. Increased life expectancy was mainly due to reductions in cardiovascular disease, neoplasms, and respiratory infections. No significant differences between the national YLD or DALY rates and the corresponding age-standardised rates were reported in any of the counties in 2019; however, Troms og Finnmark had a higher age-standardised YLL rate than the national rate (8394 per 100 000 [95% UI 7801-8944] vs 7536 per 100 000 [7391-7691]). Low inequality between counties was shown for life expectancy, HALE, all level-1 causes of DALYs, and exposure to level-1 risk factors. INTERPRETATION: Over the past 30 years, Norway has reduced inequality in disease burden between counties. However, inequalities still exist at a within-county level and along other sociodemographic gradients. Because of insufficient Norwegian primary data, there remains substantial uncertainty associated with regional estimates for non-fatal disease burden and exposure to risk factors. FUNDING: Bill & Melinda Gates Foundation, Research Council of Norway, and Norwegian Institute of Public Health.