Categorisation of continuous risk factors in epidemiological publications: a survey of current practice
Tóm tắt
Reports of observational epidemiological studies often categorise (group) continuous risk factor (exposure) variables. However, there has been little systematic assessment of how categorisation is practiced or reported in the literature and no extended guidelines for the practice have been identified. Thus, we assessed the nature of such practice in the epidemiological literature. Two months (December 2007 and January 2008) of five epidemiological and five general medical journals were reviewed. All articles that examined the relationship between continuous risk factors and health outcomes were surveyed using a standard proforma, with the focus on the primary risk factor. Using the survey results we provide illustrative examples and, combined with ideas from the broader literature and from experience, we offer guidelines for good practice. Of the 254 articles reviewed, 58 were included in our survey. Categorisation occurred in 50 (86%) of them. Of those, 42% also analysed the variable continuously and 24% considered alternative groupings. Most (78%) used 3 to 5 groups. No articles relied solely on dichotomisation, although it did feature prominently in 3 articles. The choice of group boundaries varied: 34% used quantiles, 18% equally spaced categories, 12% external criteria, 34% other approaches and 2% did not describe the approach used. Categorical risk estimates were most commonly (66%) presented as pairwise comparisons to a reference group, usually the highest or lowest (79%). Reporting of categorical analysis was mostly in tables; only 20% in figures. Categorical analyses of continuous risk factors are common. Accordingly, we provide recommendations for good practice. Key issues include pre-defining appropriate choice of groupings and analysis strategies, clear presentation of grouped findings in tables and figures, and drawing valid conclusions from categorical analyses, avoiding injudicious use of multiple alternative analyses.
Tài liệu tham khảo
Altman DG, Royston P: The cost of dichotomising continuous variables. Br Med J 2006, 332:1080.
Royston P, Altman DG, Sauerbrei W: Dichotomizing continuous predictors in multiple regression: a bad idea. Stat Med 2006, 25:127–141.
Cochran WG: The effectiveness of adjustment by subclassification in removing bias in observational studies. Biometrics 1968, 24:295–313.
Connor RJ: Grouping for testing trends in categorical data. J Am Stat Assoc 1972, 67:601–604.
Cox DR: Note on grouping. J Am Stat Assoc 1957, 52:543–547.
Lagakos SW: Effects of mismodelling and mismeasuring explanatory variables on tests of their association with a response variable. Stat Med 1988, 7:257–274.
Morgan TM, Elashoff RM: Effect of categorising a continuous covariate on the comparison of survival time. J Am Stat Assoc 1986, 81:919–921.
Taylor JMG, Yu M: Bias and efficiency loss due to categorising an explanatory variable. J Multivar Anal 2002, 83:248–263.
Zhao PZ, Kolonel LN: Efficiency loss from categorising quantitative exposures into qualitative exposures in case-control studies. Am J Epidemiol 1992, 136:464–474.
Altman DG: Categorizing continuous variables. In Encyclopedia of Biostatistics. Edited by: Armitage P, Colton T. Chicester: John Wiley and Sons; 1998:563–567.
Dinero TE: Seven Reasons why you should not categorise continuous data. J Health Soc Policy 1996, 8:63–72.
Greenland S: Dose-response and trend analysis in epidemiology: alternatives to categorical analysis. Epidemiology 1995, 6:356–365.
Greenland S: Avoiding power loss associated with categorisation and ordinal scores in dose-response and trend analysis. Epidemiology 1995, 6:450–454.
Pocock SJ, Collier TJ, Dandero KJ, de Stavola BL, Goldman MB, Kalish LA, Kasten LE, McCormack VA: Issues in the reporting of epidemiological studies: a survey of recent practice. Br Med J 2004, 329:883–888.
von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP: The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet 2007, 370:1453–1457. STROBE Initiative
Vidula H, Tian L, Liu K, Criqui MH, Ferrucci L, Pearce WH, Greenland P, Green D, Tan J, Garside DB, Guralnik J, Ridker PM, Rifai N, McDermott MM: Biomarkers of inflammation and thrombosis as predictors of near-term mortality in patients with peripheral arterial disease: a cohort study. Ann Intern Med 2008, 148:85–93.
Rosenlund M, Picciotto S, Forastiere F, Stafoggia M, Perucci CA: Traffic-related air pollution in relation to incidence and prognosis of coronary heart disease. Epidemiology 2008, 19:121–128.
Tsai SP, Ahmed FS, Wendt JK, Bhojani F, Donnelly RP: The impact of obesity on illness absence and productivity in an industrial population of petrochemical workers. Ann Epidemiol 2008, 18:8–14.
Matsunga I, Miyake Y, Yoshida T, Miyamoto S, Ohya Y, Sasaki S, Tanaka K, Oda H, Ishiko O, Hirota Y, The Osaka Maternal and Child Health Study Group: Ambient formaldehyde levels and allergic disorders among Japanese pregnant women: baseline data from the Osaka Maternal and Child Health Study. Ann Epidemiol 2008, 18:78–84.
Catov JM, Bodnar LM, Ness RB, Barron SJ, Roberts JM: Inflammation and dyslipidemia related to risk of spontaneous preterm birth. Am J Epidemiol 2007, 166:1312–1319.
Leonard H, Nassar N, Bourke J, Blair E, Mulroy S, de Klerk N, Bower C: Relation between intrauterine growth and subsequent intellectual disability in a ten-year population cohort of children in Western Australia. Am J Epidemiol 2008, 167:103–111.
Cauley JA, Hochberg MC, Lui L-Y, Palermo L, Ensrud KE, Hillier TA, Nevitt MC: Long-term risk of incident vertebral fractures. JAMA 2007, 298:2761–2767.
Fang F, Ye W, Fall K, Lekander M, Wigzell H, Sparen P, Adami H-O, Valdimarsdóttir U: Loss of a child and the risk of amyotrophic lateral sclerosis. Am J Epidemiol 2008, 167:203–210.
Bartali B, Frongilo EA, Guralnik JM, Stipanuk MH, Allore HG, Cherubini A, Bandinelli S, Ferrucci L, Gill TM: Serum micronutrient concentrations and decline in physical function among older persons. JAMA 2008, 299:308–315.
Inskip HM, Dunn N, Godfrey KM, Cooper C, Kendrick T, Southampton Women's Survey Study Group: Is birth weight associated with risk of depressive symptoms in young women? Evidence from the Southampton women's survey. Am J Epidemiol 2008, 167:164–168.
Chen H, O'Reilly EJ, Schwarzschild MA, Ascherio A: Peripheral inflammatory biomarkers and risk of Parkinson's disease. Am J Epidemiol 2008, 167:90–95.
Tworoger SS, Lee I-M, Buring JE, Hankinson SE: Plasma androgen concentrations and risk of incident ovarian cancer. Am J Epidemiol 2008, 167:211–218.
Brunner Huber LR, Toth JL: Obesity and oral contraceptive failure: findings from the 2002 national survey of family growth. Am J Epidemiol 2007, 166:1306–1311.
Roddam AW, Neale R, Appleby P, Allen NE, Tipper S, Key TJ: Association between plasma 25-hydroxyvitamin D levels and fracture risk: the EPIC-Oxford study. Am J Epidemiol 2007, 166:1327–1336.
Park Y, Mitrou PN, Kipnis V, Hollenbeck A, Schatzkin A, Leitzmann MF: Calcium, dairy foods, and risk of incident and fatal prostate cancer: the NIH-AARP diet and health study. Am J Epidemiol 2007, 166:1270–1279.
Kifley A, Liew G, Wang JJ, Kaushik S, Smith W, Wong TY, Mitchell P: Long-term effects of smoking on retinal microvascular caliber. Am J Epidemiol 2007, 166:1288–1297.
Mukamal KJ, Kennedy M, Cushman M, Kuller LH, Newman AB, Polak J, Criqui MH, Siscovick DS: Alcohol consumption and lower extremity arterial disease among older adults: the cardiovascular health study. Am J Epidemiol 2008, 167:34–41.
Auchincloss AH, Diez R, Ana V, Brown DG, Erdmann CA, Bertoni AG: Neighborhood resources for physical activity and healthy foods and their association with insulin resistance. Epidemiology 2008, 19:146–157.