A New Approach to Age-Period-Cohort Analysis Using Partial Least Squares Regression: The Trend in Blood Pressure in the Glasgow Alumni Cohort

Due to a problem of identification, how to estimate the distinct effects of age, time period and cohort has been a controversial issue in the analysis of trends in health outcomes in epidemiology. In this study, we propose a novel approach, partial least squares (PLS) analysis, to separate the effects of age, period, and cohort. Our example for illustration is taken from the Glasgow Alumni cohort. A total of 15,322 students (11,755 men and 3,567 women) received medical screening at the Glasgow University between 1948 and 1968. The aim is to investigate the secular trends in blood pressure over 1925 and 1950 while taking into account the year of examination and age at examination. We excluded students born before 1925 or aged over 25 years at examination and those with missing values in confounders from the analyses, resulting in 12,546 and 12,516 students for analysis of systolic and diastolic blood pressure, respectively. PLS analysis shows that both systolic and diastolic blood pressure increased with students' age, and students born later had on average lower blood pressure (SBP: −0.17 mmHg/per year [95% confidence intervals: −0.19 to −0.15] for men and −0.25 [−0.28 to −0.22] for women; DBP: −0.14 [−0.15 to −0.13] for men; −0.09 [−0.11 to −0.07] for women). PLS also shows a decreasing trend in blood pressure over the examination period. As identification is not a problem for PLS, it provides a flexible modelling strategy for age-period-cohort analysis. More emphasis is then required to clarify the substantive and conceptual issues surrounding the definitions and interpretations of age, period and cohort effects.

[1]  R. O’Brien,et al.  The age–period–cohort conundrum as two fundamental problems , 2011 .

[2]  E. Steyerberg,et al.  [Regression modeling strategies]. , 2011, Revista espanola de cardiologia.

[3]  Ashutosh Kumar Singh,et al.  The Elements of Statistical Learning: Data Mining, Inference, and Prediction , 2010 .

[4]  P. Baxter,et al.  Assessing the Impact of Body Size in Childhood and Adolescence on Blood Pressure: An Application of Partial Least Squares Regression , 2010, Epidemiology.

[5]  Guohua Li,et al.  What is a cohort effect? Comparison of three statistical methods for modeling cohort effects in obesity prevalence in the United States, 1971-2006. , 2010, Social science & medicine.

[6]  H. Abdi.  Partial least squares regression and projection on latent structure regression (PLS Regression) , 2010 .

[7]  André I. Khuri,et al.  Linear Model Methodology , 2009 .

[8]  P. Gluckman,et al.  Towards a new developmental synthesis: adaptive developmental plasticity and human disease , 2009, The Lancet.

[9]  Kent L Thornburg,et al.  Effect of in Utero and Early-life Conditions on Adult Health and Disease Epidemiol Ogic a Nd Clinic a L Observations , 2022 .

[10]  S. Wold,et al.  Partial Least Squares (PLS) in Cheminformatics , 2008 .

[11]  Wenjiang J. Fu,et al.  The Intrinsic Estimator for Age‐Period‐Cohort Analysis: What It Is and How to Use It1 , 2008, American Journal of Sociology.

[12]  J. Fox,et al.  Applied Regression Analysis and Generalized Linear Models , 2008 .

[13]  Anne-Laure Boulesteix,et al.  Partial least squares: a versatile tool for the analysis of high-dimensional genomic data , 2006, Briefings Bioinform..

[14]  D. Bertrand,et al.  Application of PLS‐DA in multivariate image analysis , 2006 .

[15]  D. Lawlor,et al.  Adult blood pressure and climate conditions in infancy: a test of the hypothesis that dehydration in infancy is associated with higher adult blood pressure. , 2006, American journal of epidemiology.

[16]  G. Davey Smith,et al.  Could dehydration in infancy lead to high blood pressure? , 2006, Journal of Epidemiology and Community Health.

[17]  R. Lin,et al.  Modelling the Age‐Period‐Cohort Trend Surface , 1996 .

[18]  Roman Rosipal,et al.  Overview and Recent Advances in Partial Least Squares , 2005, SLSFS.

[19]  Wenjiang J. Fu,et al.  2. A Methodological Comparison of Age-Period-Cohort Models: The Intrinsic Estimator and Conventional Generalized Linear Models , 2004 .

[20]  Yu-Kang Tu,et al.  Collinearity in linear regression is a serious problem in oral health research. , 2004, European journal of oral sciences.

[21]  P. McKinney,et al.  Type 1 diabetes in Yorkshire, UK: time trends in 0–14 and 15–29‐year‐olds, age at onset and age‐period‐cohort modelling , 2003, Diabetic medicine : a journal of the British Diabetic Association.

[22]  M. Barker,et al.  Partial least squares for discrimination , 2003 .

[23]  N. Glenn.  Distinguishing Age, Period, and Cohort Effects , 2003 .

[24]  Robert Tibshirani,et al.  The Elements of Statistical Learning: Data Mining, Inference, and Prediction, 2nd Edition , 2001, Springer Series in Statistics.

[25]  S. Wold,et al.  PLS-regression: a basic tool of chemometrics , 2001 .

[26]  R. Burdick.  Linear Models in Statistics , 2001, Technometrics.

[27]  G. Smith,et al.  Changes in blood pressure among students attending Glasgow University between 1948 and 1968: analyses of cross sectional surveys , 2001, BMJ : British Medical Journal.

[28]  Frank E. Harrell,et al.  Regression Modeling Strategies: With Applications to Linear Models, Logistic Regression, and Survival Analysis , 2001 .

[29]  G. Smith,et al.  Life course exposure and later disease: a follow-up study based on medical examinations carried out in Glasgow University (1948-68). , 1999, Public health.

[30]  P. Boyle,et al.  Age-period-cohort models: a comparative study of available methodologies. , 1999, Journal of clinical epidemiology.

[31]  S. Ebrahim,et al.  Lowering blood pressure: a systematic review of sustained effects of non-pharmacological interventions. , 1998, Journal of public health medicine.

[32]  D. Harville.  Matrix Algebra From a Statistician's Perspective , 1998 .

[33]  Thomas Mathew,et al.  Mathematical Tools for Applied Multivariate Analysis , 1998 .

[34]  P. Boyle,et al.  Age-period-cohort analysis of chronic disease rates. I: Modelling approach. , 1998, Statistics in medicine.

[35]  R. F. Ling,et al.  Some cautionary notes on the use of principal components regression , 1998 .

[36]  A. Phatak,et al.  The geometry of partial least squares , 1997 .

[37]  J. Witteman,et al.  Long-term effects of neonatal sodium restriction on blood pressure. , 1997, Hypertension.

[38]  R. Lin,et al.  Autoregressive age-period-cohort models. , 1996, Statistics in medicine.

[39]  W. H. Ray,et al.  Partial least squares modelling as successive singular value decompositions , 1993 .

[40]  I. Wakeling,et al.  A test of significance for partial least squares regression , 1993 .

[41]  S. D. Jong.  SIMPLS: an alternative approach to partial least squares regression , 1993 .

[42]  T. Holford.  Understanding the effects of age, period, and cohort on incidence and mortality rates. , 1991, Annual review of public health.

[43]  S. R. Searle.  Restrictions and Generalized Inverses in Linear Models , 1984 .

[44]  M J Gardner,et al.  Age, period and cohort models applied to cancer mortality rates. , 1982, Statistics in medicine.

[45]  Sati Mazumdar,et al.  Correspondence between a Linear Restriction and a Generalized Inverse in Linear Model Analysis , 1980 .

[46]  H. Goldstein.  Age, Period And Cohort Effects — A Confounded Confusion , 1979 .

[47]  N. Glenn,et al.  Cohort Analysts' Futile Quest: Statistical Attempts to Separate Age, Period and Cohort Effects , 1976 .

[48]  N. S. Urquhart,et al.  Generalized Inverse Matrices (with Applications to Statistics) , 1973 .

[49]  D. S. Tracy,et al.  Generalized Inverse Matrices: With Applications to Statistics , 1971 .

[50]  R. Ristl,et al.  Does the choice of intraoperative fluid modify abdominal aneurysm repair outcomes? , 2019, Medicine.