Individual and collective bodies: using measures of variance and association in contextual epidemiology

Abstract

Both measures of variance and means-centric measures of association need to be included when performing contextual analyses. The variance approach, a new aspect of contextual analysis that cannot be interpreted in means-centric terms, allows perspectives to be expanded.

Bringing the individual back to small-area variation studies: A multilevel analysis of all-cause mortality in Andalusia, Spain

“…That is, if the geographical administrative boundaries actually capture a relevant context that conditions individual health, one should expect not only a statistically significant spatial variation (as frequently detected in "small-area variation studies") (Ibanez et al, 2009), but also that this spatial variation conditions a meaningful proportion of the total individual-level variation (Boyle MH & Willms JD, 1999;Merlo, 2003;Merlo et al, 2004;Merlo et al, 2009;Subramanian SV et al, 2007). Pioneer in interpreting general contextual effects was also the work by Duncan, Jones, and Moon referred above (Duncan et al, 1993).…”

“…Since many social processes take place over space (Cummins et al, 2007;Kaplan, 1999;Macintyre S & Elleway A, 2000;Macintyre et al, 2002;Merlo J, 2011), it is a fundamental issue in public health to identify the social and geographical environments that condition individual health variance. Moreover, is necessary to ascertain the specific characteristics of these contexts that explain such variance and are associated individual disease risk (Cummins et al, 2007;Merlo et al, 2009).…”

“…However, many multilevel analyses performed so far have been mainly focused on the study of associations between contextual variables and individual health, considering the analysis of variance as secondary information (Blakely & Woodward, 2000;Diez Roux, 2008). In contrast, others scholars have explicitly concluded that the analysis of variance provides indispensable information for understanding place effects on health (Boyle MH & Willms JD, 1999;Clarke P & Wheaton B, 2007;Duncan et al, 1993;Merlo, 2003;Merlo et al, 2004;Merlo et al, 2009;Riva et al, 2007) Moreover, all around the world, a persistent amount of observational information on place effects is still being obtained from ecological/spatial studies of "small-area variations" , frequently in the form of coloured atlases and disease maps (Benach et al, 2003;Benach et al, 2004;Borrell et al, 2010;Collaboration, 2010;MacNab & Dean, 2002;Middleton et al, 2008;Ocana-Riola & Mayoral-Cortes, 2010;Ocana-Riola et al, 2008a;Pickle et al, 1999;Shaw, 2008;Turrell & Mengersen, 2000). From an empirical perspective, the advantages of multilevel versus ecological regression analyses were clearly identified by the seminal work performed by Aitkins and Longford (Aitkin M & Longford N, 1986) as well as by Jones,4 Duncan, Moon, Subramanian and colleagues (Bullen et al, 1996;Duncan et al, 1993Duncan et al, , 1995Duncan et al, , 1996Duncan et al, , 1998Duncan et al, , 1999Jones et al, 1991;Subramanian et al, 2009;Twigg et al, 2000).…”

“…Beyond the conventional study of specific contextual effects (i.e., measures of association between particular contextual variables and individual outcomes), the use of multilevel regression and analogous techniques of analyses (Katz et al, 1993) has allowed us to identify general contextual effects that are based on measures of variance and clustering rather than on measures of association (Merlo, 2003;Merlo et al, 2009;Merlo et al, 2001;Petronis & Anthony, 2003;Subramanian SV et al, 2007). However, while the explicit distinction between these two types of effects enriches our understanding on contextual influences on individual health, this approach is still rather infrequent in social epidemiology ).…”

“…Regarding ecological studies of small-area variations the concerns are not only against the legendary "ecological fallacy" (Morgenstern, 1998;Robinson, 2009) and the advantages of the multilevel analysis for preventing it (Merlo et al, 2001;Subramanian et al, 2009), but rather on the incapacity of ecological analyses for quantifying general contextual effects because they lack information on how variance is partitioned across the different levels of analysis (e.g., individual and areas) (Merlo, 2003;Merlo et al, 2009;Subramanian SV et al, 2007).…”