Table 1

Characteristics of studies used for meta-regression

Study regionFirst author, year publishedFollow-up years or time lagArea-level variableData yearAverage population (1000)Median Gini coefficient
Cohort studies
 Copenhagen, DenmarkOsler, 2002113–28149 parishes1976–1978/1964–19927.50.258
 FinlandBlomgren, 200412684 NUT4 regions1990470.220
 NorwayKravdal, 2008131–22431 municipalities1980–2002100.370
 New ZealandBlakely, 200314335 subregions1991750.338
 SwedenHenriksson, 2006152–7170 parities/municipalities1990140.251
 SwedenGerdtham, 20041610–1624 counties / 284 municipalities1980–1986140.183
 USAFiscella, 2000/199717 182–16105 primary sampling units1971–1975700.370
 USALochner, 2001191–648 states1987–199448800.340
 USABacklund, 2007204.75–10.7550 states1979–198544400.380
Cross-sectional studies
 Ontario, CanadaXi, 200521No42 public health units1996–19972560.370
 ChileSubramanian, 200322No68 communities2000440.475
 Aichi, JapanIchida, 200523No25 kyuson communities2003240.320
 JapanShibuya, 200224No46 prefectures199526120.361
 ScotlandCraig, 200525No32 local authorities1999–20001570.282
 UKWeich, 200226No18 regions199131910.284
 USALopez, 200427NoMetro areas1993–19948200.406
 USAKennedy, 199828Yes50 states200049700.332
 USAShi, 200031No50 states199550000.430
 USAKahn, 200032Yes50 states198848000.423
 East EuropeBobak, 200034No7 nations1996/199830 9000.342
 Middle and East EuropeBobak, 200733No13 nations200423 7000.340
 Multiple countriesTorsheim, 200635No27 nations1997–199829 5000.308