Table 1 Characteristics of studies included in analysis
From: Disparities in diabetes mellitus among Caribbean populations: a scoping review
Author/Year | Study Design | Study Characteristics | Ethnic group/Location | Country/Region | Setting | Reported Quality/Limitations |
|---|---|---|---|---|---|---|
Abbott, 2005 [13] | Cross-sectional study | 15,646 | Afro-Caribbean vs. Asians vs. Whites | United Kingdom | Community health center | No limitations reported. Direct standardization use to calculate age adjusted rates. |
Men: 8574 | ||||||
Women: 6892 | ||||||
Type 1&2 DM | ||||||
Abbott, 2011 [14] | Cross-sectional study | 15,692 | Afro-Caribbean vs. South Asians vs. Whites | United Kingdom | Community based population study | No limitations reported. |
Male: 8448 | ||||||
Female: 7236 | ||||||
Mean age: 61 ± 14.0y | ||||||
Type 1 & 2 DM | ||||||
Admiraal, 2011 [15] | Cross-sectional study | 1,443 | Hindustani Surinamese vs. African Surinamese vs. Dutch Caucasians | Netherlands | Community based | Discrepancy in measurement of physical activity as a confounder between ethnic groups. No adjustment for other known confounders (diet). |
Age: 35-60y | ||||||
Type 2 DM | ||||||
Agyemang, 2011 [16] | Cross-sectional study | 3,386 | South East Asian Indians vs. Afro-Caribbean in England and Netherlands | United Kingdom, Netherlands | Population based | Lack of data on all the important explanatory variables that might contribute to the observed differences, such as diet, psychosocial stress, and early-life exposures. Furthermore, there was a lack of valid data on other types of physical activity and socioeconomic position measures. |
Male: 1474 | ||||||
Female: 1912 | ||||||
Type 2 DM | ||||||
Babwah, 2006 [17] | Cross-sectional study | 360 | Trinidad and Tobago | Trinidad and Tobago | Urban clinic | Reporting bias, lack of multivariate analysis to adjust for known confounders (SES) |
Male: 93 | ||||||
Female: 267 | ||||||
Age >13Â years | ||||||
Type 2 DM | ||||||
Barcelo, 2006 [18] | Cross-sectional study | 10,587 | Barbados vs. Mexico | Caribbean, Latin America | Population-based | No limitation stated. Prevalence adjusted for known confounders. |
Male: 4041 | ||||||
Female: 6546 | ||||||
DM unspecified | ||||||
Baskar, 2006 [19] | Cross-sectional study | 6,047 | Afro-Caribbean vs. Caucasian vs. Indo-Asian | United Kingdom | Community based | No limitation stated. Analysis adjusted for known confounders |
Male: 3359 | ||||||
Female 2688 | ||||||
Type 1 & 2 DM | ||||||
Cappuccio, 1997 [20] | Cross-sectional study | 1,578 | Afro-Caribbean vs. West African vs. UK Whites vs. Asians | United kingdom | Community based, general practice | Selection bias in Caribbean group and low response rate. Prevalence rates age standardized by direct method. |
Age: 40-59 | ||||||
Type 2 DM | ||||||
Chaturvedi, 1996 [21] | Cohort study | 227 | Afro-Caribbean vs. European (UK) | United Kingdom | Hospital based | Small sample size particularly among African Caribbeans. Inability to conduct sex specific analysis. No collection of important confounders. Prevalence rates were age standardized. |
Male : 122 | ||||||
Female: 105 | ||||||
Age: 35-55 | ||||||
Type 2 DM | ||||||
Conway, 2003 [22] | Cross-sectional study | 832 | Afro Caribbean vs. Whites vs. Indo-Asian | United Kingdom | Hospital based study | Adjustment for known confounders carried out. |
Male: 449 | ||||||
Female: 383 | ||||||
Age: 74 ± 12y | ||||||
DM unspecified | ||||||
Cooper, 1997 [23] | Cross-sectional study | 4,823 | African origin populations in Nigeria, St. Lucia, Barbados, Jamaica, the United States, and the United Kingdom | Barbados, Jamaica, Nigeria, St Lucia, United Kingdom, United States of America | Community based | Limited sample size in some sites. |
Age: 25-74y | ||||||
Type 2 DM | ||||||
Cox, 2011 [24] | Cross-sectional study | 87 | Jamaica | Jamaica | Hospital based | No limitations stated. |
Male: 35 | ||||||
Female: 52 | ||||||
Age 40-90y | ||||||
DM | ||||||
Creatore, 2012 [25] | Cohort Study | 3,927,059 | Immigrant populations in Canada | Canada | Population based | Due to data restrictions analyses were not adjusted for risk factors. Immigration data restricted sample to immigrants to Canada between 1985 and 2000. |
Male: 2,094,042 | ||||||
Female: 1,833,017 | ||||||
Age : >40Â yrs. | ||||||
Cruickshank, 1980 [26] | Cross-sectional study | 27,667 | Jamaican vs. White vs. West Indian Black | United Kingdom | Hospital based | No limitations stated. |
Male: 11,157 | ||||||
Female: 9,235 | ||||||
Age 30-59y | ||||||
DM unspecified | ||||||
Cruickshank, 1987 [27] | Case–control study | 282 | Afro-Caribbean vs. Caucasian vs. Asian | United Kingdom, Jamaica | Hospital based clinic attendees | No limitations stated. No clear description of statistical technique. |
Men: 119 | ||||||
Women: 163 | ||||||
DM unspecified | ||||||
Eldemire, 1996 [28] | Cross-sectional study | 1,318 | Jamaica | Jamaica | Population based | No limitations stated and no clear description of statistical techniques. |
Male: 649 | ||||||
Female: 669 | ||||||
Age >60 | ||||||
Type 2 DM | ||||||
Ferguson, 2011 [29] | Cross-sectional study | 2,848 | Jamaica | Jamaica | Community based | No limitations stated. Appropriate adjustment for confounders. |
Age: 15-74 | ||||||
DM unspecified | ||||||
Florey, 1972 [30] | Cross-sectional study | 696 | Jamaica | Jamaica | Community based | No limitation stated and no clear description of data analysis technique. |
Male: 329 | ||||||
Female: 367 | ||||||
Age 25-64y | ||||||
DM unspecified | ||||||
Gill, 2011 [31] | Cross-sectional study | 5,354 | Afro-Caribbean vs. South Asians | United Kingdom | Clinic based screening programme | Low response rate (49.6%). Age sex adjustments were not conducted due to small number of cases. |
Male: 2544 | ||||||
Female: 2810 | ||||||
Age > 45 y | ||||||
DM unspecified | ||||||
Goyal, 2007 [32] | Cohort study | 271 | Afro Caribbean vs. Whites vs. South Asians | United Kingdom | Community clinic setting | No limitations stated. |
Male: 184 | ||||||
Female: 87 | ||||||
Age | ||||||
Type unspecified | ||||||
Gulliford, 1997 [33] | Cross-sectional study | 1,149 | Afro-Trinidadian vs. Indo-Trinidadian | Trinidad and Tobago | Hospital based | Evidence of selection bias with more ill patients less likely to provide interview data. |
Male: 454 | ||||||
Female: 695 | ||||||
Age >15y | ||||||
DM unspecified | ||||||
Gulliford, 1998 [34] | Cross-sectional study | 622 | Afro-Trinidadian vs. Indo-Trinidadian | Trinidad and Tobago | Health center | Sample biased to socially less advantage individuals. |
Male: 204 | ||||||
Female: 418 | ||||||
DM type 2 | ||||||
Gulliford, 2001 [35] | Cross-sectional study | 2,117 | Afro-Trinidadian vs. Indo-Trinidadian | Trinidad and Tobago | Government health centres | Large geographically representative sample. Reporting bias; over-reporting of private utilization in older age group. |
Male: 633 | ||||||
Female: 1484 | ||||||
DM unspecified | ||||||
Gulliford, 2004 [36] | Cross-sectional study | 548 | Indo Trinidadian vs. Afro Trinidadian vs. mixed Trinidadian | Trinidad and Tobago | Population based community study | Higher non-response among affluent groups. Appreciable risk of type II error in findings among men. |
Male: 250 | ||||||
Female: 298 | ||||||
Age >25 | ||||||
DM type 2 | ||||||
Gulliford, 2010 [37] | Cross-sectional study | 31,484 | Afro-Caribbean vs. Whites vs. Africans vs. Other blacks | United Kingdom | Clinic based screening programme | Missing data. Analysis adjusted for multiple factors. |
Male: 16,145 | ||||||
Female: 15,339 | ||||||
DM type 1 & 2 | ||||||
Khattar, 2000 [38] | Cohort study | 688 | Afro-Caribbean vs. South Asians vs. Whites | United Kingdom | Hospital and community based | Retrospective design with some degree of information bias from missing data. No mention of statistical procedures for missing data. |
Male 436 | ||||||
Female: 249 | ||||||
DM unspecified | ||||||
Leggetter, 2002 [39] | Case–control study | 528 | Afro-Caribbean vs. European | United Kingdom | Hospital based | Limitations to the quality of data collected retrospectively. |
Age >30 | ||||||
DM type 1 & 2 | ||||||
Leske,1999 [40] | Cross-sectional study | 4,631 | Black vs. White vs. Mixed | Barbados | Community based population | No stated limitations. Limited description of fata analysis. |
Male ; 1991 | ||||||
Female: 2640 | ||||||
Age: 40-84y | ||||||
DM unspecified | ||||||
Markus, 2007 [41] | Cohort study | 1,200 | African vs. Afro-Caribbean | United Kingdom | Hospital based | Case Ascertainment bias in study population. Adjustment for known confounders such as socioeconomic status reported. |
Male: 671 | ||||||
Female: 529 | ||||||
DM unspecified | ||||||
Mbanya, 1999 [42] | Cross-sectional study | 1,481 | African vs. Afro-Caribbean | Jamaica, United Kingdom, Cameroon | Community based | Relatively small sample available for British African-Caribbeans. Overall response rate of 66%. Age standardization of data for comparison across populations. |
Male: 706 | ||||||
Female: 775 | ||||||
Age: 27-74y | ||||||
DM unspecified | ||||||
Miller, 1996 [43] | Cohort study | 2,491 | Trinidad and Tobago | Trinidad and Tobago | Population based | No limitations stated. Sex specific incidence rates calculated with adjustment for age and ethnic group alone and then with additional adjustment for other factors. |
Male: 1386 | ||||||
Female:1105 | ||||||
Age 35-69y | ||||||
DM type 2 | ||||||
Molokhia, 2011 [44] | Cohort study | 832 | Trinidad and Tobago | Trinidad and Tobago | Population based | Authors reported cohort study design as the only limitation due to single village cohort. Analyses were adjusted for known risk factors and survival analysis adjusted for age and sex. |
Male 349 | ||||||
Female: 483 | ||||||
Age >20 | ||||||
DM unspecified | ||||||
Mungrue, 2011 [45] | Cohort study | 81 | Trinidad and Tobago | Trinidad and Tobago | Hospital based | Major limitation was poor record keeping and therefore the unavailability of all the data which also in part contributed to restricting the study to only one site. No survival analysis reported due to small sample size. Relevant confounders were collected and included in analysis. |
Male: 44 | ||||||
Female: 37 | ||||||
Age 10-79 | ||||||
Prasad, 2004 [46] | Cohort study | 465 | Afro-Caribbean vs. South Asians vs. Whites | United Kingdom | Clinic or hospital based study | No limitation stated. Statistical methods vaguely described. |
Male: 288 | ||||||
Female: 177 | ||||||
DM unspecified | ||||||
Riste, 2001 [47] | Cross-sectional study | 1,022 | Afro-Caribbean vs. Whites vs. Pakistani | United Kingdom | Population based register | Statistical methods included standardization for cross comparisons and log transformation carried out for variables with clearly skewed distribution. No limitations were reported. |
Male:502 | ||||||
Female: 520 | ||||||
Age 25-79 | ||||||
DM type 2 | ||||||
Sedgwick, 2003 [48] | Cross-sectional study | 1,899 | Afro-Caribbean vs. Whites vs. Black African | United Kingdom | Clinic or hospital based study | Subjects were preferentially selected from GP practices in areas with a high proportion of ethnic minorities in order to increase the representation of these groups. There was some evidence of differential non-response by ethnic minority subjects. |
Male: 409 | ||||||
Female: 390 | ||||||
Age | ||||||
DM type 2 | ||||||
Shantsila, 2011 [49] | Cross-sectional study | 128 | Afro-Caribbean vs. South Asians vs. Whites | United Kingdom | Not stated | One limitation of the study is the relatively few Afro-Caribbean subjects. There were difficulties in Afro-Caribbean subjects who met inclusion criteria, and many of them were reluctant to participate in this research. Analyses were adjusted for clinical and demographic variables. |
Male :110 | ||||||
Feamle:18 | ||||||
Age | ||||||
DM unspecified | ||||||
Sharp, 2008 [50] | Randomized controlled study | 509 | Afro-Caribbean vs. Whites | United Kingdom | clinical trial | No limitations stated. |
Male: 441 | ||||||
Female: 68 | ||||||
Age: 40-79 | ||||||
DM type 2 | ||||||
Sosin, 2008 [51] | Cross-sectional study | 108 | Afro-Caribbean vs. South Asians vs. Whites | United Kingdom | Clinic or hospital based study | Recruitment of African Caribbean subjects fell short of the numbers required from our power calculation. Limitation of cross sectional study design. |
Male: 85 | ||||||
Female: 23 | ||||||
DM unspecified | ||||||
Sargeant, 2002 [52] | Cohort Study | 728 | Jamaica | Jamaica | Population based | Lack of data for two important confounders, physical activity and diet. |
Male: 290 | ||||||
Female: 438 | ||||||
Age: 25-74 | ||||||
DM type 2 | ||||||
UKPDS-32, 1998 [53] | Cohort study | 4,974 | Mixed | United Kingdom | Clinic or hospital based study | No stated limitations. Analyses adjusted for known confounders. |
Men: 2920 | ||||||
Women: 2054 | ||||||
Age 25-65y | ||||||
Wilks, 1999 [54] | Cross-sectional study | 1,303 | Jamaica | Jamaica | Population based | No limitations stated. |
Male: 520 | ||||||
Female: 783 | ||||||
Age:25-74y | ||||||
DM type 2 | ||||||
Wilks, 1998 [55] | Other | 9772 | Nigeria vs. Caribbean vs. United Kingdom vs. United States of America | Barbados, Jamaica, Nigeria, St Lucia, United Kingdom, United States of America | Population survey | No limitations stated. |
Male:4581 | ||||||
Female: 5191 | ||||||
Age >25 | ||||||
DM type 2 |