Table 2.

Ethnical and geographical factors investigated in the systematic review

Ethnical and geographical factors	Subgroups	Direction of the association with antidiabetic medication adherence
		Increase	Equal	Decrease
Ethnicity/race	Ethnicity	CS: 1 [121]	CS: 5 [39, 43, 72, 107, 122] RCT: 2 [123, 124]
	Non-white			CS: 1 [60] C: 2 [104, 125]
	Non-European			C: 1 [126]
	African-American			C: 1 [127]
	Black			C: 4 [83, 86, 88, 128] CC:1 [90] I: 2 [51, 89]
	Asian	CS: 1 [36]
	Malay			CS: 1 [111]
	Indian	CS: 1 [111]
	Chinese		C: 1 [124]	CS: 1 [99]
	Japanese		C: 1 [124]
	Filipino			C: 1 [124]
	Saudi Arabian	CS: 1 [101]
	Arab non-Emirati	CS: 1 [36]
	Pacific Islander			C: 3 [35, 122, 124]
	Maori			C: 3 [35, 37, 122]
	Latin-American/Hispanic			CS: 1 [121] C: 2 [88, 129]
	Native Hawaiian			CS: 1 [124]
Country of birth	Foreign-born		CS: 1 [85]
	US-born			CS: 1 [71]
Acculturation	Acculturation	CS: 1 [39]
	Believe in traditional Chinese medicine			CS: 1 [39]
Accessibility to health care	Accessibility to health care*
	Distance to healthcare provider			CS: 3 [31, 61, 101]
Area of residence	Area of residence		CS: 5 [31, 80, 98, 100, 130] C: 1 [122]
	Rural	C: 1 [104]
	Urban			CS: 2 [45, 58]
	IMD quintile	C: 1 [125]
	Neighbourhood deprivation			CS: 2 [37, 77]
	Socioeconomic living area		C: 1 [126]
Regional differences	Geographical area	C: 1 [104]	CS: 2 [41, 110]
	Southern United States			C: 1 [95]

Factors marked with *—a study has found an association, but did not state in which direction [105]

Type of study—CS cross-sectional, C Cohort, I interventional, RCT randomized controlled trial

Bolded numbers—Number of articles

Number in brackets—References to the articles

IMD indices of multiple deprivation