Skip to main content


Table 2 Included studies

From: Salt intakes in sub-Saharan Africa: a systematic review and meta-regression

Country Period of data collection Population and Quality of Sample Representativeness Method and Quality of Data Collection References
Benin January-February 1996 Male school children aged 6–12 and their mothers in rural Penessoulou, Atacora. (B). U3 Melse-Boonstra et al, 1998 [19]
Botswana October 1967, May 1968 Male adults recruited from the Kung bushmen in Northwestern Ngamiland. (B). U4 Truswell et al, 1972 [20]
Cameroon 1993-1995 Random sample recruited from the civil service in Yaounde (affluent urban) and villages in a district in the same region (rural poor). Men and women aged 25–74. (A). U4 Cooper et al, 1997 [21]
Cote d’Ivoire Not reported Families chosen by the investigators as representative of the populations recruited from 3 Northern villages (Koro, N’deo and Maranama) and from Abidjan. Children aged 2–12 and adults. (B). D4 (all food eaten by the families for 3 consecutive days was weighed by investigators). U3 (adults only). Hess et al, 1999 [22]
Ghana February-April 2007 Random sample selected from a population register recruited from rural Kassena-Nankana District. (A). D4 (household inventories) Kunutsor and Powles, 2009 [23]
Ghana June 2001-June 2002 Random sample selected from census, recruited from 12 villages, 6 semi-urban, 6 rural in the Ejisu-Juabeng and Kumasi Districts. Adults aged 40–75, 92 % Ashanti tribe, 94 % Twi-speaking. (A). U4. Cappuccio et al, 2006 [24]; Kerry et al, 2005 [25]
Ghana June 1997-June 1999 Patients aged 5–12 in the paediatric surgery unit of the Korle-Bu Teaching Hospital, Accra, with surgical conditions that did not affect urine volume (herniae, hydroceles, undescended testes, hypospadias and tumours) and post-operative patients after surgery for acute appendicitis who had fully recovered. All eating a normal diet. Ready for discharge but detained for 24 h urine to be collected. (B). U4 Badoe et al, 2005 [26]
Ghana Not reported Patients recruited from Korle-Bu Teaching Hospital, Accra, suffering from uncomplicated hernias, nodular goitre, breast tumours and simple tumours eating a full hospital diet. Adults aged 20–69, all ambulant, no evidence of urinary disease. (B). U4 Badoe and Osafo, 1971 [27]
Kenya 1986 Random sample drawn from 320 households in the rural villages of Rambugu and Ndori, north of Lake Victoria in Western Kenya. Exclusively from the Luo tribe. Pregnant women excluded. (Part of INTERSALT). (A). U3 Carvalho et al, 1989 [28]
Kenya September 1980-November 1984 Recruited from two rural populations (the Luo tribe and the Kamba tribe) and migrants from the Luo community to Nairobi (urban migrants). Only spot urine done on Kamba population, therefore not included in this systematic review. (B). U4 Poulter et al, 1985 [29]
Malawi Not reported All inhabitants from the first building estate in Lilongwe (urban, mainly white collar workers) and from the villages of Gunde and Msinje (rural, mainly farmers). Urine requested from men aged 15+ with even subject numbers. (B). U4 Simmons et al, 1986 [30]
Nigeria Not reported Recruited from residential communities in South-western Nigeria. Excluding pregnant women and those on anti-hypertensives. (B). U3 Tayo et al, 2012 [31]
Nigeria November 2006-January 2007 All hypertensive patients attending the hypertensive clinic and selected normotensive controls attending the outpatient departments of two specialist hospitals in Akure and Ondo towns. Adults aged 20+. Normotensive controls only considered in this review. (B). D3. U5. Ijarotimi and Keshinro, 2008 [32]
Nigeria Not reported Recruited from two rural communities in Southwest Nigeria Igbo-Ora and Idere. Excluding pregnant or breastfeeding women and people with diabetes, kidney disease or atherosclerosis and BMI over 40. Normotensive men and women. Aged 25–55. (B). U4 Forrester, 2005 [33]
Nigeria Not reported Free-living adult volunteers aged 18–48. University staff and/or their dependents of lower to high socio-economic status by Nigerian standards. (B). D1 Smith, 1988 [34]
Nigeria 1993-1995 Random sample recruited from the rural village of Idere and 2 sites within Ibadan: Idikan traders and crafters (urban poor) and male pensioners of the Nigerian Railway Corporation (urban salaried). All Oyo Yoruba aged 25–74. (A). U4 Kaufman et al, 1999 [35]; Cooper et al, 1997 [21]; Kaufman et al, 1996 [36]
Nigeria Summer 1990 All civil servants in Sokoto, Northern Nigeria excluding pregnant women. Aged 18–66. (B). U5 Bunker et al, 1996 [37]
Nigeria Not reported Random sample from two secondary day schools and one primary day school in Calabar (urban) and one co-educational secondary day school and one primary day school in Akpabuyo (rural). Boys and girls aged 12–14. (A). U3 Ekpo et al, 1990 [38]
South Africa Not reported Random sample selected from housing map, recruited from rural communities of Empangeni, KwaZulu-Natal. Adults 19+ excluding pregnant or breast feeding women. (A). D3 Kolahdooz et al, 2013 [39]
South Africa 2002-2006 Random sample selected from recent census, recruited from metropolitan areas of Johannesburg. Adults aged 17+ of black African descent. (A). U2 Maseko et al, 2006 [40]; Millen et al, 2013 [41]; Redelinghuys et al, 2010 [42]
South Africa 2002 Convenience sample recruited from staff of the Cape Town City Council offices. Including hypertensive and normotensive adults aged 20–65 years old from three ethnic groups (black, white, mixed). Normotensive population only considered in this review. (B). D1. U1. Charlton et al, 2013 [43]; Charlton et al, 2008 [44]; Charlton et al, 2005 [45]; Charlton et al, 2005 [46]
South Africa March 1981-October 1982 Adult patients with hypertension and normotensive controls recruited from two light industrial firms (urban Zulus and urban Indians), from the Lamontville township and the outpatient department of King Edward VIII hospital (urban Zulus), from a satellite clinic of KEVIII hospital (urban Indians) from Bethesda hospital in Ubombo, Benedictine hospital in Nongoma and others from the same communities (rural Zulus). Patients with complications of hypertension or major concomitant illness were excluded from the study. Normotensive controls only considered in this review. (B). U5 Hoosen et al, 1985 [47]
South Africa Not reported Healthy male volunteers resident in Johannesburg. Aged 20–30 years. (B). U2 Barlow et al, 1985 [48]
South Africa 1980 Random sample selected from male employees of a wire-rope manufacturing company near Johannesburg. Aged 30–50 years. (A). U2 Barlow et al, 1982 [49]
South Africa Not reported Normotensive factory workers and hospital staff and hypertensive patients. Normotensive population only included in this analysis. (B). U3 Cohen et al, 1982 [50]
South Africa August-September 1978 Urban, apparently healthy, informed volunteers from domestic and clinical staff of the Groote Schuur Hospital, Cape Town and rural Xhosa volunteers from villages surrounding St Lucy’s Mission Hospital in the Transkei. Excluding those taking any drug (including oral contraceptives). (B). D4 (“a full dietary history”) Sever et al, 1980 [51]
Tanzania Not reported 13–21 year old Bantu-speaking boys/young men from a secondary boarding school in Mafinga, Iringa District. (B). D4 (analysis of weekly menu) Rebacz-Maron et al, 2013 [52]
Tanzania Not reported 20–50 year old healthy men living in Mwanza (urban). (B). U2 Hamada et al, 2010 [53]
Tanzania 1998 Random sampling from administrative lists recruited from urban (Dar es Salaam), rural (Handeni) and semi-nomadic (Monduli) communities. Adults aged 47–57. (A). U2 Njelekela, 2001 [54]
Tanzania 1987 Random sample recruited from Dar es Salaam (urban), Handeni (rural) and Moduli (nomadic-rural). Aged 30–54. (A). U4 Mtabaji et al, 1990 [55]
Tanzania and Uganda Not reported General population cohorts from Lugarawa district in Tanzania and Lugbara in Uganda. (B). D4 (detailed questionnaire about dietary habits) Pavan et al, 1997 [56]
The Democratic Republic of the Congo December 1983-May 1984 10 % random sample of a quarter of Kinshasa. Aged 10+. Subjects on anti-hypertensives were excluded. (A). U2 M’Buyamba-Kabangu et al, 1986 [57]; M’Buyamba-Kabangy et al, 1986 [58]
Zimbabwe Not reported Male school children in rural Zimbabwe. (B). U4 Matthews and Pegge, 1997 [59]
Zimbabwe Not reported Volunteer first year medical students. Male and female, average age 20 years. (B). U2 Mufunda et al, 1992 [60]