Table 1

Description of interventions, effectiveness estimates and uncertainty assumptions

InterventionsDescription and data sourcesEffect sizes and uncertainty
Mass media campaignThis was geared towards informing populations on the ills of excess salt (sodium) consumption, the need to make and adopt healthier food choices and practices, such as reducing the amount of salt and stock cubes used during cooking and none at table (given most salt in the diet is discretionary), and where possible, select low-salt containing prepackaged foods. This would be rolled out nationwide using various media outlets, including national and local television and radio adverts three times per day; permanent billboards carrying salt awareness and risk messages; and print press, including weekly newspaper inserts and flyers.
The majority of salt reduction campaigns have been carried out in non-African settings or in combination with other strategies. Hence, to obtain the possible effect of such an intervention in Cameroon, a systematic review and Delphi study was conducted (described in detail in online supplemental file 1).
Delphi estimate for average annual reduction in salt (grams/day):
  • Men: mean=0.19, SD=0.07.

  • Women:

    mean=0.16, SD=0.04.

Uncertainty distribution: normal
School-based education programmeThe school-based education programme evaluated here is modelled from the novel and successful cluster randomised trial in China.17 In this trial, primary school children were educated fortnightly during their regular health education classes on the following key messages: harms of salt with messages such as ‘more salt less health; less salt, more health’, the recommended salt intake level of <5 g/day, and on methods to reduce salt intake such as adding less during cooking and eating less pickles. They were then encouraged to persuade adult family members, especially the one cooking the home meal, to reduce the amount of salt used; and to use alternative condiments for flavouring and replace regular salt with mineral salt. They were also given biweekly educational materials and flyers carrying the health messages of the ills of excess salt intake to take home as well.
We modelled the potential impact of this intervention in Cameroon, for all schools nationwide, and scaled the effect for all adults, including those without children.
Estimated salt reduction (g/day) from 24 hours urine in adults (difference between groups):
95% CI −3.7 to −2.2
Reported to cause a 25% relative reduction in salt intake
Given the difference in average salt intake levels between Cameroon and China (higher), we used the relative reduction in our model.
Uncertainty distribution: normal
Salt substituteThis intervention essentially assessed the impact of switching from a 100% NaCl or salt to a mix of NaCl and KCl. Evidence from a meta-analysis of randomised controlled trials (Peng et al 2014)44 was used to inform our effectiveness estimate. Four out five trials included in this meta-analysis compared normal salt (100% NaCl) with a salt substitute (65% NaCl, 25%KCl, 10%MgSO4). Given that about 75.4% of sodium in Cameroonian diet is from salt and stock cubes,12 we assume the intervention applies to just cooking salt and stock cubes (and not bread, cereals and milk).Given the sodium content in the trials was reduced from 100% to 65%, we estimate a 35% reduction, adjusting to target sources (salt and stock cubes) in Cameroon.
Final effect: −26.4%
SD=20% of point estimate
Uncertainty distribution: normal
  • KCl, potassium chloride; NaCl, sodium chloride.