Table 2.

Secondary studies with behavioural interventions included in the umbrella review.

Author (Year)	Description of the Intervention	Results	Implications
Allan et al. (2016) [15]	Environmental intervention (environmental intervention) affecting eating habits	For behavioural endpoints, 13 of 22 studies showed a significant effect on primary endpoints. For physical endpoints, some studies showed no difference in BMI or body weight, while others confirmed it.	The current state of knowledge does not allow for clear recommendations for introducing environmental interventions to change eating habits in the workplace.
Brown et al. (2017) [30]	Workplace well-being programs to prevent or treat diabetes (nutrition—cooking workshops, individual dietary consultations, dietary changes; physical activity pedometers, workout plans; smoking cessation; usually in combination)	The study demonstrated a steady improvement in health in biological measures, self-reported behavioural adherence measures, and psychosocial variables. The authors presented data that showed improvement in most cases.	Workplace diabetes prevention programs can be useful in reducing disease occurrence and progression, but better design of interventions is needed. Employer education and further research in this area are crucial.
Geaney et al. (2013) [16]	Change in the composition of meals available at work, change in portion sizes (usually reduction), changes in access to healthy products for employees.	All of the included studies showed changes in fruits and vegetables intake, but none showed an effect size greater than a half-portion increase in fruit and vegetables consumption.	Modification of workers’ meals may increase fruit and vegetables intake, but the strength of evidence is low.
Park et al. (2019) [37]	A nutritional intervention that limits the intake of energy and certain nutrients (carbohydrates or fats) or a balanced diet that ensures a normal supply of all nutrients	Employees’ body weight decreased significantly: WMD of −4.37 kg (95% CI −6.54 to −2.20; Z = 3.95, p < 0.001), so did BMI: WMD of −1.26 (95% CI −1.98 to −0.55) kg/m2, but it was statistically significant (Z = 3.47, p = 0.001), blood cholesterol and blood pressure values also declined—but the problem is the duration of the study and the quality of the data.	It is challenging to definitely state the effectiveness of interventions, but it is a good start for further research.
Sawada et al. (2019) [17]	Discounts on healthy food products or for a smaller portion ordered in the employee cafeteria, colour-coding of dishes (yellow, green and red),	No significant changes in BMI, blood cholesterol levels or changes in diet	Link between the intervention and the outcomes cannot be established; poor quality of evidence; a need for further research in this area.

CI—confidence interval; p—probability value; WMD—weighted mean difference; BMI—body mass index; Z—z-score statistics.