Reducing health harms of foods high in fat, sugar, or salt: consultation analysis

1. Introduction and background 

1.1. The Scottish Government published A Healthier Future: Scotland’s Diet and Healthy Weight Delivery Plan (the Delivery Plan) in July 2018^[4] following consultation.

1.2. The Delivery Plan set an ambition to halve child obesity in Scotland by 2030 and to significantly reduce health inequalities. It set out a range of actions to improve the nation’s diet-related health outcomes, working towards achieving five key outcomes: 

i. Children have the best start in life – they eat well and have a healthy weight
ii. The food environment supports healthier choices
iii. People have access to effective weight management services
iv. Leaders across all sectors promote healthy weight and diet
v. Diet-related health inequalities are reduced.

1.3. In October 2018, the Scottish Government consulted on plans to restrict the promotion and marketing of targeted foods high in fat, sugar or salt (HFSS) where those foods are sold to the public. This consultation closed in January 2019 and its analysis is the subject of this report.

Policy Context

1.4. The Scottish Health Survey^[5] provides information on the health and factors relating to health of people living in Scotland. In 2017, it found that

• 65% of adults (aged 16 and over) are considered overweight of which almost half are obese (29% of adults); 
• 26% of children (aged 2 to 15) are at risk of being overweight of which almost half are at risk of obesity (13% of children);
• 24% of adults and 15% of children^[6] eat the recommended 5-a-day portions of fruit and vegetables; and
• Children in Scotland consume higher fat and/or sugary food/drink more often than adults.

1.5. The Foods Standards Scotland Situation Report, The Scottish Diet: It needs to change (2018 update)^[7], highlighted that too much of the Scottish diet is made up of discretionary foods:

• 20% of all calories and fat consumed comes from discretionary foods;
• 50% of the sugar consumed comes from discretionary foods; and
• 21% of sugar intake comes from sugary drinks.

1.6. Poor diet and being overweight have profound impacts on health. Obesity is the second largest preventable cause of cancer, linked to over 2,100 cases of cancer per year in Scotland.^[8] Being overweight and obese is the most significant risk factor for developing Type 2 diabetes^[9], which can result in an increased risk of other types of diseases, notably cardiovascular disease and hypertension.^[10] Diet-related diseases pose a substantial burden to the NHS, with the annual cost estimated to be between £363 million to £600 million. Factoring in labour-market related costs (e.g. loss of productivity), the total annual cost to the Scottish economy is estimated much higher at between £0.9 billion and £4.6 billion.^[11] As well as reducing life expectancy (anywhere from 2-10 years depending on body mass index^[12]), overweight and obesity are considered to hinder the population’s ability to live happy and fulfilling lives.^[13]  

The Consultation 

1.7. The consultation paper set out proposals for restricting the promotion and marketing of targeted HFSS foods where those foods are sold to the public. The primary aim of the policy is to reduce the public health harms associated with the excessive consumption of calories, fat, sugar and salt. The consultation paper set out the rationale for this approach, including that evidence suggests promotions and marketing encourage purchases that would otherwise not have occurred.

1.8. The consultation paper proposed to restrict ‘discretionary’ foods, defined as food and drink categories typically high in fat, sugar or salt, but providing little or no nutritional benefit necessary for a healthy diet. The proposals relate to the in-store promotion and marketing of discretionary foods where sold to the public, including retail and the Out of Home (OoH)^[14] sector.  The consultation paper asked whether restrictions should also apply online.

1.9. The consultation opened on 2^nd October 2018 and closed on 9^th January 2019. The consultation contained 17 questions – 10 were open questions and 7 were multiple choice (closed) questions. For all questions, space was provided for respondents to provide an explanation for their answer. The questions covered a number of themes:

• Whether mandatory measures should be introduced to restrict the promotion and marketing of HFSS foods (Question 1)
• Identifying categories of food and drink that would be targeted (Questions 2, 3 and 4)
• Forms of promotion and marketing that would be restricted (Questions 5 and 6)
• Places the restrictions would apply (Questions 7 and 8)
• Exemptions to the restrictions (Questions 9, 10 and 11) 
• Enforcement and implementation (Question 12)
• Legislative framework (Question 13)
• Impact and support (Questions 14, 15 and 16)
• Any other comments (Question 17).

1.10. Annex 1 contains a full list of the consultation questions.

Aims of this report

1.11. The overall aim of this report was to gain some sense of the balance of opinion among respondents towards the proposals, whilst also attempting to come to a clear realisation about the breadth and detail of arguments put forward both for and against the proposals.^[15]

1.12. The structure of the report follows the structure of the consultation paper, considering the response to each question in turn. Section 2 provides an overview of the respondents to the consultation.^[16] Sections 3-10 summarise the views submitted by respondents in answer to the consultation questions. Finally, Annexes 1-6 provide further detail on the consultation questions, the responses, the respondents and issues with interpretation of the consultation questions. 

Approaches to analysis 

Quantitative analysis 

1.13. Frequency analysis was undertaken for all multiple choice (closed) questions. The figures are included within the findings and presented with additional detail in Annex 2.^[17] Further analysis was also undertaken to compare how respondents’ tick box response to the first question (whether or not to restrict mandatory measures) related to their responses to other tick box questions.^[18] These figures can be found in Annex 2 and a brief summary of the findings in Annex 6.

1.14. Prior to the main analysis, responses were reviewed for blanks, duplicates, campaign responses, missing answers and any problems with responses such as misinterpretation of the questions. The purpose of this was both to avoid the findings being skewed and to ensure the views of respondents were represented as accurately as possible.^[19]

1.15. The full justification for cleaning the data and a full list of issues with the questions, and the solutions applied is provided in Annex 3. As part of the process to ensure that the views of respondents were represented as accurately as possible, it should be noted that the frequencies and percentages in the tables for each question in the main report (sections 3-10) include answers that were corrected. To be clear, corrections took place where there were clear discrepancies between the tick box answer and the comments that were provided alongside it. This includes instances where the respondent had clearly inferred a view in their comments that did not match the tick box answer they had selected (e.g. where a respondent had found elements of the wording for the closed questions difficult to understand, and provided a tick box answer which did not match their clearly expressed written view below). In these cases, the data cleaning process corrected the tick box response to be in line with the comment, and incorporated these responses into the tables in the main report. The tables also include an additional category referred to as ‘conflicting answers’ or ‘possible conflicting answers’. These reflect the number of responses where there was doubt about the certainty of the tick box response on the basis that it did not match the point of view expressed in the comments but was not sufficiently clear to enable correction. Further information on these responses is provided in Annex 3 and a more detailed breakdown of these responses is provided for each question in the tables in Annex 2.

1.16. The overall impact of cleaning the data was reviewed. The impact varied from question to question - sometimes increasing the number of responses in favour of proposals; sometimes increasing the number of responses against proposals. However, data cleaning did not result in an overall shift in stance among respondents (i.e. the overall proportion for and against restrictions), nor did it alter the trends in the data (i.e. the relative popularity of each answer option within a question remained the same). 

Qualitative analysis

1.17. The responses given to the open questions and the comment sections of the closed questions were analysed to identify the main themes emerging, the range of views expressed for each question and how views varied by respondent type. In general the responses submitted by organisations were more detailed and lengthy than those submitted by individuals.^{[20, 21]}

1.18. Due to the varying numbers of responses to each question and each answer option, descriptors are used to represent the proportion of respondents that hold a particular reason for each answer option. Table 1.1 displays the descriptors used.

Table 1.1 Descriptors

Descriptor	Value
Most	Over 75%
Many	51 - 74%
Half	50%
Some	25 - 49%
Few	Under 25%

1.19. For example, 40% may have disagreed with a given question. Among those that expressed a reason for disagreement, if over 75% shared the same reason for disagreement this would be referred to as ‘most’ followed by an explanation of the reason e.g. ‘of those that disagreed, most thought it was not needed’.

Comment on the generalisability of the consultation findings

1.20. It should be noted that the purpose of consultation is to understand the range of views expressed and the reasons these views are held. However, caution is needed when interpreting the responses. Anyone could submit their views, and individuals and organisations who have a keen interest in the topic and the capacity to respond were more likely to participate than those who do not. This self-selection means that the balance of views of consultation participants cannot be generalised to the wider population. It is not a reliable measure of public opinion.

Interpretation of the findings

1.21. While efforts have been made to explain and summarise the questions posed within the consultation paper, it should be noted that the full rationale and evidence base for the proposals is not presented within this report. It is therefore advised that the full consultation paper is read in advance of this document to aid interpretation of the findings presented here. It should also be noted that generally those agreeing with the proposals within the consultation paper were less likely to explain their reasoning and gave less detail than those who disagreed with proposals.^[22]