Is a diet low in greenhouse gas emissions a nutritious diet? – Analyses of self-selected diets in the LifeGene study
© The Author(s). 2017
Received: 1 June 2016
Accepted: 16 February 2017
Published: 10 April 2017
Climate change is an urgent global issue and the food sector is a major contributor to greenhouse gas emissions (GHGE). Here we study if a diet low in GHGE could be a nutritious diet compared to the Nordic Nutrition Recommendations (NNR).
The environmental impact of foods from Life Cycle Assessment (LCA) data was linked to a food frequency questionnaire (FFQ) filled out by 5,364 participants in the Swedish LifeGene study. Thereafter, we calculated the daily emission of CO2 equivalents (CO2e) as well as the intake of selected nutrients associated with vegetables, fruits, meat and dairy products. The CO2e was divided into quartiles were quartile 1 corresponds to a diet generating the lowest CO2e, and quartile 4 corresponds to a diet with the highest CO2e.
The overall diet-related emission was 4.7 kg CO2e/day and person, corresponding to 1.7 ton CO2e/year. In general, there were only small differences in nutrient intake between groups of varying levels of CO2e, regardless if the intake was analyzed as absolute intake, energy percent or as nutrient density. Moreover, adherence to NNR was high for the group with the lowest CO2e, except for saturated fat where the intake was higher than recommended for all CO2e groups. On the other hand, only the group with the lowest CO2e fulfilled recommended intake of fiber. However, none of the CO2e groups reached the recommended intake of folate and vitamin D.
Here we show that a self-selected diet low in CO2e provides comparable intake of nutrients as a diet high in in CO2e.
KeywordsDiet Nutrients Carbon dioxide equivalents Greenhouse gas emission Life cycle assessment Nordic Nutrition Recommendations LifeGene study
Substantial reductions of greenhouse gas emissions (GHGE) are needed if the global warming should be limited to the UN target of a maximum of 2 °C, compared to pre-industrial times, and dietary shifts towards a more climate friendly diet are one of several strategies to reduce emissions from the food sector [1–3]. However, when promoting a diet low in GHGE, it is important to consider health aspects of the diet and adherence to dietary recommendations. Previous studies based on simulated dietary scenarios have reported that a diet low in GHGE can be nutritious at the same time [2, 4, 5], but among studies based on self-selected diets [6–9], only one reported GHGE in relation to the intake of macronutrients and fiber . Therefore, more research is needed to study intake of vitamins, minerals, macronutrients and fiber among individuals with a diet low in GHGE .
The production of food is estimated to contribute to 25% of the total GHGE in the world, more than the total emissions from the transport sector . However, there is a substantial variation in GHGE for different food products and the production of food items from animal origin is generally associated with higher GHGE than plant-based food items, such as vegetables, whole grains and legumes. Particularly, meat from ruminants, such as cattle and sheep, are associated with high emissions due to their less efficient feed-conversion rate and to the methane produced during digestion [13, 14]. Our collective food choices have significant impact on global GHGE and a dietary shift may significantly impact public health.
Here we estimate diet-related GHGE in the Swedish LifeGene study using a Food Frequency Questionnaire (FFQ) linked to Life Cycle Assessment (LCA) data of carbon footprint for food products representative for Swedish food habits. The overall aim is to investigate if a diet low in GHGE can be a nutritious diet. To the best of our knowledge this is the first Swedish study reporting the intake of several nutrients from a self-selected diet low vs. high in GHGE. Second, it is the first study relating GHGE to the Nordic Nutritional Recommendations (NNR)  on an individual level for nutrients associated with meat, dairy, fruits and vegetables.
The LifeGene study is a prospective Swedish cohort study aiming at combining advances in modern biotechnology with information on individual’s health and lifestyle . The target enrollment in LifeGene is 300,000 Swedes, with the projected follow up of 20 years . The present study is based on the pilot phase of the LifeGene study, which was launched in Stockholm in October 2009, followed by Umeå in November and Alingsås in January 2010, and invitations were sent out until March 31, 2010. In total, 42,700 women and men age 18–45 years old were randomly selected through the national population registry and invited to the study. An invitation letter with personal login information was sent out, including up to three reminders. After the study participants agreed to participate and left consent on the LifeGene web page, they were asked to respond to a comprehensive interactive web questionnaire at home, including questions on lifestyle factors, self-care, women’s/men’s health, living habits, health history, injuries, asthma and allergy, mental health, home and work. Thereafter, an appointment at one of the test centers was made for in-person testing. 7,818 filled out part of, or the whole questionnaire, and 6,633 visited the test center. The Research Ethics Review Board at Karolinska Institutet approved the present study.
Dietary assessment method
Diet was assessed using the interactive web-and meal based FFQ called Meal-Q, described in detail elsewhere [18, 19]. Meal-Q assesses habitual dietary intake during the previous months and includes 102–174 food items, dishes, and beverages, depending on the number of follow-up questions, as well as questions about supplement use, meal patterns, and eating behavior. Participants were instructed to choose among predefined food items and intake frequencies and report on all items that were consumed at least once a month. Five photos of different portion sizes were included for 1) rice, potatoes and pasta, 2) meat, chicken, fish and vegetarian substitutes and 3) vegetables (raw or cooked) and used to calculate portion sizes for cooked dishes and vegetables whereas a standard portion sizes were used for all other food items. A program tailored for Meal-Q called NutriCalc was used to link dietary data to the national food composition table from the Swedish National Food Agency  to generate the daily intake of energy and nutrient per person. Meal-Q has been validated using 7-day weighed food records with regards to nutrients and energy as well as doubly labeled water with regards to energy, demonstrating good validity and reproducibility [18, 19].
We identified published LCA data for 65 food items and food groups representative of typical food consumption in Sweden described in detailed elsewhere and about half of the LCA data came from the same source . These 65 food items and groups matched the food items/questions in Meal-Q, for example, we asked for 3 kinds of bread in Meal-Q (white, whole grain and crisp, respectively), but applied the same LCA data for all 3 kinds of bread. Assessment of the combined impact of different greenhouse gases was achieved using Global Warming Potential (GWP) with a 100 years perspective expressed as kg carbon dioxide equivalents (CO2e) per kg of food product. The GWP used to calculate the CO2e was 1 for carbon dioxide, 34 for methane and 296 for nitrous oxide .
The GHGE include emissions from agriculture and its inputs, food processing, distribution and retailing . If LCA studies did not include emissions from distribution and retailing, emission were imputed by adding emissions linked to retail, transportation and packaging using Swedish data . Emissions after the retail phase were not included, such as transports to the household, storing and cooking, as well as from waste management.
Portion sizes in Meal-Q were based on food ready to be eaten and therefore we recalculated LCA data for uncooked food to CO2e per kg cooked food when needed, considering both hydration, i.e. cooking of rice, and dehydration, i.e. cooking of meat . In addition, we adjusted for unavoidable food losses (i.e. shell and bone) using data from the Swedish food composition database  and avoidable food waste both before and after food preparation using data from the British Waste and Resources Action Programme  and a FAO report .
CO2e for mixed dishes was based on up to three main food products or groups and weighed using standard recipes from the Swedish food composition database , for example, lasagna was based on weighted LCA data from ground meat, milk and tomato, respectively. Thereafter, data on CO2e per kg food item were linked to all food items in Meal-Q by the NutriCalc program to calculate daily CO2e per person. The assessment of CO2e by Meal-Q was validated using 7-day weighed food records and the Spearman correlation coefficient between CO2e from Meal-Q and the 7-day weighed food records were r = 0.70 (95% CI 0.61–0.77), whereas 90% were categorized into the same/adjacent quartile in cross-classification analyses . The intraclass correlation coefficient for the reproducibility of Meal-Q was 0.81(95% CI 0.73–0.87), and 94% were categorized into the same/adjacent quartile in cross-classification analyses .
Nordic nutritional recommendations (NNR)
Description of the recommendations in the Nordic Nutrition Recommendations (NNR) from 2012
Goal for menu planningb
Saturated fat (E%)d
Monounsaturated fat (E%)d
Polyunsaturated fat (E%)d
Vitamin C (mg)
Vitamin B12 (μg)
Vitamin D (μg)
Retinol equivalents (RE)h
The in-person clinical testing included measurements of weight, height, waist, hip and chest circumference, bioimpedance, heart rate and blood pressure along with audiometry and spirometer. Blood and urine samples were taken for analyses and biobanking.
Participants with energy intake less than 3,300 or more than 21,000 kJ were excluded (n = 212). The purpose of the cut-off is to exclude participants with implausibly high or low total calorie intake, thus, improving the quality of the data that is being analyzed. Quartiles were used to split CO2e into four groups, both for crude values and energy adjusted values using the residual method  and quartile 1 corresponds to the group with the lowest CO2e, and quartile 4 to the group with the highest CO2e. Median and interquartile range (25th–75th percentile) of the nutrient intake divided by CO2e groups was calculated and the difference tested with Kruskal-Wallis test. To show the distribution within the energy adjusted CO2e groups, the nutrients are presented as boxplots in Figs. 2 and 3. The notch corresponds to the median, the edges of the box correspond to the first quartile (q 1) and third quartile (q 3). The vertical lines at the end of the dotted line are the lower and upper adjacent value (LAV and UAV) here calculated as follows: LAV = smallest value which is ≥ q 1−4 IOR; UAV = largest value which is ≤ q 3 + 4 IOR, where IQR is the interquartile range (IOR = q 3−q 1). Due to large sample size we chosen 4 instead of commonly used 1.5 to highlight extreme observations to make more distinguishable graphs. The extreme observations, values below the LAV or above UAV, are marked as circles. All analyses were performed in the statistical software STATA version 13.1. Significance level was set to α = 0.05.
Characteristics of the participants in the Swedish LifeGene study in 2009–10
(n = 3,239)
(n = 2,125)
(n = 5,364)
Servings of beef/daya
Servings of meat/dayb
Servings of dairy products/dayc
Servings of dairy products and dishes/dayd
Median nutrient intake according to daily diet-related greenhouse gas emissions in the LifeGene study, 2009–10, Sweden
215. 9 (97.6)
Saturated fat (g/d)
Monounsaturated fat (g/d)
Polyunsaturated fat (g/d)
Vitamin C (mg/d)
302. 3 (146.8)
Vitamin B12 (μg/d)
Vitamin D (μg/d)
Retinol equivalents (RE/d)
The group with the lowest CO2e emissions had adequate intake of most nutrients, and the intake was comparable with the nutrient intake among people in the highest CO2e emissions, suggesting that a diet low in GHGE can be nutritious at the same time. Although the intake of some important nutrients increased with increasing emissions in our study, there were only small differences between groups with varying CO2e levels and the effect was less pronounced compared to a study by Vieux et al 2013 where an indicator of good nutritional quality dietary was significantly higher for a diet high in GHGE compared to a diet low in GHGE .
Compared to the NNR , the median intake of micronutrients was generally near or above the recommended intakes when analyzed as nutrient density in all CO2e groups in the present study. The trend was similar for energy percent of macronutrients, where the intake was within recommended levels, except for saturated fat which was higher than recommended. In general, all CO2e groups fulfilled the recommended intake of iron, B12, zinc, retinol equivalents, nutrients mainly found in animal products. Highest intake of B12 and zinc were found in the group with the highest CO2e, whereas there were no differences for iron and retinol equivalents. However, the intake of folate did not reach the recommended level for women in reproductive age, which is a majority of the females in this study. Nor did any of the groups have enough intake of vitamin D, a common problem in the general population in Sweden.
This study has several strengths and limitations that should be considered. A main strength is the use of individually assessed dietary information from a large population-based study, which allows for analyses of individual variability in CO2e, whereas many studies in this area rely on simulated dietary scenarios, such as comparisons between meals, dietary patterns or national consumption statistics [4, 26, 27]. Also, the magnitude of the CO2e reported in this study is comparable to Bryngelsson et al 2016 , where similar LCA data were linked to consumption data based on national statistics from Sweden. Moreover, the daily emissions of CO2e was comparable to other studies based on individually assessed dietary intake using FFQ [6–8]. The dietary assessment method that was used, Meal-Q, was developed and validated by us for the LifeGene study [18, 19], and showed strong validity and reproducibility . A limited number of studies have analyzed CO2e from self-selected diets assessed in surveys or cohort studies [6–9], but to our knowledge, this is the first study based on a Swedish population. Also, this is the first observational study looking at the relation between CO2e, and adherence to NNR with regards to individual nutrients. This study therefore contributes with substantial new knowledge about a diet low in GHGE based on a self-selected diet.
Differences in assumptions and methodologies between LCA studies, such as allocations and system boundaries, make comparisons between studies complicated. Product specific variations, such as use of fuel for transports, fodder for the animals, electricity mix etc. have impacts on emissions linked to each product. Moreover, the functional unit may differ, for instance, if CO2e is expressed for meat with or out without bones. In this study, we have taken several steps to ensure that the LCA data accurately represent food consumption in Sweden. First, the result is based on a large number of LCA studies (65 food groups) that matches the food items in dietary questionnaire . Second, most LCA data used in this study have the same system boundaries and surrounding system. Thirdly, the LCA data were re-calculated taking into account weight change during cooking. Finally, we included avoidable and unavoidable waste on a household level using different values for different products [23, 24], for example, the proportion of waste is greater for fresh foods than for staple food. Although, the total estimated emissions per person may be somewhat biased, it allows us to rank individuals and make comparisons between groups of people with high and low emissions, respectively  and extensive sensitivity analyses show that the results are robust and did not differ when CO2e was divided into tertiles, quartiles or quintiles.
The cross-sectional design of the study is a limitation. Also, systematic bias may be introduced when linking data on emissions to dietary information. The FFQ is designed to capture most of the diet and the food items listed in the questionnaire is representative for main stream food habits in the population, i.e. it does not capture the whole diet . Moreover, for mixed dishes, such as lasagna, we rely on standard recipes  and used weighted LCA data for up to three main food products or groups . Also, underreporting is one of the most challenging problems with all self-reported dietary assessments, and some foods are thought to be underreported to a greater extent than others, as well as vary between groups . Thus, the absolute CO2e is therefore underestimated in the present study as compared to if a more extensive dietary method had been used .
In conclusion, the magnitude of the diet-related CO2e in the present observational study is in line with result from other studies, both scenarios and self-selected diets. Moreover, it shows that a self-selected diet low in CO2e provides comparable intake of nutrients associated with vegetables, fruits, meat and dairy, as a diet high in CO2e. Also, it shows that a diet low in CO2e adhere to dietary guidelines for most nutrients. This opens up for a future win-win situation between a diet low in GHGE and a nutritious diet. Our collective food choices have significant impact on global GHGE and in order to reduce climate impact from food, the consumption of meat, in particular beef, should be reduced and the consumption of plant-based foods, such as whole grains, legumes, vegetables and fruit increase.
Body mass index
Carbon dioxide equivalents
Food frequency questionnaire
Greenhouse gas emission
Global warming potential
Life cycle assessment
Nordic nutrition recommendations
We would like to thank the participants in the LifeGene study as well as Jonas Karlström for invaluable assistance with the calculations of CO2e and nutrients.
The present work was supported by grants from the Swedish Research Council, The Alpro Foundation and Riksbankens Jubileumsfond, and the data collection of LifeGene was supported by Torsten and Ragnar Söderberg’s Foundation, and AFA Insurances.
Availability of data and supporting materials
The database on CO2e used to calculated daily CO2e has been published by us and is found in Sjörs et al. Environmental Health (2016).
The dataset, i.e. data based on questionnaires, used for this manuscript is not publically available at this time. However, the LifeGene study is national resource and researchers may apply directly to LifeGene to access the data from the ongoing study.
CS and FH were responsible for collecting LCA data; CS recalculated LCA data and linked LCA data to food items; KB developed and validated of the calculation program NutriCalc; AT, AS, and KB was responsible for statistical analyses and interpretation of results; KB drafted the manuscript; KB, CS, FH, CG, AT and AS contributed substantially to the revision of the manuscript and all authors read and approved the final manuscript.
The authors declare that they have no competing interests.
Consent for publication
The manuscript does not contains any individual person’s data in any form (including individual details, images or videos), and therefore, we have not requested consent to publish from any person.
Ethics approval and consent to participate
The Research Ethics Review Board at Karolinska Institutet approved the present study and each participant signed a consent forms.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Bryngelsson D, Wirsenius S, Hedenus F, Sonesson U. How can the EU climate targets be met? A combined analysis of technological and demand-side changes in food and agriculture. Food Policy. 2016;59:152–64.View ArticleGoogle Scholar
- Carlsson-Kanyama A, Gonzalez AD. Potential contributions of food consumption patterns to climate change. Am J Clin Nutr. 2009;89(5):S1704–9.View ArticleGoogle Scholar
- Hedenus F, Wirsenius S, Johansson DJA. The importance of reduced meat and dairy consumption for meeting stringent climate change targets. Clim Chang. 2014;124(1–2):79–91.View ArticleGoogle Scholar
- Saxe H, Larsen TM, Mogensen L. The global warming potential of two healthy Nordic diets compared with the average Danish diet. Clim Chang. 2013;116(2):249–62.View ArticleGoogle Scholar
- Davis J, Sonesson U, Baumgartner DU, Nemecek T. Environmental impact of four meals with different protein sources: Case studies in Spain and Sweden. Food Res Int. 2010;43(7):1874–84.View ArticleGoogle Scholar
- Biesbroek S, Bueno-de-Mesquita HB, Peeters PH, Verschuren WM, van der Schouw YT, Kramer GF, et al. Reducing our environmental footprint and improving our health: greenhouse gas emission and land use of usual diet and mortality in EPIC-NL: a prospective cohort study. Environ Health. 2014;13(1):27.View ArticlePubMedPubMed CentralGoogle Scholar
- Scarborough P, Appleby PN, Mizdrak A, Briggs AD, Travis RC, Bradbury KE, Key TJ. Dietary greenhouse gas emissions of meat-eaters, fish-eaters, vegetarians and vegans in the UK. Clim Change. 2014;125(2):179–92.View ArticlePubMedPubMed CentralGoogle Scholar
- Soret S, Mejia A, Batech M, Jaceldo-Siegl K, Harwatt H, Sabate J. Climate change mitigation and health effects of varied dietary patterns in real-life settings throughout North America. Am J Clin Nutr. 2014;100 Suppl 1:490S–5S.View ArticlePubMedGoogle Scholar
- Vieux F, Soler L-G, Touazi D, Darmon N. High nutritional quality is not associated with low greenhouse gas emissions in self-selected diets of French adults. Am J Clin Nutr. 2013;97(3):569–83.View ArticlePubMedGoogle Scholar
- Temme EH, Toxopeus IB, Kramer GF, Brosens MC, Drijvers JM, Tyszler M, et al. Greenhouse gas emission of diets in the Netherlands and associations with food, energy and macronutrient intakes. Public Health Nutr. 2015;18(13):2433–45.View ArticlePubMedGoogle Scholar
- Temme EH, van der Voet H, Thissen JT, Verkaik-Kloosterman J, Van Donkersgoed G, Nonhebel S. Replacement of meat and dairy by plant-derived foods: estimated effects on land use, iron and SFA intakes in young Dutch adult females. Public Health Nutr. 2013;16(10):1900–7.View ArticlePubMedGoogle Scholar
- Edenhofer O, Pichs-Madruga R, Sokona Y, Kadner S, Minx JC, Brunner S, Agrawala G. S, et al. Technical Summary. In: Edenhofer O, Pichs-Madruga R, Sokona Y, Farahani E, Kadner S, Seyboth A. K, et al., editors. Climate Change 2014: Mitigation of Climate Change. Cambridge: Contribution of Working Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change; 2014.
- Intergovernmental panel on climate change. [http://www.ipcc.ch/]
- Commission E: Environmental Impact of Products (EIPRO). Analysis of the Life Cycle Environmental Impacts Related to the Final Consumption of the EU 25 2006. In: Contract No: European Commission Technical Report EUR 22284 EN. 2006.Google Scholar
- Nordic Council of Minister. Nordic Nutrition Recommendations: integrating nutrition and physical activity. 5th ed. Copenhagen: Nordic Council of Minister; 2012.Google Scholar
- LifeGene [www.lifegene.se]
- Almqvist C, Adami HO, Franks PW, Groop L, Ingelsson E, Kere J, et al. LifeGene-a large prospective population-based study of global relevance. Eur J Epidemiol. 2011;26(1):67–77.View ArticlePubMedGoogle Scholar
- Christensen SE, Moller E, Bonn SE, Ploner A, Wright A, Sjolander A, et al. Two new meal- and web-based interactive food frequency questionnaires: validation of energy and macronutrient intake. J Med Internet Res. 2013;15(6), e109.View ArticlePubMedPubMed CentralGoogle Scholar
- Christensen SE, Moller E, Bonn SE, Ploner A, Balter O, Lissner L, et al. Relative validity of micronutrient and fiber intake assessed with two new interactive meal- and Web-based food frequency questionnaires. J Med Internet Res. 2014;16(2), e59.View ArticlePubMedPubMed CentralGoogle Scholar
- Food composition table [https://www.livsmedelsverket.se/en/]
- Sjors C, Raposo SE, Sjolander A, Balter O, Hedenus F, Balter K. Diet-related greenhouse gas emissions assessed by a food frequency questionnaire and validated using 7-day weighed food records. Environ Health. 2016;15(1):15.View ArticlePubMedPubMed CentralGoogle Scholar
- Myhre GD, Shindell FM, Bréon W, Collins J, Fuglestvedt J, Huang D, et al. Anthropogenic and Natural Radiative Forcing. In: Stocker TFD, Qin G-K, Plattner M, Tignor SK, Allen J, Boschung A, et al., editors. Climate Change 2013: The Physical Science Basis Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge: Cambridge University Press; 2013.
- Quested TJH. Household Food and Drink Waste in the UK. In: Report prepared by Waste and Resources Action Programme (WRAP) ISBN: 1-84405-430-6. Banbury. 2009.Google Scholar
- Gustavsson JCC, Sonesson U, Emanuelsson A. The methodology of the FAO study. Göteborg: Global Food Losses and Food Waste - extent, causes and prevention; 2013.Google Scholar
- Willett W. Nutritional Epidemiology. 3rd ed. Oxford: Monographs in Epidemiology and Biostatistics; 2013.Google Scholar
- Wallen A, Brandt N, Wennersten R. Does the Swedish consumer’s choice of food influence greenhouse gas emissions? Environ Sci Pol. 2004;7(6):525–35.View ArticleGoogle Scholar
- Risku-Norja HKS, Helenius J. Impact of consumers’ diet choices on greenhouse gas emissions. In: Future of the Consumer Society 2009. Tampere: Finland Futures research Center; 2009. p. 159–70.Google Scholar