The world globalisation creates a more uniform nutritional pattern, englobing local cultures and traditional dietary patterns. Nutritional ethnography, the science behind the interactions between humans and environment, has an essential role to play in describing traditional dietary patterns(Reference Ottrey, Jong and Porter1). Nutritional ethnography implicates not only work in the field but also a (re)analysis and standardised reporting of historical nutritional data.
Inuit traditional nutrition was unique because the nutritional resources of the environment were limited and the almost non-existent agriculture due to weather conditions(Reference Bjerregaard and Jeppesen2). As stated by Kuhnlein et al.(Reference Kuhnlein3), who studied Canadian Inuit nutrition, the traditional dietary pattern played a central role in the cultural identity. The traditional dietary pattern changed slowly to more Western dietary patterns, associated with an increase in chronic diseases such as obesity, diabetes and CVD(Reference Kuhnlein, Receveur and Soueida4).
In Inuit traditional nutrition, two micronutrients were subject of intense discussions and controversy in the past, i.e. Ca and vitamin C. Ca is a mineral for optimal bone and teeth health and plays a critical role in nerve and muscle function(Reference Heaney, Dowell and Hale5,Reference Centeno, de Barboza and Marchionatti6) . The Inuit traditional dietary pattern has been described as low in Ca, mainly because marine food is a poor source of Ca(Reference Andersen, Noahsen and Rex7). When excluding famine, scurvy, a disease due to a low vitamin C intake, is probably the nutritional deficiency disease that has caused the highest mortality in human history(Reference Carpenter8). The arctic environment, with long winters and an almost total lack of fresh vegetable foods, was challenging to avoid scurvy. Questions arise about the sources of vitamin C in the Inuit traditional food system(Reference Carpenter8).
An important barrier in reanalysing traditional nutrition is that nutrition is often recorded anecdotally by early explorers and researchers. This was not the case for the Høygaard et al. expedition(Reference Høygaard9). Høygaard et al. left Copenhagen in August 1936 and stayed in East-Greenland until August 1937. The nutritional data were recorded by two researchers, but analysed on a household level only. Luckily, Høygaard et al. published the raw data as supplement, allowing a reanalysis on individual basis(Reference Høygaard9).
The aim of the present study was to reanalyse the Høygaard et al. nutritional data. In previous publications, we focused on adults, and some data for adults in Table 3 have been reported in the International Journal of Circumpolar Health(Reference Mullie, Deliens and Clarys10,Reference Mullie, Deliens and Clarys11) . In this publication, we focus on Ca and vitamin C in daily nutritional patterns of adults and children in a mid-1930 Inuit settlement in East-Greenland.
Methods
Participants
In 1937, the old ‘Angmagssalik’ district counted 756 inhabitants living in seventy-three houses (Fig. 1). Høygaard et al. selected ninety participants for the study, but for this analysis, we excluded fourteen participants less than two years old (nine boys and five girls). According to Høygaard et al., infants were suckled until they were two years old. In total, seventy-six participants clustered in thirteen families were selected: four individuals from Amituarsuk, twelve from Ikateq, ten from Kulusuk, twelve from Qernertivartivit, seventeen from Sermiligaq and fifteen from Titeqilaq; six participants lived near trading centres, three from Igdlumuit and three from Tasisaq (trading centre). Participating families were offered one krone per family.
Locations in East-Greenland of the Høygaard et al. study 1936–1937.
Assessment of nutrition
At baseline of the Høygaard et al. study, two researchers living with the families weighed all the imported and traditional foods in the houses. During the study, all given or received foods were registered, together with the blubber quantities used for the lamps, the dog food and the food supply stored. At endpoint, food quantities in the houses were weighed again. Høygaard et al. stated that they tried to limit interference of the researchers to induce the participants to other food than usually consumed. Food was weighed in a raw state. After cooking the remnants, mainly bones were weighed again, and the weight was subtracted from the total weight before cooking.
The total number of days that food was recorded varied between 6 and 225 per family. To estimate individual food consumption in this study from the Høygaard et al. data, the adult-equivalent conversion factors were used from Claro et al.(Reference Claro, Levy and Bandoni12) (online Supplementary Table 1S). A total adult-equivalent conversion factor for a whole family was estimated from the Claro et al. data, for example, for a family of one male adult of 43 years and one male adult of 18 years, this was 1·1 plus 1·2 = 2·3. The total food consumption of the family was divided by the total conversion factor, and the individual consumption was calculated by multiplying this by the individual conversion factor. Total traditional and imported food estimates from households were divided by adult-equivalents(Reference Høygaard9).
Estimation of the nutritional composition of traditional foods
To estimate the nutritional composition of traditional food, Høygaard et al. dissected fjord seals, cods, guillemot and ducks. Meat, organs and blubber were weighed and analysed for composition (online Supplementary Table 2S). The composition of imported food was calculated using McCance and Widdowson’s Composition of Foods(Reference Holland, Welch and Unwin13). Høygaard et al. determined vitamin C in traditional foods by the method suggested by Emmerie and van Eekelen(Reference Emmerie and Van Eekelen14). This method involves titration with 2–6 dichlorophenolindophenol after precipitation of cysteine, glutathione and ergothionine with mercuric acetate(Reference Emmerie and Van Eekelen14).
Body weight and height
For adults, body weight and height were measured by the Høygaard et al. team between September and October 1936. In absence of BMI data in the Høygaard et al. study, we calculated BMI for adults of 18 years or more, using the following formula: BMI = body weight (kg)/height (m)2. As for children data on weight and height were lacking, we used the standard weight and height for each age from the FAO Standards in Table 1 (15).
Baseline characteristics of the seventy-six participants
(Høygaard et al. Study 1936–1937)
IQR, interquartile range.
Assessment of rest metabolic rate, dietary-induced thermogenesis and physical activity
To estimate the rest metabolic rate in kcal per d, we used the Harris–Benedict equation adapted in 1984 for adults and the standard rest metabolism for each age from the FAO Standards for children(15,Reference Roza and Shizgal16) . Dietary-induced thermogenesis was estimated at 10 % of the rest metabolic rate in kcal per d. Physical activity was evaluated as a high physical activity level, following the data of Høygaard et al. For males, the researchers observed 8 h of sleep, 12 h of low-intensity physical activity and 4 h of moderate-to-vigorous intensity physical activity during hunting; for females this was 10 h of sleep, 8 h of low physical activity and 6 h of more intense physical activity during household activities and cooking. Physical activity level was evaluated as 2·10 for males and 1·82 for females(17).
Assessment of total energy expenditure
Total energy expenditure expressed in kcal per d was the sum of:
TEE = RM + RM/10 + EE
TEE = total energy expenditure in kcal per d,
RM = rest metabolic rate in kcal per d,
RM/10 = dietary-induced thermogenesis in kcal per d,
EE = energy expenditure by physical activity in kcal per d.
Energy balance (EB) was calculated by: EB = EI – TEE.
EI = energy intake by food and beverages.
Statistics
All descriptive data are presented as median and interquartile range (IQR) because data were not normally distributed. The data were analysed using IBM SPSS Statistics for Windows Version 27.0 (IBM Corp). All data were converted to 24 h, expressed as d–1 (day).
Results
Table 1 presents the general characteristics of the 37 (48·7 %) males and 39 (51·3 %) females of the Høygaard et al. study. Median (IQR) age in years was 35·0 (15·0) for male and 28·5 (9·0) for female adults. This was 9·5 (7·0) and 9·0 (7·0) for male and female children and adolescents, respectively. Median (IQR) BMI in kg/m2 was 24·9 (1·0) for male and 25·1 (2·9) for female adults.
Table 2 presents the energy expenditures for the seventy-six participants in the Høygaard et al. study. Median (IQR) rest metabolic rate in kcal per d was estimated at 1551 (104) for male and 1368 (25) for female adults. This was 1040 (265) and 1040 (243) for male and female children and adolescents, respectively. Median total energy expenditure in kcal per d was estimated at 2978 and 2627 for male and female adults and 1997 for children and adolescents.
Energy expenditure of the seventy-six participants (Høygaard et al. Study 1936–1937)
IQR, interquartile range.
* Rest metabolic rate estimated by FAO for ≤ 18 years and by Harris-Benedict 1984 for > 18 years (see Method section) (kcal per d).
† Dietary-induced thermogenesis estimated as 10 % of rest metabolic rate (kcal per d).
‡ Energy expenditure by physical activity estimated from the Høygaard data (kcal per d) with a physical activity level of 2·10 for males and 1·82 for females, corresponding to moderate to high physical activity level (see Methods section).
§ Total energy expenditure (kcal per d) (= sum of rest metabolic rate + dietary-induced thermogenesis + energy expenditure by physical activity).
Supplementary Table 2S estimates the nutritional composition of traditional foods. Energy content varied between 20·3 kcal for sorrel and 891·0 kcal for liveroil. Highest protein content was found in dried animal foods, and, as expected, carbohydrate content in animal foods was usually <1 mg for 100 g of food.
Table 3 presents the energy and the macronutrient consumption for the seventy-six participants in the Høygaard et al. study. Median (IQR) energy consumption in kcal per d was 3881 (1568) for male and 2910 (882) for female adults. This was 2442 (857) and 2023 (1122) for male and female children and adolescents, respectively. In energy-percent, 38 % were proteins, 39 % fats and 23 % carbohydrates. Less than 3 % of the energy intake came from imported foods.
Table 4 presents Ca and vitamin C intake. Median (IQR) Ca intake in mg.d–1 was 555 (1110) for male and 484 (883) for female adults. This was 458 (747) and 358 (838) for male and female children and adolescents, respectively. The Ca content of traditional foods varied between 3 mg for blubber to 1505 mg for dried capelin (online Supplementary Table 2S). In general, the Ca content of animal food is much lower than the cA content of marine plants. In total 19 % of the male and 43 % of the female adults had a Ca consumption below 400 mg.d–1. This was 38 % and 56 % for male and female children and adolescents.
Median (IQR) energy consumption and sources of consumption for the seventy-six participants*
(Høygaard et al. Study 1936–1937)
* All data were extracted from the Høygaard manuscript (see Method section).
Median (IQR) calcium and vitamin C consumption and sources of consumption for the 76 participants
(Høygaard et al. Study 1936–1937)
* All data were extracted from the Høygaard manuscript (see Method section). Høygaard et al. found a mean of 50 % loss when cooking.
† Muktuk or mattak was two or three days old when vitamin C was determined.
Median (IQR) vitamin C intake in mg.d–1 was 79 (77) for male and 59 (56) for female adults (Table 4). This was 44 (47) and 60 (52) for male and female children and adolescents, respectively. Approximately 40 to 50 % of the vitamin C came from animal foods, where Høygaard et al. found a 50 % loss when cooked. Narwhal skin and eyes had a marginal contribution to the vitamin C consumption, in contrast with algae. For males, 5 to 6 % of the participants had a vitamin C intake below 25 mg.d–1, this was 14 to 16 % of the females. Supplementary Table 2S presents the vitamin C content in mg per 100 g traditional Inuit foods as determined by Høygaard et al. The vitamin C content varied between 0 mg for blubber and 127 mg for adrenal glands. In general, the vitamin C content of meat is much lower than in organs: seals meat, for example, contains only 2 mg per 100 g compared with 10 to 20 mg for organs. High vitamin C concentrations can be found in some marine plants like Alaria pylaii with 44 mg for 100 g. Fjord seal eye and muktuk or mattak, traditionally seen as rich in vitamin C and antiscorbutic, contain 10 and 20 mg vitamin C for 100 g, respectively.
Figure 2 presents the individual vitamin C consumption in mg.d–1 v. total algae consumption in g.d–1. An increase in algae consumption was clearly associated with an increase in vitamin C consumption (r2 = 0·78). Six persons living near trading centres and consuming more imported foods had the lowest daily vitamin C consumption, increasing the risk of developing scurvy (marked in red in Fig. 2).
Vitamin C consumption (mg per d) v. total algae consumption (g per d) for the seventy-six participants (Høygaard Study 1936–1937).
Discussion
The main findings from the reanalysis of the Høygaard et al. data are that the food consumed in East-Greenland came mainly from traditional hunting and was low in carbohydrates. Second, Inuit traditional dietary pattern was low in Ca, and sources of Ca were algae and dried fish. Third, we emphasise the role of algae consumption in East-Greenland traditional dietary pattern as source of vitamin C. Food traditionally seen as sources of vitamin C like eyes from seals and narwhal skin played probably a minor role in avoiding scurvy.
The mean Ca intake in the Høygaard et al. study, i.e. 294 mg per 1000 kcal for males and 358 mg per 1000 kcal for females, was higher as described by Deutch et al. in 1953(Reference Deutch, Dyerberg and Pedersen18). They found a mean intake of 235 mg per 1000 kcal for Qaqortoq and 221 mg per 1000 kcal for Ilulissat. In the Høygaard et al. data, more than 95 % of foods were traditional, which was much lower in Deutch et al., i.e. 17 % for Qaqortoq and 25 % for Ilulissat. However, and for both studies, the intake of Ca was far below the recommendations of the Nordic Nutrition Recommendations 2012, i.e. 800 mg per d(19).
As noted by Høygaard et al., an often forgotten Ca source is drinking water. The geological systems south of Scoresby Sound are poor in Ca, so fresh water contains maximally 200 mg per l. However, at one liter water a day, this could be a non-negligible source of Ca with a biological availability comparable to milk(Reference Bacciottini, Tanini and Falchetti20). We have no data in Høygaard et al. to know if the low Ca intake was associated with health problems, except the statement that the team did not observe osteomalacia, a disease caused by inadequate levels of Ca and/or vitamin D. Sellers et al. gave ten healthy Inuit children (5 to 17 years of age) a standardised Ca load(Reference Sellers, Sharma and Rodd21). Dietary Ca absorption appeared to be more efficient in these Inuit children, with an increased frequency of hypercalciuria. The authors stated that a genetic adaptation to low Ca dietary patterns may be present in Inuit, with a possible risk of nephrolithiasis if standard nutritional guidelines would be followed.
A high vitamin C content of algae was confirmed by Emmerie et al. in 1933. They found 43 mg vitamin C in Fucus serratus and even 77 mg vitamin C in Fucus vesiculosis (Reference Emmerie and Van Eekelen14). According to Hoygaard et al., algae were much more consumed between September and December and plant foods during the rest of the year(Reference Høygaard9). Mattak or muktuk was seen by early researchers as an important source of vitamin C, but the consumption was very low and depending on the hunt. Second, seals eyes were also seen as source of vitamin C, but an eye, usually consumed raw, weighted only 33 g, which is 3 mg vitamin C, too low to feed a community(Reference Høygaard9). A person would need three eyes a day to reach 10 mg vitamin C and avoid scurvy, which represents 1·5 seal a day! Data on the prevalence of scurvy in East-Greenland in 1936–1937 are lacking. According to Høygaard et al., severe scurvy was unknown in the area under study, but sub-scurvy cannot be excluded in view of the chronic gingivitis observed by Høygaard et al. during spring. Høygaard et al. performed thirty-four blood determinations of vitamin C, 18 between November and April, and 16 between July and August. In total, 12 (35 %) observations were between 0·5 and 1·2 mg per dl, 16 (47 %) around 0·4 mg per dl and 6 (18 %) <0·3 mg per dl. The reported vitamin C blood determinations reflect hypovitaminosis C for 47 % of the sample and scurvy levels for 18 % of the population. It is remarkable that all six with extreme low blood concentrations of vitamin C lived near the trading centre consuming more imported food and less traditional food. Of the six persons living near trading centres, three adults lived in Igdlumuit and two adults with one child in Tasisaq. This distribution into two distinct families limits the possibility of a family clustering effect and/or of a genetic efficacy in vitamin C metabolisation(Reference Granger and Eck22).
Høygaard et al. noticed high seasonal variations in body weight, probably as consequence of variations in energy balances. Høygaard et al. recorded adult body weight of ten male and eight female Inuit. The highest mean weight for males was in September 1936 with 64·7 kg, the lowest in April 1937 with 61·6 kg, a difference of 3·1 kg or 21 700 kcal in approximately 212 d or 100 kcal per d. For females this was 63·3 kg in April 1937 and 55·0 kg in July 1937, a difference of 8·3 kg or 58 100 kcal in 122 d or 476 kcal per d, assuming no pregnancy. Høygaard et al. recorded a mean height for male adults of 1·61 m and 1·53 m for female adults. This means that the seasonal BMI (in kg/m2) varies between 23·8 and 25·0 for male and between 23·5 and 27·0 for female adults. According to the WHO, the same international cut-offs for BMI can be used for Asian populations, meaning that there was no undernutrition in this population during the time of the Høygaard et al. study(23). The different relationship in body weight between season and gender, i.e. a higher body weight for males and a lower body weight for females in the summer, might be the consequence of more hunting opportunities for males in the summer and consuming immediately meat and fat when carving up the game.
The most important limitation of the present study is that converting household consumption to individual consumption will obscure differences in intake among people of differing age and body weight. An unavoidable source of error for Høygaard et al. was eating when hunting. It was impossible to prevent the members of a family to hunt and/or to forbid eating during a successful hunting with colleagues. This quantity could not be estimated and will introduce an error mainly on male consumption. A last limitation of the present study is that we do not have information about how representative the sample of 10 % was for the total population in the district. Høygaard et al. did not describe how the families were selected, but several families had to be excluded because the nutritional data were unreliable. Again, we have no data about how many families were excluded, and a healthy volunteer effect and Hawthorne effect cannot be excluded.
As conclusion to the present study, a reanalysis of older nutritional data in a modern and standardised way can be interesting to better understand traditional nutritional systems in general, and of East-Greenlandic Inuit in particular. In this study, the importance of traditional foods in reaching an acceptable energy balance is emphasised, together with the confirmation of a low Ca intake in East-Greenland traditional dietary pattern, and the important role of algae consumption in Inuit traditional dietary pattern to avoid scurvy.
Acknowledgements
The authors are indebted to the participants of this study and to the remarkable work of the Høygaard team in 1936–1937, i.e. Arne Høygaard, Harald Waage Rasmussen, Edward Falsen Krohn and Unni Høygaard. We would also like to thank Katrien Alewaeters, librarian at the Vrije Universiteit Brussel, for finding the Høygaard manuscript.
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.
P. M. worked on the original idea for the study. P. M. analysed the data and drafted the first version of the manuscript and corrected by T. D. and P. C. All authors read and approved the final version of the manuscript. All authors had full access to all data (including statistical reports and tables) in the study and can take responsibility for the integrity of the data and the accuracy of the data analysis.
There are no conflicts of interest.
Supplementary material
For supplementary material/s referred to in this article, please visit https://doi.org/10.1017/S0007114521005055
