Iodine is an essential nutrient needed to produce thyroid hormones thyroxine (T4) and triiodothyronine (T3), and the intake of iodine is a key determinant of thyroid disease risk with both low and high intakes increasing the risk of disease(Reference Laurberg, Cerqueira and Ovesen1,Reference Taylor, Albrecht and Scholz2) . Thyroid hormones are essential for growth, neuronal development, reproduction and regulation of energy metabolism(Reference Brent3,Reference Cheng, Leonard and Davis4) . Mild and moderate iodine deficiency presents primarily as non-toxic goitre, an enlarged thyroid gland with normal production of thyroid hormones, and prolonged iodine deficiency may lead to hyperthyroidism and shorter lifespan(Reference Riis, Pedersen and Danielsen5). Furthermore, severe iodine deficiency may lead to hypothyroidism in adults and to cretinism in infants and children(Reference Laurberg, Cerqueira and Ovesen1,6) with iodine deficiency being the single most important, preventable cause of developmental brain damage worldwide(Reference Delange7,8) . Thus, the WHO recommends monitoring of all populations.
More than seventy countries have introduced iodisation programmes to combat iodine deficiency disorders(8). Denmark was mild to moderately iodine deficient and introduced iodine fortification of salt about year 2000(9) aiming to raise the iodine intake by about 50 µg/d(Reference Laurberg, Jørgensen and Perrild10). Follow-up has documented a marked reduction in the occurrence of goitre and thyroid dysfunctions(Reference Petersen, Knudsen and Carlé11,Reference Zimmermann and Boelaert12) , while iodine nutrition in pregnant women remains a continuous concern(Reference Andersen and Laurberg13–Reference Adalsteinsdottir, Tryggvadottir and Hrolfsdottir15).
The Faroe Islands are included in the Danish iodine fortification programme similar to Greenland(Reference Andersen, Hvingel and Kleinschmidt16) on a voluntary basis with non-iodised salt also available. However, no studies have been conducted on iodine status in the Faroese population. Nevertheless, several studies conducted in the Faroe Islands on contaminant exposure(Reference Weihe, Bjerregaard and Bonefeld-Jørgensen17) included dietary and urinary sampling, which may provide insight into iodine nutrition.
We thus aimed to settle the iodine nutrition status among adult Faroese aged 40 through 74 years by measuring iodine in spot urine samples to estimate the prevalence of iodine deficiency. In addition, we included an assessment of the importance of dietary peculiarities among the Faroese for iodine nutrition in order to guide dietary recommendations in the Faroe Islands. The iodine content of water varies considerably with geography but remains stable with time, and data for the Faroe Islands are included(Reference Andersen, Laurberg, Preedy, Burrow and Watson18).
Experimental methods
Area of investigation
The Faroe Islands, an archipelago located in the North Atlantic Ocean, are inhabited by close to 53 000 individuals(19). The islands are a part of the Nordic countries with similarities in terms of language, welfare system and lifestyle, including dietary intake. The food sources available in grocery stores in the Faroe Islands are like the sources available in grocery stores in the other Nordic countries. Dairy products in terms of milk and sour-milk products are produced locally and sold in all grocery stores, whereas the vast majority of cheeses are imported. Additionally, the Faroese cuisine includes local traditional foods like fermented and wind-dried lamb meat and fish, which has been cherished for centuries and still is(Reference Joensen20). Pilot whale (Globicephala melas) hunting is an old tradition too as is driving and communal sharing methods(21). In the Faroe Islands, there are two dietary recommendations for pilot whale meat and blubber intake. One published in 2008 by the Chief Medical Officer and the Department of Occupational Medicine and Public Health stating that pilot whale is considered unsuited for human consumption due to high concentrations of environmental pollutants(Reference Weihe and Joensen22) and the other published in 2011 by the Faroese Food and Veterinary Authority(23) stating that a maximum of one meal per month for the adult population, but that fertile women abstain from whale blubber until last child is born and whale meat from 3 months prior to conception and during lactation period. Fish intake has declined from three meals per week in the 1930s to one meal per week among the younger generations during the 2010s, and there are indications that dairy intake had decreased markedly, from 390 g/d in 1982 to 179 g/d in 2016(Reference Wennberg24).
Both iodised salt and non-fortified salt are available in all grocery stores, and although not mandatory, most of the commercial bread production in the Faroe Islands, but not all, uses iodised salt.
Participants and diet questionnaire
The study population was derived from a national cross-sectional population-based survey, conducted in 2011–2012 by Department of Occupational Medicine and Public Health to investigate a possible association between environmental contamination and risk of type 2 diabetes mellitus. A total of 2186 age- and sex-adjusted individuals aged 40–74 years were randomly selected from the Faroese Population Registry (comprising 10 % of the total population in each age group). The reason for including people from the age of 40 years in the original study was that the risk for type 2 diabetes increases with age and the prevalence is low before the age of 40 years. A detailed description has been published previously(Reference Veyhe, Andreassen and Halling25). At study entrance, 491 participants answered a questionnaire that pertained to personal information: schooling and education, health status and smoking habits. The questionnaire included two short sections on dietary habits. The first section posed the question: ‘How often do You consume’: potatoes; grains and cereals; cooked, fried or baked vegetables; salad and raw vegetables; fruit; fish as small meal or as bread-spread with six possible options from never/rarely to every day/several times per day. The second section posed the question ‘How often during the last year have you consumed’: fish; sea birds; pilot whale meat; pilot whale blubber with the frequency option: never; times per year; times per month; times per week. Anthropometric measures included body weight, body height, hip and waist circumference. BMI was calculated as weight in kg divided by height in metres squared.
Ethical approval
The study was approved by the local ethical review committee and the Data Protection Authority of the Faroe Islands, with participation on a voluntary basis as documented by written informed consent complying with the Declaration of Helsinki. Procedures done caused no harm to the participants. All data were pseudo-anonymised, and all analyses were conducted in accordance with the ethical approval.
Urine and tap water sampling
The participants (n 491) donated a morning spot urine sample at study entrance. All samples were stored in iodine-free containers at −80°C at the Department of Occupational Medicine and Public Health, Torshavn, the Faroe Islands until further analyses. Tap water samples were collected at twelve locations covering the geography of the Faroe Islands for measurement of iodine in drinking water to evaluate a possible contribution to the iodine intake.
Analytical methods
We used single random sampling for measuring urinary iodine concentration (UIC) to comply with the recommendations from the international organisations WHO and UNICEF to use single random sampling UIC to compare public health data among countries. UIC portrays dietary iodine intake because more than 90 % of dietary iodine is excreted in the urine and the vast majority within 24 h(Reference Keating and Albert26).
The iodine concentration in water and urine was determined by the Sandell–Kolthoff reaction modified after Wilson and van Zyl(Reference Wilson and van Zyl27). The principle is detection of the catalytic role of iodine in the reduction of ceric ammonium sulphate in the presence of arsenous acid after pre-treatment with alkaline ashing(Reference Andersen, Pedersen and Iversen28). Procedures were conducted at the iodine laboratory at Aalborg University Hospital (Aalborg Denmark)(Reference Laurberg29). Samples were analysed in random order, and repeated analyses of stored batches provided quality control.
Statistical analyses and justification of sample size
Statistical analyses were performed in Statistical Program for Social Science (version 25.0; SPSS Inc.). Continuous variables were analysed for normal distribution by visual inspection of QQ-plots and by Kolmogorov–Smirnov test. Four urine samples had UIC > 1000 mcg/l, but they did not influence the statistical significances in the results and all samples were included in the further analyses. Descriptive statistics are presented with mean and standard deviation or median and interquartile range for continuous variables and numbers and percentages for categorical variables. Intake of local traditional foods, whale meat, whale blubber and fish was quantified to times intake per year, and correlations were explored with iodine status using the Spearman’s rho for preliminary analyses (data not shown). Whole grain (bread and groats) was initially recorded with six categories from daily intake to never and after that converted to a binary variable: intake daily and more or less frequent than daily intake. Neither the iodine concentration nor the dietary variables were normally distributed, and groups were compared using the non-parametric Mann–Whitney U test (two groups) or Kruskal–Wallis test (several groups). The χ 2 test was applied for categorical variables. Univariate binary logistic regression was applied to further explore the link between the median iodine status (<101 µg/l v. ≥101 µg/l) and the following predictor variables: sex; age; place of living (city/town); BMI groups (<24·9, 25·0–29·9, ≥30·0); smoking (daily, occasionally, never) and educational level (up to 7 years, 8–11 years, upper secondary). All significant variables from the univariable logistic regression analyses were included in a multivariable binary logistic regression model using conditional backward elimination. Reported P-values were judged based on a statistical significance limit of 0·05.
When settling sample size, we aimed for a 5 % precision range with 95 % confidence in a population. The recommended number of urine samples adds up to a total of 489 in a population(Reference Andersen, Karmisholt and Pedersen30). This allows for three to four subgroups with a precision range of 10 % with 95 % confidence, and five to six subgroups with 10 % precision with 90 % confidence(Reference Andersen, Karmisholt and Pedersen30).
Results
Study population characteristics
The study group comprised 294 men and 197 women with a mean age of 59·5 (sd 9·3) and 60·4 (sd 8·8) years, respectively. Mean BMI for the whole group was 29·7 (sd 5·0), and a total of 86% were classified to be overweight (BMI > 25 kg/m2) or obese (BMI > 30 kg/m2) with significantly more men than women being overweight (P < 0·05) and obese (P < 0·05). More women smoked (P < 0·05), and 70 % were non-smokers.
Men reported a more frequent intake of the whale meat and whale blubber compared with women, 6 v. 3 times per year (P < 0·001) and 12 v. 4 times per year (P < 0·001), respectively, and 44 % of the women reported a daily intake of whole grain compared with 35 % of the men (P = 0·04). Fish intake was similar between the sexes with a frequency of 104 times per year for both men and women. There were no differences between the groups regarding food items: whale meat, whale blubber, fish and whole grain. There was a weak non-significant positive correlation between fish and whale meat intake, assessed as times per year, both overall (Spearman’s P = 0·07; P = 0·15) and stratified by sex: men, Spearman’s P = 0·08; P = 0·18; women, Spearman’s P = 0·04; P = 0·55. Fish and whale blubber intake were overall positively associated, Spearman’s P = 0·12; P = 0·01, but not when stratified by sex. Groups did not differ by educational level, place of living and smoking habits.
A total of eight participants were on thyroid supplementation medication, their UIC were in the range 50–99 µg/l (n 2), 100–199 µg/l (n 5) and 200–299 µg/l (n 1) and no participant was on anti-thyroid medication.
Iodine in urine and tap water
The median UIC was 101 µg/l (interquartile range 69–153 µg/l; range 21–1870 µg/l). A total of 37 % of urine samples were in the range of adequate status (UIC 100–199 µg/l). None of the samples from participants had UIC in the range of severe iodine deficiency (UIC < 20 µg/l), whereas 10 % were in the range of moderate deficiency (UIC 20–49 µg/l), 39 % mildly deficient (UIC 50–99 µg/l), 9 % were just above the recommended range (UIC 200–299 µg/l) and 6 % of the samples had UIC above 300 µg/l (Fig. 1). Significantly more women than men had below the median value UIC < 101 µg/l (61·9 % v. 41·5 %, P < 0·001). Conversely, a total of 8·2 % of the samples from men and 2·0 % of the samples from women had UIC > 300 µl.
Fig. 1. Proportions of urinary iodine concentrations among a population-based sample of 491 adult Faroese men and women divided into the following groups: moderate, mild deficiency, adequate, slightly increased and excess. P-values assemble sex differences assessed by one-sampleχ
2. *n 4 in the female group. 
, Male; 
, Female
Tap water iodine was low (<2·0 µg/l) on all locations (Fig. 2).
Fig. 2. Map showing the twelve locations for tap water collection in the Faroe Island. All samples had values < 2 µg/l. (1) Sørvágur, (2) Skálabotnur, (3) Við Streymin, (4) Fuglafjørður, (5) Klaksvík, (6) Runavík, (7) Gøta, (8) Kollafjørður, (9) Vestmanna, (10) Miðvágur, (11) Tvøroyri and (12) Tórshavn.
Determinants of iodine nutrition
The demographic and lifestyle characteristics presented in Table 1 do not explain any pattern regarding the UIC level except for sex, as women have significantly lower levels compared with men (P < 0·001). Although not significantly so, it seems as BMI is positively associated with UIC and the opposite direction is observed for educational attainment (see Table 1).
Table 1. Iodine urinary concentrations (µg/l) according to sex and age, demographic and lifestyle characteristics
(Numbers; median and interquartile range (IQR))
* P-values were assessed by using non-parametric test Mann–Whitney U test.
† BMI information missing 1 for one study participant.
‡ Education level missing for 14 study participants.
Intake of fish, whale meat and whale blubber was significantly associated with UIC, as the group with an intake in the fourth quartile of the respective food items had significantly higher UIC compared with low intake (see Table 2). We did not detect any association between the intake of whole grain and UIC.
Table 2 Iodine urinary concentrations (µg/l) for first and fourth quartile of three local traditional food items
(Numbers; median and interquartile range)
* Mann–Whitney U test between first and fourth quartile for each food item.
† Globicephala melas.
Men were more than twice as likely to have UIC above the median concentration compared with women in the binary logistic regression (OR = 2·30, P < 0·001); overweight (25·0 ≤ BMI < 30·0) and obese participants (BMI ≥ 30·0) were likely to have above median iodine status, although not significantly so (OR = 1·23, P = 0·047 and OR = 1·67, P = 0·07) compared with normal-weight participants. For educational attainment, participants with upper secondary school were close to half as likely to have sufficient iodine status compared with those with 7 years of schooling (OR = 0·53, P = 0·01). Regarding dietary intake, fish, whale meat and whale blubber were positively associated with iodine status (respectively: OR = 1·79, P = 0·06; OR = 1·43, P = 0·22; OR = 1·88, P = 0·02). There was no association between iodine status and age (OR = 0·99, P = 0·58); place of living (OR = 1·002, P = 0·9); smoking status (corresponding to daily: smoking occasionally, OR = 2·11, P = 0·17 and never smoking, OR = 1·17, P = 0·45) and whole grain intake (OR = 1·07, P = 0·80). Sex, BMI status, educational level and seafood intake (fish, whale meat and whale blubber) were included in the multivariable logistic regression model, and positive predictors retained were sex (male) (OR = 3·29, P = 0·03) and whale meat (fourth quartile) (OR = 2·94, P = 0·06).
Discussion
Main finding
Our study provides the first data to describe iodine nutrition status in the population in the Faroe Islands. Median UIC was 101 µg/l, which is by WHO considered to be within the recommended range. A total of 37 % of urine samples were in the iodine replete range of 100–200 µg/l, whereas 39 % were in the range of mild iodine deficiency (100 > UIC > 50 µg/l) and 10·2 % of samples were in the range of moderate iodine deficiency (50 > UIC > 20 µg/l). This fraction complied with the goal of no more than 20 % of the population samples lower than 50 µg/l and a median UIC above100 µg/l(6). Importantly, none of the participants’ samples had UIC in the range of severe iodine deficiency. Overall, women had lower UIC and compared with men with nearly half of samples suggesting mild to moderate iodine deficiency. Conversely, more than twice as many men had concentrations above the recommended range (UIC ≥ 200 µg/l), and for susceptible groups, this might pose a risk of iodine-induced thyroid dysfunction(Reference Laurberg, Cerqueira and Ovesen1,Reference Leung and Braverman31) .
Comparison with other studies
In Denmark, results from the DanThyr follow-up study conducted in 1997–1998 and again in 2008–2010 including 2465 participants (women aged 18–65 and men aged 60–65 years) reported an increase in median UIC from 61 to 83 µg/l. The increased iodine level was mainly ascribed to the iodine supplementation programme introduced in 1998. It included mandatory iodine fortification of table salt and fortified salt as a compulsory ingredient in commercial bread production from year 2000(Reference Rasmussen, Jørgensen and Perrild32). The approach was a cautious introduction of iodine supplementation(Reference Laurberg, Jørgensen and Perrild10), and the median UIC in the general population is still below the recommended level, corresponding to mild or borderline iodine deficiency(Reference Andersen, Sørensen and Krejbjerg33). In a convenience sample in Norway(Reference Brantsæter, Knutsen and Johansen34) of 276 participants aged 3+ years, the median UIC was 101 µg/l for the whole group, but for adults aged 18–64 years and elderly aged 65+ years, the median concentrations were 96 and 62 µg/l, respectively, suggesting inadequate iodine intake and a need for monitoring of iodine intake. In Sweden at the beginning of the twentieth century, the occurrence of goitre was about 18 %(Reference Greenwald35); therefore, the Swedish national iodine fortification programme for table salt started in 1936. Results from a representative national Swedish sample (n 889) of children aged 6–12 years included median UIC of 125 µg/l and thus stated a sufficient iodine status(Reference Nyström, Brantsæter and Erlund36). A newly published review on iodine status in twenty-three European countries was based on forty studies with a total of 63 000 individuals covering schoolchildren, pregnant women and adults. This review concluded that iodine deficiency is a continuous concern in Europe, especially among adults and pregnant women, and that monitoring is needed(Reference Ittermann, Albrecht and Arohonka37). Our study on iodine nutrition among Faroese aims to accommodate such recommendations to ensure good iodine health.
The sex-specific iodine intake found in our study with median UIC of 115 µg/l in males and 86 µg/l in females can partly be attributed to men’s higher whale meat intake (data not shown). Thus, markedly more men than women were iodine replete with the higher levels in the male group being comparable with findings in other European countries(Reference Ittermann, Albrecht and Arohonka37). Importantly, the lower level in women with a higher fraction being iodine deficient may be a concern with pregnancy raising the iodine need in women only(Reference Andersen and Laurberg13).
Food sources
The importance of fish as a source of iodine nutrient was mentioned as early as 1937 in a Faroese daily newspaper(38), but no further iodine investigations were undertaken. Foods of marine origin have higher iodine content because marine plants and animals concentrate iodine from seawater, as seaweed is one of the best iodine food sources. Other good sources include fish and other seafood, as well as eggs and bread(Reference Zimmermann39,40) . From early on, the major iodine source in the Faroe Islands is assumed to have been marine products and the population has been guessed to be iodine replete due to the high frequency of seafood consumption(Reference Steuerwald, Weihe and Jørgensen41). A dietary survey from 1982 including adult men and women found a daily fish intake of 78 g and a daily mean intake of iodine of 244 µg (–24 µg females; +32 µg males (n 331))(Reference Vestergaard and Zachariassen42). Almost 20 years later, in 2001, a dietary investigation among pregnant women found that the daily fish intake had come down as low as 40 g(Reference Veyhe43), corresponding to less than two meals per week. Recent data on fish intake recorded at the Department of Occupational Medicine and Public Health in the Faroe Islands indicate that the fish consumption may be even lower today(Reference Wennberg24), especially among the generation below the age of 30 years that includes the majority of pregnant women. Iodine concentration in Faroese cod is 15 µg/kg compared with 17 µg/kg in Denmark and 20 µg/kg in Iceland(44). The iodine content in fish seems low, and with the decreasing intake, we may assume that fish is not a main iodine source today. We did not detect any correlation between fish and whale meat intake but a weak positive overall association between fish and whale blubber intake. The logistic regression only retained whale meat as a positive predictor to the UIC. To our knowledge, there is no information available on the iodine content in pilot whale meat and blubber. But, a study conducted in Greenland reported the iodine content of whale meat to be 21 µg/kg and whale blubber 130 µg/kg(Reference Andersen, Hvingel and Laurberg45). The study does not state what type of whale but concludes that sea food in general has higher iodine contents compared with terrestrial animals.
Iodine content of tap water collected at twelve locations covering the geography of the Faroe Islands found a very low iodine content. Thus, iodine in drinking water does not contribute to iodine intake in the Faroe Islands in contrast to findings in example Denmark(Reference Andersen, Pedersen and Iversen28) but comparable with findings in Greenland(Reference Andersen, Hvingel and Kleinschmidt16,Reference Andersen, Hvingel and Laurberg45) . Other sources of iodine for the Faroese population include fortified salt available in grocery stores and used partly in commercial bread production. This may have compensated some of the reduction of intake of iodine-rich food items among the population of the Faroe Islands.
In regard to the dietary issues, the Faroese health authorities follow the Danish health authorities in terms of dietary recommendations. The Danish dietary recommendation, to the general public, of improving iodine intake is to increase the consumption of fish and whole grain as well as intake of lean dairy products. This could be done within the existing recommendation of healthy habits and by advising a dietary contributing 350 g of fish per week, of which preferably 200 g of fatty fish(46). Additionally, the Faroese health authorities recommend people to abstain from whale meat and blubber due to environmental contamination(Reference Weihe and Joensen22).
Strengths and limitations
This study has some limitations. The urinary samples are approximately 10 years old and were originally donated for a study concerning diet, contaminants and type 2 diabetes. The present study represents a secondary analysis of these data. The iodine status was similar across the original groups being at ‘low risk’, ‘having prediabetes’ and ‘diabetes’ and between the variables of age and BMI in the original group and in this present study is very similar with R 2 = 0·024 in the former and R 2 = 0·008 in the latter. The similarity across participant groups in the primary study supports the reliability of our findings.
Urine iodine concentrations vary markedly with both short-term variations related to dietary peculiarities(Reference Noahsen, Kleist and Larsen47) and long-term variation, and we do not have dietary information for the day of the urine sample. Diurnal variation in urinary iodine has been reported with lower concentration in morning urine and a 24 h sampling giving a better excretion estimate and thus thereby suggest a slight underestimation of our results(Reference Andersen, Waagepetersen and Laurberg48). However, with the sample size of 491 participants, we were able to determine the median UIC with 5 % precision, as the required minimum of 489 samples was accommodated in accordance with recommendations for number of urine samples. Moreover, the sample size allowed for confidence when evaluating subgroups(Reference Andersen, Karmisholt and Pedersen30). Furthermore, the study population reflects the general Faroese population in the respective age group 40–74 years, with the sex and age-adjusted sampling(Reference Veyhe, Andreassen and Halling25).
The questionnaire was made to investigate the association between environmental contamination and risk of diabetes mellitus type 2 and therefore limited to food that might contain contamination, thus not containing iodine-specific questions, but 67 % of the study population reported to eat whole grain daily to several times per day. The dietary information was based on reporting of food frequency without information on portion size and therefore carries some imprecision in reflecting the iodine intake, and we have no information about dairy intake nor on the intake of iodine containing supplements.
It is important to consider not only that the median iodine intake is sufficient but also to consider the total iodine exposition in the population. Pregnant and lactating women have greater need of iodine, and the recommended intake is 175–250 µg/d to target a median UIC of 150–249 µg/l during pregnancy and >100 µg/l during lactation(40,Reference Zimmermann49) . Our results indicate an increased risk of insufficient iodine intake among women of child-bearing age, pregnant and lactating women in the Faroe Islands, in keeping with indications that the younger generations consume less sea food and dairy products(Reference Adalsteinsdottir, Tryggvadottir and Hrolfsdottir15,Reference Nyström, Brantsæter and Erlund36) .
Concluding remarks
This study among 40–74 years old randomly selected participants suggests that the iodine intake in The Faroe Islands overall is adequate. Some groups may still be iodine deficient though none was severely iodine deficient with UIC below 20 µg/l. Samples from 48 % of the population surveyed were in the range of insufficiency. Importantly, we found the lowest iodine excretion and thus the highest risk of iodine deficiency in Faroese women, which calls for studies on vulnerable groups.
Future perspective
Iodine intake in the Faroe Islands might be a cause of concern, and further studies are needed to substantiate the results and to examine the iodine status in younger generations, including young women likely to have a lower intake of iodine-rich food items. Pregnant and lactating women are a key focus as the developing brain is particularly vulnerable to the negative effects of iodine deficiency. The systematic monitoring of iodine status in different groups is important to follow trends in iodine nutrition. This is an important task for the health authorities, and there is a need for public health strategies to monitor and secure adequate iodine status, especially in young women and during pregnancy.
Furthermore, a randomised controlled trial in a population with mild iodine deficiency is warranted to provide authorities with guidance on recommendations for iodine supplementation during pregnancy.
Acknowledgements
This study is in collaboration with The Thyroid Society Faroe Islands. Special thanks to Marita Hansen for helping and organisation.
This work was supported by the Research Council Faroe Islands (The Thyroid Society Faroe Islands and H. L. J. grant number 3118), The Research Council at the Faroese National hospital (H. L. J. grant number 18/00133-19) and Aalborg Hospital (S. A. grant number 3118) Dancea research programme, The Danish Environmental Protection Agency, grant number 112-00292.
The authors contributions were as follows: H. L. J. and P. W. were involved in conception of the study, H. L. J., S. A., P. W. and The Thyroid Society Faroe Islands were grant holders, H. L. J., G. S. K., A. S. V., P. W. and S. A. contributed to design and implementation of the study and preparation of the manuscript. S. A. did the iodine analyses. H. L. J., G. S. K. and A. S. V. did the data analysis.
There are no conflicts of interest.
Open access
