Dietary vitamin D intake and sun exposure are not associated with type 1 diabetic schoolchildren and adolescents: A first report in Algeria 1

Abstract

BACKGROUND:

A large number of children and adolescents worldwide suffer from physiological vitamin D (VD) deficiency, which has been associated with sun exposure and, consequently, the risk of developing various autoimmune diseases, including type 1 diabetes (T1D). However, the association of the disease with VD intake and sun exposure has yet to be explored.

MATERIALS AND METHODS:

We conducted a food frequency questionnaire and a 24-hour food recall survey, using “Ciqual table 2016” in 335 type 1 diabetic and age- and gender-matched healthy Algerian school children and adolescents from sunny Saharan and relatively less sunny Northern regions, aged between 5 and 19 years.

RESULTS:

Both dietary VD intake and VD levels were similar in T1D patients when comparing northern and southern regions (for both comparisons, p > 0.05). Neither sun exposure nor VD intake was associated with the disease (respectively, relative risk [RR] = 1.050, p = 0.680; RR = 1.082, p = 1.000. For Cochran and Mantel-Haenszel analysis; RR = 0.841, p = 0.862). VD intake showed a significant difference between diabetics and non-diabetics in the sunny region (p = 0.022). Additionally, significant differences were found between normal and T1D schoolboys (p = 0.038), and when comparing the two groups according to the dry areas (p = 0.016). Moreover, in contrast to circulating VD levels, which were lower in T1D patients than in healthy controls, those of VD intake were significantly higher (p < 0.05), especially in male patients and in those with balanced diet, low protein or carbohydrate consumption, specific food intolerances, and regular meals (p < 0.05), as well as in patients with a moderate or low consumption of cooked meals or steamed foods (p < 0.01). Conversely, VD intake was markedly lower in type 1 diabetics than in controls for dry and sunny areas, including the region of Adrar, as well as for consumption of low-fat foods and eggs (p < 0.05 for all comparisons). Nevertheless, the relative risk of sun exposure and dietary vitamin D intake according to the World Health Organization (WHO) standard did not show a significant association with T1D (common Mantel-Haenszel estimation, RR = 0.841, 95% CI 0.118–5.973, p > 0.05).

CONCLUSIONS:

T1D does not appear to be associated with VD intake and sun exposure in the Algerian Sahara region. Therefore, the consumption of VD in T1D patients in the Algerian Sahara would suspect that its association with the disease would be related to its synthesis alteration.

Keywords

Algeria adolescents children diet survey Dietary Vitamin D nutrition assessment schools Type 1 Diabetic

1 Introduction

Type 1 diabetes (T1D) is an autoimmune disease in which insulin-producing pancreatic beta-cells are progressively and selectively damaged [1] in a process mediated primarily by autoreactive T-cells [2]. However, an alternative view has recently been proposed according to which the beta-cell could be the origin of the disease, due to its nature and function making it susceptible to stress and, therefore, to an autoimmune reaction [3]. No matter what, it is clear that a combination of genetic susceptibility and environmental factors, including dietary factors, can lead to the activation of autoimmune cells that can potentially be harmful to pancreatic beta-cell antigens [4].

The impact of particular dietary components, such as micronutrients, on the immune system has been investigated through a range of investigations, including in vitro cell cultures and animal models, to assess their influence on various diseases, including autoimmune disorders. Of the micronutrients, VD has been the focus of significant study recently, given its capacity to influence the immune system as immunomodulator [5].

From the biochemical point of view, VD exists in two forms, vitamin D₂ (VD₂, ergocalciferol) and vitamin D₃ (VD₃, cholecalciferol) [6]. VD₂ comes from vegetables and fungi, and is produced by irradiating plants and mushrooms with ultraviolet B (UVB) rays [7, 8], whereas VD₃ has two main origins, i.e., dietary origin, mainly supplied by fatty fish, liver, butter, cheese and egg yolk, etc., and endogenous origin, in which it is produced from cholesterol, photolyzed by solar UVB rays [9]. VD₂ and VD₃ are both absorbed to reach the circulation, bound to the vitamin D-binding protein (VDBP), then undergo two main changes in the liver to produce calcidiol and then in the kidneys to generate the calcitriol [5 , 10–12].

Evidence seems to indicate that VD might be one of the major factors involved in the development of T1D due to the fact that there is a disparity in the incidence of the disease between the Northern and Southern areas, which could be associated with lower levels of exposure to sunlight and VD levels [13]. Of note, a VD-poor diet and insufficient sun exposure leads to VD deficiency at all stages of life [14], and this has been linked to the onset of T1D during childhood and adolescence [13].

Despite its year-round sunny climate, Algeria experiences a high prevalence of VD deficiency among children and teenagers [15, 16]. Additionally, the number of children under the age of fifteen developing T1D is increasing, with almost 3,000 newly diagnosed patients with T1D annually [17]. Moreover, when compared to other Mediterranean countries such as France, Greece and Italy, Algeria’s prevalence of T1D has increased steadily to reach 31.12±3.60 (IC 95%, 27.90–34.23), between 2013 and 2017 [18].

In view of the aforementioned elements, we examined for the first time whether both dietary VD intake and sunlight exposure might be associated with T1D. To this end, we assessed VD intake levels in patients with recent-onset T1D and age- and gender-matched healthy school children and teenagers from two Algerian regions, taking into account the sunshine level, i.e., the Northern region which has relatively low sunshine, and the Saharan region which has relatively high sunshine.

2 Population and methods

2.1 Study design and participants

The present study was a 24-hour recall food survey. The estimation of dietary VD intake was done using the Ciqual Table 2016 [19]. Three hundred and thirty-five children and adolescents, including diabetics and non-diabetics aged 5 to 19 years and age- and gender-matched, were recruited from primary, middle and secondary schools, registered in Screening and Follow-Up Health School Units (SFHSU) (Fig. 1).

Fig. 1

Summary and flowchart of the current study.

Schools in urban and rural areas were randomly selected in each district, in accordance with the mix of social classes. In Algeria, schooling is mandatory for the corresponding age groups. These schools are balanced in terms of gender and age groups, and the selection of age groups was made for practical and physiological reasons. Additionally, the WHO and EURODIAB diagnostic criteria were utilized by clinicians from the SFHSU to assist in the selection of pupils with diabetes [20].

2.2 Study geolocation and nutritional assessment

The importance of VD is acknowledged for its myriad of vital functions, particularly in immunomodulation [5]. The two main ways that guarantee the required levels of this nutrient are dietary intake and exposure to sunlight. Given that VD deficiency is frequently observed in Algeria, we have attempted to determine whether VD intake and sunshine are both associated with T1D. Therefore, this study was conducted in two areas of Algeria based on the level of sunshine: the Sahara (Naama, Bechar and Adrar) with relatively high sunshine, and the North (Oran, Ain Temouchent, and Sidi Bel-Abbès) with relatively low sunshine (Fig. 2). Pupils with diabetes over the age of 19, as well as those who provided incomplete or incorrect data, particularly concerning food intake, were excluded.

Fig. 2

Geographical location of the study areas. With an area of 2,381,741 km² (919,595 sq miles), Algeria is the largest country in Africa and is estimated to have a population of over 44 million people. Approximately 70% of the population resides in the coastal region, while the remaining 30% is concentrated in the Sahara region, mainly in oases. It is noteworthy that nearly 30% of Algerians are under the age of fifteen [67].

The questionnaire consisted of two sections, the first included personal information, socio-demographic, clinical and anthropometric parameters (age, gender, education level, height, weight, family history of diabetes, physical activity, eating and smoking habits). Skin phototypes were assessed according to Fitzpatrick’s classification. The student clinical files are used to determine weight and height. The second part of the questionnaire comprises two sections: (i) a 24-hour recall that contains the dietary intake with details on all foods, water and beverages taken the day before, and (ii) a meal frequency questionnaire in which the informants indicate the snack and certain type of food consumed, the cooking method used, whether they use VD supplements, and VD-rich foods. Concerning the amount of the foods, we used the usual culinary instruments such as spoons, bowls, cups, plates, slice (of bread/cake), etc., and a food scale, moreover, each pupil was asked to indicate whether the portion was large, medium, or small.

2.3 Body mass index

After obtaining the data using the Centers for Disease Control and Prevention (CDC) calculator [21], the sample was divided into four groups of body mass index (BMI) percentiles: underweight (<5^th percentile), normal weight (5^th –85^th percentile), overweight (85^th –95^th percentile), and obese (≥95^th percentile) (Fig. 3).

Fig. 3

Percentile growth charts [21].

2.4 Basal VD

Basal circulating VD levels were recorded in all participants to check for any abnormality in serum vitamin D concentration prior to nutritional survey.

2.5 Ethical statement

The study was approved by the Institutional Review Board of Biology Department, Faculty of Nature and Life Sciences, Djillali Liabes University of Sidi-Bel-Abbès, Algeria and authorized by the Direction of Educational Authorities of each district, in collaboration with SFHSU Physicians. Students and even their parents were consented to fill out the questionnaire.

In our country, studies similar to ours, specifically the nutrition survey, are registered and approved by the Institutional Review Board. The registration code for this study is 168/DS/FSNV/2011. The initial approval date by the Ethical Oversight Authorities (the Direction of Educational Authorities of each district, in collaboration with SFHSU Physicians) was in October 2016.

2.6 Statistical analyses

The collected data were represented as mean±standard deviation, and the statistical analyses was performed using IBM SPSS v 26 software. The median was calculated at 95% confidence interval, and the interquartile range was determined by the 25^th and 75^th percentiles.

(i) The required sample size was determined using the Scalex SP calculator, based on type of the study (cross sectional survey) [22], the statistic for a level of confidence (Z, 1.96 for 95% confidence level), (ii) the prevalence of the population of children with T1D (p, 0.207%) [18], and (iii) the precision/acceptable margin of error corresponding to the size of the workforce (d, 5% as a starting precision), as reported [23 –25]. In order to increase the sample size and to improve the precision of the results, the margin of error was lowered from 5% to 1.15%. To reach the desired precision level, the number of participants required was estimated at 60, which corresponds to a total of 360 participants, considering the highest prevalence of 0.207% of eligible type 1 diabetic children across the six departments of the North and South region. Ultimately, a number of 335 participants was included after excluding those who did not meet the eligibility criteria or had declined to participate to the study (25 participants; 16 from Oran and 9 from Naama). The comparison of VD intake between groups was done by Mann-Whitney U test. The association between T1D and dietary VD intake was assessed using χ² or Fisher exact test, and the relative risk was estimated with 95% confidence interval. Cochran and common Mantel-Haenszel analysis was performed to assess combined the relative risk. Pearson or Spearman correlation was used, appropriately, to evaluate the relationship between dietary VD intake and latitude, sun exposure duration and body weight. A two-sample z-test was used for comparison of proportions. The differences were considered significant when p < 0.05.

3 Results

As shown in Fig. 4, dietary VD intake percentiles were not similar between the two groups. In fact, the difference was observed upper the 35^th percentile and was clearly evident in subjects having more than 90^th percentile. This result indicates that less than 5% of all subjects have a sufficient dietary VD intake.

Fig. 4

Dietary vitamin D intake percentiles in both groups.

The comparison of VD intake between type 1 diabetic patients and non-diabetic subjects is presented in Table 1.

Table 1

Dietary vitamin D intake in diabetic and non-diabetic pupils

Groups	Min.	Max.	Mean±SD	IQR		Median	95% CI		p-value
				25th	75th		LL	UL
ND	0.03	20.25	1.48±3.32	0.30	1.09	0.75	0.60	0.76	0.073
T1D	0.03	20.35	1.43±2.94	0.30	1.44	0.80	0.76	0.98

The amount of vitamin D intake is expressed as micrograms per 100 grams of food daily. The comparison of differences between the two independent groups was done using the Mann-Whitney U test. 95% CI: confidence interval of the median, IQR: Interquartile range; LL: lower limit, ND: non-diabetics, SD: standard deviation, T1D: type 1 diabetics, UL: upper limit.

In contrast to the basal circulating VD levels, according to the WHO standard, which were decreased in T1D patients compared to healthy controls, regardless of geographic locality with respect to sun exposure (p < 0.05) (data not shown), our results did not show a significant difference between VD intake distribution in diabetic and non-diabetic (ND) pupils (Table 1).

The levels of sun exposure relative to time spent outdoors in type 1 diabetic patients and non-diabetic subjects, and in the two groups combined from the North and South geographic localities are presented in Tables 2 3, respectively.

Table 2

Levels of sun exposure relative to time spent outdoors in type 1 diabetic patients and non-diabetic subjects

Sunshine duration	Non-Diabetics		Type 1 diabetics		p-value
	Proportion	%	Proportion	%
5 min	8	61.5%	5	38.5%	0.721
10 min	22	56.4%	17	43.6%	0.565
l5 min	49	57.0%	37	43.0%	0.484
30 min	71	56.3%	55	43.7%	0.457
>30 min	30	42.3%	41	57.7%	0.290

A two-sample z-test was used for comparison of proportions.

Table 3

Levels of sun exposure relative to time spent outdoors according to the North and South geographic localities

Sunshine duration	Southern region		Northern region		p-value
	Proportion	%	Proportion	%
5 min	3	23.1%	10	76,9%	0.040*
10 min	14	35.9%	25	64,1%	0.044*
l5 min	47	54.7%	39	45,3%	0.438
30 min	77	61.1%	49	38,9%	0.004*
>30 min	30	42.3%	41	57,7%	0.095

*p < 0.05, **p < 0.01. A two-sample z-test was used for comparison of proportions.

As depicted in Table 2, there was no significant difference in sun exposure duration between type 1 diabetics and non-diabetics in the regular range from 5 minutes to 30 minutes (p > 0.05). However, the number of type 1 diabetics and non-diabetics exposed to the sun for 5 minutes or 10 minutes was significantly increased in the northern region than in the southern region (for both comparisons, p < 0.05) (Table 3). In contrast, the number of subjects exposed to the sun for 30 minutes was significantly higher in the Southern region than in the Northern region (77 versus 49, p < 0.01). Finally, no significant difference was observed between the two regions for sun exposure duration of 15 minutes or more than 30 minutes (for the two comparisons, p > 0.05).

As shown in Table 4, skin, age, educational level, and sports practicing, did not make any significant difference between T1D patients and healthy controls (for all comparisons, p > 0.05). Additionally, we observed that dietary VD intake median was significantly higher in T1D patients than in ND subjects, according to the region with relatively high sun exposure (p = 0.018).

Table 4

Dietary vitamin D intake between diabetic and non-diabetics pupils, according to general characteristics, geographic location, eating habits, cooking methods, and vitamin D-rich foods

Variable		Diabetics (n = 155)		Non-diabetics (n = 180)		p-value
		n	X±SD	n	X±SD
Socio-demographic	Gender
	Male	68	1.92±2.43	87	1.09±2.43	0.038*
	Female	87	1.06±1.83	93	1.85±3.95	0.456
	Skin color
	Black	7	1.52±1.13	5	0.45±0.21	0.381
	White	55	1.41±3.20	73	1.39±3.15	0.061
	Dark-brown	92	1.45±2.93	102	1.58±3.50	0.519
	Age
	5–10	28	0.91±0.60	5	0.54±0.45	0.268
	10–15	61	1.29±2.87	92	1.50±3.34	0.388
	>15	66	1.83±3.58	83	1.51±3.39	0.185
	BMI
	Underweight	9	0,66±0,59	11	2,24±4,33	0,824
	Normal weight	73	1,56±3,28	116	1,51±3,38	0,118
	Overweight	13	1,87±3,74	5	0,49±0,32	0,289
	Obese	18	1,00±0,79	7	0,49±0,28	0,074
	Study Level
	Primary	36	0.90±0.75	33	1.14±2.82	0.134
	Middle	66	1.26±2.88	71	1.47±3.31	0.412
	Secondary	53	2.09±3.91	76	1.63±3.55	0.157
	Sport
	No	33	0.23±0.61	16	0.59±0.45	0.344
	Yes	122	1.59±3.26	164	1.87±3.47	0.091
Geographical location	Regions
	Region with relative	74	1.61±3.89	90	1.56±1.08	0.906
	low sun exposure “North”
	Ain Témouchent	30	1.55±4.23	30	1.86±4.36	0.560
	Sidi Bel Abbes	30	2.11±4.34	30	1.64±3.03	0.823
	Oran	14	0.60±0.60	30	1.12±2.74	0.878
	Region with relative	81	1.28±1.80	90	1.39±1.14	0.022*
	high sun exposure “South”
	Naama	21	1.77±3.30	30	1.85±2.98	0.946
	Bechar	30	1.25±0.65	30	0.63±0.51	0.018*
	Adrar	30	1.19±0.93	30	1.76±4.94	0.027*
	Climate
	Dry	83	1.35±1.87	90	1.39±3.13	0.016*
	Humide	72	1.53±3.88	90	1.56±3.48	0.777
Diet habits	balanced diet
	No	56	1.01±1.32	30	1.25±2.80	0.647
	Yes	99	1.65±3.47	150	1.53±3.43	0.036*
	Rich in carbohydrates
	No	118	1.58±3.27	160	1.49±3.33	0.016*
	Yes	37	0.99±1.50	20	1.42±3.38	0.787
	Rich in protein
	No	116	1.57±3.29	166	1.56±3.43	0.034*
	Yes	39	0.84±1.48	14	0.52±0.40	0.456
	Rich in lipids
	No	132	1.50±3.08	167	1.45±3.26	0.090
	Yes	23	0.84±0.70	13	1.89±4.11	0.519
	Intolerance to a particular food
	No	135	1.32±2.51	153	1.07±2.32	0.016*
	Yes	20	2.41±5.43	27	3.96±6.24	0.431
	The habit of skipping a meal
	No	56	1.81±3.43	68	1.12±2.25	0.010*
	Yes	98	1.18±2.56	112	1.68±3.76	0.857
Cooking method	Grilled
	No	148	1.32±2.76	165	1.31±2.83	0.100
	Yes	7	3.39±5.24	15	3.06±6.14	0.207
	Fried
	No	87	1.26±2.51	69	1.63±3.37	0.679
	Yes	68	1.65±3.41	111	1.41±3.31	0.086
	Boiled
	No	115	1.40±2.69	114	1.06±2.38	0.009**
	Yes	40	1.52±3.67	66	2.13±4.33	0.759
	Steamed
	No	129	1.58±3.26	108	1.31±3.13	0.005**
	Yes	26	1.18±2.13	72	1.71±3.57	0.356
	In Sauce
	No	21	2.95±5.70	19	0.78±0.62	0.257
	Yes	134	1.18±2.13	161	1.59±3.55	0.134
Foods rich in VD	Fish
	No	88	0.94±0.74	89	0.95±1.91	0.074
	Yes	66	2.07±4.32	91	1.95±4.14	0.373
	Fats
	No	111	1.16±1.66	131	1.40±3.14	0.049*
	Yes	43	2.15±4.90	49	1.66±3.72	0.848
	Eggs
	No	61	1.55±3.62	46	1.26±3.39	0.633
	Yes	93	1.36±2.46	134	1.56±3.31	0.042*
	Cheese
	No	40	1.62±2.73	31	0.92±1.35	0.178
	Yes	114	1.38±3.01	149	1.59±3.56	0.204
	Cereals
	No	54	1.46±3.25	68	1.15±2.56	0.132
	Yes	100	1.42±2.78	112	1.71±3.74	0.320

Data are presented as mean (X)±standard deviation (SD). The amount of vitamin D intake is expressed as micrograms per 100 grams of food daily. The comparison of differences between the two independent groups was done using the Mann-Whitney U test. *p < 0.05, **p < 0.01. BMI: body mass index classes.

As depicted in Table 5, there was a moderate positive and highly significant correlation between VD intake and body weight gain in healthy population (n = 61, Spearman’s Rho = 0.372, p = 0.003). Conversely, no significant correlation was observed between dietary VD intake, weight loss, latitude, and sun exposure duration for diabetic and non-diabetic pupils (Fig. 5). A significant difference in dietary VD intake was shown in southwestern Algeria (p = 0.022), particularly in Adrar (p = 0.027), and Bechar (p = 0.018); whereas this difference was not observed in the northern region. Additionally, except for humidity, there was a significant difference in VD intake between normal and T1D schoolboys (p = 0.038); while schoolgirls showed a similar VD intake (p = 0.456). Furthermore, a significant difference in the dry areas was observed (p = 0.016), more specifically in Adrar and Bechar regions.

Table 5

Relationship between latitude, sun exposure duration or body weight and vitamin D intake

Variable	Non-diabetics			Type 1 diabetics
	n	r	p	n	r	p
Latitude	180	–0.062	0.409	155	–0.035	0.667
Sun exposure duration (min)	180	0.000	0.995	155	–0.092	0.256
Body weight loss (kg/cm²)	54	–0.072	0.605	49	–0.031	0.833
Body weight gain (kg/cm²)	61	0.372**	0.003	65	–0.017	0.893

The amount of vitamin D intake is expressed as micrograms per 100 grams of food daily. Bivariate correlation analysis was carried out using Pearson’s or Spearman’s correlation coefficients, appropriately, according to the normality of the distribution. **p < 0.01.

Fig. 5

Levels of vitamin D intake in type 1 diabetes and healthy controls according to sun exposure.

As for diet habits, schoolchildren with a balanced diet differed significantly in terms of dietary VD intake. Learners who did not consume a diet rich in carbohydrate or protein marked a significant difference in terms of dietary VD intake (p = 0.030). Additionally, for pupils who did not report an intolerance to a specific food such as fish, fat, eggs, an important difference was noted (p = 0.016). Regarding the habit of skipping meals, a significant difference was clearly indicated for those who did not skip their meals (p = 0.010). Taking into account the cooking method, a high significant difference was observed in subjects who did not employ the boiled method (p = 0.009). For the subjects who did not use the streamed method, a high significant difference was observed (p = 0.005). As for the other cooked methods (grilled, fried, and in sauce), no significant difference was observed (p > 0.05). The results indicate that consuming foods rich in VD such as fats displays a difference in dietary VD intake (p = 0.049), contrary to fish, cheese and cereals, no significant difference was observed. For the individuals who consumed the eggs, a significant difference was noted (p = 0.042).

Finally, no association was observed for common Mantel-Haenszel estimation when considering both sun exposure and dietary VD intake according to the WHO standard (relative sun exposure; RR, 1.050, 95% CI 0.833–1.323, dietary VD intake; RR, 1.082, 95% CI 0.403–2.902, MH; RR, 0.841, 95% CI 0.833–1.323, 0.118–5.973, p = 0.862) (Table 6).

Table 6

Relative risk of sun exposure and dietary vitamin D intake according to the World Health Organization standard

Factors	RR	95% CI		p-value
		LL	UL
Relative sun exposure (min)	1.050	0.833	1.323	0.680
Dietary VD intake (micrograms/100 g of food/day	1.082	0.403	2.902	1.000
MH	0.841	0.118	5.973	0.862

The relative risk was calculated using Cochran and common Mantel-Haenszel analysis. 95% CI: 95 % confidence interval, LL: lower limit, MH: Mantel-Haenszel common relative risk estimate, RR: relative risk, UL: upper limit.

4 Discussion

VD can be obtained from sunlight, and it varies related to endogenous factors such as skin pigmentation, genetics, physiology (bioavailability, bioequivalence and safety), adiposity, and bile acid secretion, as well as exogenous factors, such as socio-demographic (age, income, race/ethnicity, gender), physical inactivity, nutrition (food VD intake (food sources (foods rich in VD, competing foods (cholesterol, tocopherol, phytosterols [26]), dietary habits (cooking method, including skipping a meal, dairy and animal product exclusion (lactose intolerance, cow’s milk food allergy [27], ovo-vegetarianism, veganism), VD supplementation and fortification, sun exposure (latitude, season, outdoor physical activity, use of sunscreens, cultural differences in dress, tanning bed), drugs (anticonvulsants (phenobarbital, dilantin, tegretol), bisphosphonates [29], ketoconazole, thiazides [30] and toxic habits including smoking. The classical role of VD to promote the human growth is very important during childhood and adolescence, when an individual spends considerable time in education and playing sports. In Algeria, studies on VD status in children and adolescents are scarce. To our knowledge, this is the first research conducted in Algeria that reports on the dietary VD intake in schoolchildren and adolescents, in both urban and rural areas of many districts. Through this study, we attempted to investigate the difference in VD intake in food consumed by healthy and T1D patient pupils.

4.1 VD deficiency

In 2005, a national survey, entitled Transition and Health Impact in North Africa (TAHINA), was conducted by the National Institute of Public Health in Algiers (INSP) in collaboration with the European Union in 16 districts. The findings of this survey showed that the daily food consumption of the studied population did not comply with international health recommendations [31], and, the present study found that the reported food intake of VD was low. This situation may be explained by the low socioeconomic status and a diet low in VD [32, 33]. People may rely on diet to satisfy their VD requirements, but according to some reports, a high rate of VD insufficiency in Algerian children and teenagers has been noticed [16, 34], which is consistent with the results obtained in other regions of the world such as, Spain, Greece, and Africa [35, 36]. In fact, the consumption of food rich in VD is little or almost non-existent and even fortified foods are negligible, for instance, the milk is generally not fortified with VD, unlike in Western countries [37], such as Finland [38], Germany, Ireland, the Netherlands, the United Kingdom, and the United States, where higher intakes are remarked [39, 40]. In our sample, cheese did not make a significant difference between the normal and diabetic pupils. Although Algeria is one of the main countries consuming milk and dairy products, this consumption is primarily a source of calcium and protein [41].

4.2 Regions

According to O’Mahony, et al., 2011, dietary VD intake differ across regions. In contrast to Mediterranean countries, intakes are higher in Scandinavian region [42].

Of note, the lack of VD in foods may be related to the availability and bioavailability of food containing VD [43]. Fish are the major source of dietary vitamin D intake in Japan, Norway, United Kingdom, and Spain [43], while in the Algerian Sahara, fresh fish is difficult to obtain due to the long distance between the fishing port and this region, as well as the distribution of small and distant settlements, the high cost in winter, and the rapid deterioration in quality, especially especially in the heat of summer, making it rare and expensive.

Fat are also the most common food sources in Norway, that’s why VD status is adequate in a large part of the population [44, 45]. Comparing with our sample, there was no difference in VD intake in pupils eating diet rich in lipids, so we hypothesise that in these diets the VD intake was very low.

4.3 Eating habits

Dietary intake habits may vary between countries, necessitating region-specific and population-specific research on VD [46]. In the United Kingdom, cereal and cereal products contribute significantly to the average daily VD intakes in adults [47], whereas our study showed that patients and healthy learners who did not follow a high-carbohydrate or high-protein diet had different in VD intake levels. In Japan, eggs are the second major sources of dietary VD intake after fish [48], whereas in patients and healthy pupils who did not consume a high-fat diet there were marked differences in VD intake. On top of that, there were certain foods that could not be tolerated by the diabetics and controls, such as milk, eggs, fish, etc., but, in fact, even these pupils had no intolerance or allergy to any particular food, a significant difference was noticed. For diabetic and healthy schoolchildren and adolescents who had a balanced diet, VD intake differed significantly.

Skipping meal as breakfast or lunch among adolescents and children worldwide is becoming a public health concern, resulting in inadequate intake of essential nutrients such like VD [49 –51]. For our sample, the significant difference in VD intake for those who did not skip any meals may indicate that their diets were not sufficient in VD intake, since breakfast and lunch are important sources to supply this nutrient.

For the cooking methods, previous studies revealed that cooking may affect the retention of VD in the diet, as VD is a cooking-sensitive micronutrient, and its loss is generally higher for longer cooking processes or higher heating temperature [52]. VD intake was significantly different between the two groups except for the boiling and streaming methods, which are not commonly used in Algerian cooking.

4.4 Gender

From our findings, it appears that dietary intake of VD is low in both genders, with a significant difference between normal and T1D schoolboys. Previous studies found that males had higher intakes of VD [53, 54]. In Ireland, Black et al., 2014 [55] found that teenage boys had the highest intakes (2.7μg/d), which were significantly higher than in teenage girls (1.9μg/d) with no significant difference between boys and girls in users of VD-containing supplements. According to the Canadian Community Health Survey (CCHS), the average daily intake of VD was 7.3 and 5.4μg for boys and girls aged 9-18-years, respectively Vatanparast, et al., 2010, Murphy et al., compared milk intake, which is the main source of VD in the USA, and revealed gender differences [56].

4.5 Skin

For all researchers, the major and the essential source of VD is the skin synthesis [57 –59], which represents 90 % of individual’s VD needs [60, 61]. Within 20 minutes, it can be obtained at least 10 000 IU, while nutrients can only provide 40–400 IU [62], i.e., less than 10% [63]. From this, dietary sources represent small amount. Nevertheless, dietary intake may be still be important, particularly in areas with limited sun exposure.

In Norway, low-pigmentation skin facilitates adequate VD intake from sunlight, whereas in the Algerian population, this is not the case. Data from countries with similar climates and the same human composition indicated that VD intake was low [45]. However, the results of the present study showed that there was no significant difference in VD intake between the control group and the patients.

4.6 BMI

VD intake classed by BMI percentile did not differ between normal and T1D pupils. Moore et al., (2014) reported that 15.2% were overweight (≥85^th to < 95^th BMI percentile) and 19.1% were obese (BMI≥95^th percentile), but did not provide details on VD intake for each group. Increased adiposity is known to cause lower serum 25OHD levels [64]. Nonetheless, our study showed that VD intake was higher in obese persons compared to their leaner counterparts, and we can generally classify them as follow: underweight pupils (9.3%), normal weight (75.5%), overweight (7.2%), and obese (8.1%). Therefore, we should promote VD intake through dietary sources for all of these categories, since they are all below the recommendation.

4.7 Diet survey

It is common that foods eaten less than once or twice a week may not be captured in a single day dietary record, due mainly to subject recall errors and underreporting [65]. However, a large group is sufficient to report mean usual intake using a single 24-hour recall [57]. Therefore, this survey was supplemented with a food frequency questionnaire [66].

4.8 Limitations

One of the limitations is the uncertainties surrounding the Dietary Reference Intake (DRI) values for VD which leads to an inability to characterize and integrate sun exposure with dietary intake recommendations as much as may be appropriate [64]. However, this was a food-based VD intake assessment and cannot be compared to DRI, since dietary supplementation was not recorded to estimate total VD intake.

5 Conclusions and future prospects

In the current study, although the basal circulating VD levels were lower in T1D patients than in healthy controls, our results did not demonstrate a difference in VD intake and sun exposure between T1D patients and healthy controls. However, even then we would recommend VD supplementation for children and adolescents, especially in schools, which are an ideal setting to implement many nutritional strategies, such as a VD campaign by vaccination, or a focus on VD-rich foods in the school canteen program to encourage pupils to make healthy choices. This remains as an alternative, but not a definitive strategy, because alteration in VD levels may be due to an abnormality of enzyme bioconversion to the VD active form, i.e., 1,25(OH)₂D₃, which can consequently lead to its usually immunomodulatory effect.

Conflict of interest

The authors have no conflicts of interest to report.

Footnotes

Acknowledgments

First of all, the authors would like to express their deep gratitude to the Algerian Ministry of Higher Education and Scientific Research, as well as the General Direction of Scientific Research and Technological Development (DGRSDT) for their support of this work. They are also thankful to all the learners and their parents who contributed to this study, as well as the Educational Institutions for their valuable assistance.

References

Ysmail-Dahlouk

, Nouari

, Aribi

. 1,25-dihydroxyvitamin D3 down-modulates the production of proinflammatory cytokines and nitric oxide and enhances the phosphorylation of monocyte-expressed STAT6 at the recent-onset type 1 diabetes. Immunol Lett. 2016;179:122–30. https://doi.org/10.1016/j.imlet.2016.10.002

Ahmed

, Cerosaletti

, James

, Long

, Mannering

, Speake

, Nakayama

, Tree

, Roep

, Herold

, Brusko

. Standardizing T-Cell Biomarkers in Type 1 Diabetes: Challenges and Recent Advances. Diabetes. 2019;68:1366–79. https://doi.org/10.2337/db19-0119

Roep

, Thomaidou

, van Tienhoven

, Zaldumbide

. Type 1 diabetes mellitus as a disease of the β-cell (do not blame the immune system?). Nat Rev Endocrinol. 2021;17:150–61. https://doi.org/10.1038/s41574-020-00443-4

Bener

, Alsaied

, Al-Ali

, Al-Kubaisi

, Basha

, Abraham

, Guiter

, Mian

. High prevalence of vitamin D deficiency in type 1 diabetes mellitus and healthy children. Acta Diabetol. 2009;46:183–9. https://doi.org/10.1007/s00592-008-0071-6

Aribi

, Mennechet

FJD

, Touil-Boukoffa

. Editorial: The role of vitamin D as an immunomodulator. Front Immunol. 2023;14:1186635. https://doi.org/10.3389/fimmu.2023.1186635

Bendik

, Friedel

, Roos

, Weber

, Eggersdorfer

. Vitamin D: a critical and essential micronutrient for human health. Front Physiol. 2014;5:248. https://doi.org/10.3389/fphys.2014.00248

Hossein-nezhad

, Holick

. Vitamin D for Health: A Global Perspective. Mayo Clin Proc Mayo Clin. 2013;88:720–55. https://doi.org/10.1016/j.mayoc2013.05.011

O’Brien

, Jackson

. Vitamin D and the immune system: beyond rickets, Vet J Lond Engl. 1997.. 2012;194:27–33. https://doi.org/10.1016/j.tvjl.2012.05.022

van Schoor

, Lips

. Global Overview of Vitamin D Status. Endocrinol Metab Clin North Am. 2017;46:845–70. https://doi.org/10.1016/j.ecl.2017.07.002

10.

Heyden

, Wimalawansa

. Vitamin D: Effects on human reproduction, pregnancy, and fetal well-being. J Steroid Biochem Mol Biol. 2018;180:41–50. https://doi.org/10.1016/j.jsbmb.2017.12.011

11.

Kovacs

. Maternal vitamin D deficiency: Fetal and neonatal implications. Semin Fetal Neonatal Med. 2013;18:129–35. https://doi.org/10.1016/j.siny.2013.01.005

12.

Pike

, Christakos

. Biology and Mechanisms of Action of the Vitamin D Hormone. Endocrinol Metab Clin North Am. 2017;46:815–43. https://doi.org/10.1016/j.ecl.2017.07.001

13.

Aribi

. Autoimmunity and Immunotherapy of Type 1 Diabetes, Type 1 Diabetes - Pathog. Genet. Immunother. (2011). https://doi.org/10.5772/20674

14.

Mallet

. Vitamine D, J. Pédiatrie Puériculture. 2014;27:29–38. https://doi.org/10.1016/j.j2013.10.007

15.

Chakhtoura

, Rahme

, El-Hajj Fuleihan

. Vitamin D Metabolism in Bariatric Surgery. Endocrinol Metab Clin North Am. 2017;46:947–82. https://doi.org/10.1016/j.ecl.2017.07.006

16.

Djennane

, Lebbah

, Roux

, Djoudi

, Cavalier

, Souberbielle

J-C

. Vitamin D status of schoolchildren in Northern Algeria, seasonal variations and determinants of vitamin D deficiency, Osteoporos. Int J Establ Result Coop Eur Found Osteoporos Natl Osteoporos Found USA. 2017;25:1493–502. https://doi.org/10.1007/s00198-014-2623-7

17.

Malek

. Épidémiologie du diabète en Algérie: revue des données, analyse et perspectives. Médecine Mal. Métaboliques. 2008;2:298–302. https://doi.org/10.1016/S1957-2557(08)71686-9

18.

Touhami

, Zennaki

, Bouchetara

, Naceur

, Aoui

, Gharnouti

, Latroch

, Bouziane-Nedjadi

, Boudraa

. Évolution épidémiologique du diabète de type 1 chez l’enfant: données du registre du département d’Oran, Algérie, 1973-2017. Rev DÉpidémiologie Santé Publique. 2019;67:369–74. https://doi.org/10.1016/j.respe.2019.08.001

19.

Ciqual Table, Ciqual.Anses. (2016). https://ciqual.anses.fr/?fbclid=IwAR2dTb0npMv-HYORL9t4hx0IVHDPeUhPyqRAfIKIf3AD2uv_tN1InPVHJlQ#/cms/telechargement/node/20 (accessed October 31, 2020).

20.

Liese

, Ma

, Mittleman

, The

, Standiford

, Lawrence

, Pihoker

, Marcovina

, Mayer-Davis

, Puett

. Dietary Quality and Markers of Inflammation: No Association in Youth with Type 1 Diabetes in the SEARCH for Diabetes in Youth Study, J. Diabetes Complications. 2018;32:179–84. https://doi.org/10.1016/j.jdiacom2017.10.015

21.

CDC, Achieve and Maintain a Healthy Weight, Cent. Dis. Control Prev. (2020). https://www.cdc.gov/healthyweight/index.html (accessed October 29, 2020).

22.

Charan

, Biswas

. How to calculate sample size for different study designs in medical research? Indian J Psychol Med. 2013;35:121–6. https://doi.org/10.4103/0253-7176.116232

23.

Naing

, Nordin

, Abdul Rahman

, Naing

. Sample size calculation for prevalence studies using Scalex and ScalaR calculators. BMC Med Res Methodol. 2022;22:209. https://doi.org/10.1186/s12874-022-01694-7

24.

Daniel

, Cross

. Biostatistics: a foundation for analysis in the health sciences, Eleventh edition, Wiley, Hoboken, NJ, 2019.

25.

Pourhoseingholi

, Vahedi

, Rahimzadeh

. Sample size calculation in medical studies, Gastroenterol. Hepatol. Bed Bench. 2013;6:14–17.

26.

Goncalves

, Gleize

, Bott

, Nowicki

, Amiot

M-J

, Lairon

, Borel

, Reboul

. Phytosterols can impair vitamin D intestinal absorption in vitro and in mice, Mol. Nutr. Food Res. 2011;55(Suppl 2):S303–311. https://doi.org/10.1002/mnfr.201100055

27.

Ercan

, Bostancib

, Ozmen

, Tekindal

. Is there an association between vitamin D levels and cow’s milk protein allergy at infancy? Arch Argent Pediatr. 2019;117. https://doi.org/10.5546/aa2019.eng.306

28.

Is there an association between vitamin D levels and cow’s milk protein allergy at infancy?

Arch Argent Pediatr. 2019;117. https://doi.org/10.5546/aa2019.eng.306

29.

Gröber

, Kisters

. Influence of drugs on vitamin D and calcium metabolism, Dermatoendocrinol. 2012;4:158–66. https://doi.org/10.4161/derm.20731

30.

Wakeman

, A Literature Review of the Potential Impact of Medication on Vitamin D Status, Risk Manag. Healthc. Policy. 2021;14:3357–81. https://doi.org/10.2147/RMHP.S316897

31.

Salemi

. Pratiques alimentaires des diabétiques. Étude de quelques cas à Oran (Algérie), Économie Rurale Agric. Aliment. Territ. 2010;80-95.https://doi.org/10.4000/economierurale.2816

32.

El-Hajj Fuleihan

, Nabulsi

, Choucair

, Salamoun

, Hajj Shahine

, Kizirian

, Tannous

. Hypovitaminosis D in healthy schoolchildren. Pediatrics. 2001;107:E53. https://doi.org/10.1542/peds.107.4.e53

33.

Puri

, Marwaha

, Agarwal

, Tandon

, Agarwal

, Grewal

, Reddy

DHK

, Singh

. Vitamin D status of apparently healthy schoolgirls from two different socioeconomic strata in Delhi: relation to nutrition and lifestyle. Br J Nutr. 2008;99:876–82. https://doi.org/10.1017/S0007114507831758

34.

Djeraba

, Benlabidi

, Djaballah-Ider

, Medjeber

, Arroul-Lammali

, Belguendouz

, Otmani

, Touil-Boukoffa

. Vitamin D status in Algerian Behçet’s disease patients: an immunomodulatory effect on NO pathway, Immunopharmacol. Immunotoxicol. 2017;39:243–50. https://doi.org/10.1080/08923973.2017.1327967

35.

Amegah

, Nsoh

, Ashley-Amegah

, Anaman-Togbor

. What factors influences dietary and non-dietary vitamin D intake among pregnant women in an African population? Nutr Burbank Los Angel Cty Calif. 2018;50:36–44. https://doi.org/10.1016/j.nut.2017.11.003

36.

Bwibo

, Neumann

. The need for animal source foods by Kenyan children. J Nutr. 2003;133:3936S–3940S. https://doi.org/10.1093/jn/133.11.3936S

37.

Maliou

, Belmadi

, Saadi

, Mahfouf

, Benzidane

, Bitam

. Effect of dairy products intake on breast cancer risk: A case-control study in Algeria. Nutr Clin Métabolisme. 2018;32:187–94. https://doi.org/10.1016/j.nupar.2018.04.001

38.

Pietinen

, Paturi

, Reinivuo

, Tapanainen

, Valsta

. FINDIET Survey: energy and nutrient intakes. Public Health Nutr. 2010;13:920–4. https://doi.org/10.1017/S1368980010001102

39.

Moore

, Murphy

, Keast

, Holick

. Vitamin D intake in the United States. J Am Diet Assoc. 2004;104:980–3. https://doi.org/10.1016/j.jada.2004.03.028

40.

Spiro

, Buttriss

. Vitamin D: An overview of vitamin D status and intake in Europe. Nutr Bull Bnf. 2014;39:322–50. https://doi.org/10.1111/nbu.12108

41.

Derouiche

. Consumption of milk and dairy products in rural environment. Tebessa, Algeria, http://www.lrrd.org/lrrd28/5/meri28089.html (2016) https://lrrd.cipav.org.co/lrrd28/5/meri28089.html (accessed October 31, 2020).

42.

Freisling

, Fahey

, Moskal

, Ocké

, Ferrari

, Jenab

, Norat

, Naska

, Welch

, Navarro

, Schulz

, Wirfält

, Casagrande

, Amiano

, Ardanaz

, Parr

, Engeset

, Grioni

, Sera

, Bueno-de-Mesquita

, van der Schouw

, Touvier

, Boutron-Ruault

M-C

, Halkjaer

, Dahm

, Khaw

K-T

, Crowe

, Linseisen

, Kröger

, Huybrechts

, Deharveng

, Manjer

, Agren

, Trichopoulou

, Tsiotas

, Riboli

, Bingham

, Slimani

. Region-specific nutrient intake patterns exhibit a geographical gradient within and between European countries. J Nutr. 2010;140:1280–6. https://doi.org/10.3945/jn.110.121152

43.

O’Mahony

, Stepien

, Gibney

, Nugent

, Brennan

. The potential role of vitamin D enhanced foods in improving vitamin D status. Nutrients. 2011;3:1023–41. https://doi.org/10.3390/nu3121023

44.

Calvo

, Whiting

, Barton

. Vitamin D intake: a global perspective of current status. J Nutr. 2005;135:310–6. https://doi.org/10.1093/jn/135.2.310

45.

Lips

, Cashman

, Lamberg-Allardt

, Bischoff-Ferrari

, Obermayer-Pietsch

, Bianchi

, Stepan

, El-Hajj Fuleihan

, Bouillon

. Current vitamin D status in European and Middle East countries and strategies to prevent vitamin D deficiency: a position statement of the European Calcified Tissue Society. Eur J Endocrinol. 2019;180:P23–P54. https://doi.org/10.1530/EJE-18-0736

46.

Karlsson

, Andersson

, Hussain

, Bosaeus

, Jansson

, Osmancevic

, Hulthén

, Holmäng

, Larsson

. Lower vitamin D status in obese compared with normal-weight women despite higher vitamin D intake in early pregnancy. Clin Nutr Edinb Scotl. 2015;34:892–8. https://doi.org/10.1016/j.clnu.2014.09.012

47.

Henderson

, Irving

, Gregory

, Bates

, Prentice

, Perks

, Swan

, Farron

. National Diet and Nutrition Survey: Adults Aged 19 to 64 Years — Health Economics Research Centre (HERC), Off. Natl. Stat. (2003). https://www.herc.ox.ac.uk/downloads/health_datasets/browse-data-sets/national-diet-and-nutrition-survey-adults-aged-19-to-64-years (accessed November 1, 2020).

48.

Labour and Welfare (Japan). National Health and Nutrition Survey, Ministry of Health, Japan, 2010.

49.

Kuroda

, Onoe

, Yoshikata

, Ohta

. Relationship between skipping breakfast and bone mineral density in young Japanese women. Asia Pac J Clin Nutr. 2013;22:583–9. https://doi.org/10.6133/apjcn.2013.22.4.10

50.

Mathias

, Jacquier

, Eldridge

. Missing Lunch Is Associated with Lower Intakes of Micronutrients from Foods and Beverages among Children and Adolescents in the United States. J Acad Nutr Diet. 2016;116:667–76.e6. https://doi.org/10.1016/j.jand.2015.12.021

51.

Matsumoto

, Hatamoto

, Sakamoto

, Masumoto

, Ikemoto

. Breakfast skipping is related to inadequacy of vitamin and mineral intakes among Japanese female junior high school students: a cross-sectional study. J Nutr Sci. 2020;9:e9. https://doi.org/10.1017/jns.2019.44

52.

Kondjoyan

, Portanguen

, Duchène

, Mirade

, Gandemer

. Predicting the loss of vitamins B3 (niacin) and B6 (pyridoxamine) in beef during cooking. J Food Eng. 2018;238:44–53. https://doi.org/10.1016/j.jfoodeng.2018.06.008

53.

Moore

, Radcliffe

, Liu

. Vitamin D intakes of children differ by race/ethnicity, sex, age, and income in the United States, to 2010. Nutr Res N. Y. N. 2014;34:499–506. https://doi.org/10.1016/j.nutres.2014.06.002

54.

Wallace

, Reider

, Fulgoni

. Calcium and vitamin D disparities are related to gender, age, race, household income level, and weight classification but not vegetarian status in the United States: Analysis of the NHANES 2001-2008 data set. J Am Coll Nutr. 2013;32:321–30. https://doi.org/10.1080/07315724.2013.839905

55.

Black

, Walton

, Flynn

, Kiely

. Adequacy of vitamin D intakes in children and teenagers from the base diet, fortified foods and supplements. Public Health Nutr. 2014;17:721–31. https://doi.org/10.1017/S1368980013000359

56.

Murphy

, Douglass

, Johnson

, Spence

. Drinking Flavored or Plain Milk Is Positively Associated with Nutrient Intake and Is Not Associated with Adverse Effects on Weight Status in US Children and Adolescents. J Am Diet Assoc. 2008;108:631–9. https://doi.org/10.1016/j.jada.2008.01.004

57.

Thompson

, Byers

. Dietary assessment resource manual. J Nutr. 1994;124:2245S–2317S. https://doi.org/10.1093/jn/124.suppl_11.2245s

58.

Jablonski

, Chaplin

. The roles of vitamin D and cutaneous vitamin D production in human evolution and health. Int J Paleopathol. 2018;23:54–9. https://doi.org/10.1016/j.ij2018.01.005

59.

Tuffaha

, El Bcheraoui

, Daoud

, Al Hussaini

, Alamri

, Al Saeedi

, Basulaiman

, Memish

, AlMazroa

, Al Rabeeah

, Mokdad

. Deficiencies Under Plenty of Sun: Vitamin D Status among Adults in the Kingdom of Saudi Arabia, North Am. J. Med. Sci. 2015;7:467–75. https://doi.org/10.4103/1947-2714.168675

60.

Poduje

, Sjerobabski-Masnec

, Ozanić-Bulić

. Vitamin D–the true and the false about vitamin D, Coll. Antropol. 2008;32(Suppl 2):159–62.

61.

Holick

. Sunlight and vitamin D for bone health and prevention of autoimmune diseases, cancers, and cardiovascular disease. Am J Clin Nutr. 2004;80:S1678–88S. https://doi.org/10.1093/ajcn/80.6.1678S

62.

Khosravi-Largani

, Pourvali-Talatappeh

, Rousta

, Karimi-Kivi

, Noroozi

, Mahjoob

, Asaadi

, Shahmohammadi

, Sadeghi

, Shakeri

, Ghiyasvand

, Tavakoli-Yaraki

. A review on potential roles of vitamins in incidence, progression, and improvement of multiple sclerosis, ENeurologicalSci. 2018;10:37–44. https://doi.org/10.1016/j.ensci.2018.01.007

63.

Webb

, Kline

, Holick

. Influence of season and latitude on the cutaneous synthesis of vitamin D exposure to winter sunlight in Boston and Edmonton will not promote vitamin D3 synthesis in human skin. J Clin Endocrinol Metab. 1988;67:373–8. https://doi.org/10.1210/jcem-67-2-373

64.

Institute of Medecine, Dietary Reference Intakes for Calcium and Vitamin D, National Academies Press (US), Washington (DC), 2011. http://www.ncbi.nlm.nih.gov/books/NBK56070/ (accessed October 31, 2020).

65.

Willett

, Nutritional Epidemiology, Oxford University Press, 1998. https://oxford.universitypressscholarshicom/view/10.1093/acprof:oso/9780195122978.001.0001/acprof-9780195122978 (accessed October 31, 2020).

66.

Ortega

, Pérez-Rodrigo

, López-Sobaler

. Dietary assessment methods: dietary records, Nutr. Hosp. 2015;31(Suppl 3):38–45. https://doi.org/10.3305/nh.2015.31.sup3.8749

67.

Lamri

, Gripiotis

, Ferrario

. Diabetes in Algeria and challenges for health policy: a literature review of prevalence, cost, management and outcomes of diabetes and its complications, Glob. Health. 2014;10:11. https://doi.org/10.1186/1744-8603-10-11