Introduction

Fish oil (FO) consumption carries a noteworthy historical heritage long before they were considered a dietary supplement (DS) [1]. For example, cod-liver oil has been used to treat non-specifically a variety of diseases since 1700, it was given to children in northern Europe and England for centuries, and at the end of the eighteenth century the rheumatism healing effects of cod-liver oil were first mentioned [1].

Regarding their definition, FO are derived from the tissues of oily fish and contain essential omega-3 fatty acids (Ω-3), i.e., eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) [2], although some FO such as cod-liver oil, are also a source of vitamins A and D [3]. Unlike omega-6 fatty acids (Ω-6), Ω-3 are considered essential because humans lack the necessary enzymes required to synthesize them [4], and their intake must therefore be ensured through diet and/or DS. In addition, Ω-3 fatty acids, i.e., EPA and DHA, compared to Ω-6, i.e., arachidonic acid, can lead to less inflammatory products [5].

Interestingly, it has been shown that the ratio of Ω-6 to Ω-3 fatty acids in the diet of humans has changed significantly through the years. Namely, the ratio of Ω-6/Ω-3 was 0.79 in the Paleolithic period [6], 1.00–2.00 in Greece prior to 1960 [7], 4:1 in Japan during 1975 to 1994 [8], an average of 7.0 in Britain during 1986–1987 [9], and 16.74 in the USA of 1998 [6].

Subsequently, concern has been raised due to associations between an increased ratio and chronic diseases, while simultaneously pointing to the benefits of FO consumption [10].

FO, or Ω-3 to be more precise, can be found both as a DS and as a medically prescribed product, yet they are not interchangeable. A key difference between these two lies in that for the latter (i), their manufacturers are required to prove their safety and efficacy before releasing it in the market and (ii) they hold the ability to treat or mitigate a disease [11]–[13]. Meanwhile, several health claims for EPA and DHA have been authorized by the European Food Safety Authority (EFSA) to accompany DS that contain them in certain concentrations, e.g., maintenance of normal blood pressure, heart function, and triglyceride levels [14]. On the other hand, Ω-3, as an active medical ingredient, are approved to treat hypertriglyceridemia [15].

Indeed, even nowadays, FO use is particularly pronounced in countries like England [16]–[18] and Norway [17]; meanwhile a tenfold increase in their use from 1.3 to 12.0% was observed in the USA, from 1999 to 2012 [19]. While history of FO accounted for about 6.5% of our DS users, we detected significant differences between DS users who had used FO (among other DS) at least once in their lifetime (DS + FO) and those who had used DS other than FO (DS − FO) regarding their stance towards DS, with DS + FO having more favorable attitudes towards them [20].

The literature suggests that consumers seem to use FO for a variety of reasons including both specific, such as heart health [21], and non-specified ones, such as general health [22, 23]. Meanwhile, studies in which FO use was considered when dividing their participants have come across significant results regarding the relationship between FO use and the presence of certain diseases, e.g., arthritis [18, 23].

Building on our previous study [20], and due to the scarcity of relevant research data in Greece, we sought to further our understanding of the behavioral differences between DS + FO and DS − FO by setting a number of aims, namely, (i) What are the reasons and duration regarding DS use and its perceived positive and negative outcomes by DS + FO? (ii) What beliefs are held by DS + FO with regards to the efficacy of DS as opposed to a balanced diet? (iii) How many DS + FO are going through medical examinations or are supervised before and after using DS, respectively? (iv) How many DS + FO report adherence towards the recommended dietary allowance (RDA) and are subjectively cognizant of the dangers of DS overuse?

Material and methods

Study design

Details of the methodology of this study (e.g., approval of the research protocol) are presented in our previously published work [24]. However, key points are reported below. Our sample was collected in 2018–2019 and was representative of the general Greek population in terms of sex. According to the 2011 census, the Greek population consisted of 51.0% women and 49.0% men. Accordingly, the collected sample consisted of 53.0% women and 47.0% men.

The initial sample size was 31,824 Greek citizens. Incomplete questionnaires were discarded, leaving 28,491 respondents. Those who had used a DS at least once in their lifetime were defined as DS users (15,608; 54.8%), while those who had never used a DS (12,494; 43.9%) were not included in the analysis. The remaining 389 (1.4%) respondents who could not recall whether they had used a DS or not were also not included in the analysis, resulting in a final sample size of 15,608. Among DS users, if FO were among their reported DS, they were defined as DS + FO (1001; 6.4%); otherwise, they were defined as DS − FO (14,607; 93.6%).

The assumed definition of DS was as follows: “foodstuffs the purpose of which is to supplement the normal diet and which are concentrated sources of nutrients or other substances with a nutritional or physio-logical effect, alone or in combination, marketed in dose form, namely forms such as capsules, pastilles, tablets, pills and other similar forms, sachets of powder, ampoules of liquids, drop dispensing bottles, and other similar forms of liquids and powders de-signed to be taken in measured small unit quantities” (Directive 2002/46/EC, European Parliament and Council, 2002) [25].

Statistical analysis

For all of the categorical variables, frequencies were presented overall and according to DS use. Pearson’s chi-square test was used to detect the existence (or non-existence) of statistically significant dependence between the categorical variables [26]. The analysis was performed using IBM SPSS Statistics v26.0. The significance level was set at α = 0.05.

Results

The total analyzed sample consisted of 15,608 respondents. There were totally 1001 (6.4%) DS + FO and 14,607 (93.6%) DS − FO.

Reasons for DS use

According to Table 1, the top 6 reasons for DS use, regardless of viewpoint, i.e., overall or separately for DS + FO and DS − FO, are the same. Namely, in a descending order, they were the improvement of physical condition (39.7%; 49.9 vs 39.0%), nutrient deficiency/ies (35.9%; 48.1 vs 35.0%), pathological conditions (24.8%; 33.2 vs 24.3%), prevention of health problems (22.0%; 29.8 vs 21.5%), increase in muscle mass (16.0%; 24.1 vs 15.4%), and increase in sports performance (15.6%; 24.1 vs 15.0%). Also, every one of the reasons (without considering the subcategories of “other reasons”) was selected by a significantly higher percentage of DS + FO, compared to DS − FO, except for “pregnancy/birth,” “anemias,” “allergies and flues,” “autoimmune disorders,” and “other reasons,” where significant differences were absent (p > 0.05).

Table 1 Absolute and relative frequencies in parenthesis (%) of the reasons for DS use. The p-values of chi-square tests are presented. Statistically significant dependence between DS use status and selection of answer is indicated with asterisks (DS + FO, DS users who had used fish oils (FO) among other DS; DS − FO, DS users who had used DS other than FO; **p < 0.01, ***p < 0.001)
Full size table

Additionally, almost half of DS + FO (47.6%) selected 3 or more reasons for DS use compared to 25.4% of DS − FO (Supplementary table 1), that is, if the subcategories of “pathological conditions” and “other reasons” are not taken into account. Hence, around one third of DS + FO (34.0%) selected 2 or more pathological conditions as opposed to 20.8% of DS − FO (Supplementary table 2).

Medical assessment and knowledge/opinions regarding DS use

According to Table 2, significant dependence (p < 0.05) was detected between the DS + FO and DS − FO subgroups and the answers given on the above outcome variables, except for the one exploring whether DS users underwent medical or hematological examinations before DS use (p > 0.419).

Table 2 Absolute and relative frequencies in parenthesis (%) regarding medical assessment, knowledge, and opinions revolving around DS use. The p-values of chi-square tests are presented. Statistically significant dependence between DS use status and selected answers is indicated with asterisks (RDA, recommended dietary allowance; DS + FO, DS users who had used fish oils (FO) among other DS; DS − FO, DS users who had used DS other than FO; **p < 0.01; *** p < 0.001)
Full size table

While 48.3% go through medical examinations before using DS, 50.6% remain medically unsupervised during DS use. Despite the lack of a statistically significant difference between the percentages of DS + FO and DS − FO that are medically examined before DS use, significant differences were detected when the examined dependent variable was medical supervision during DS use (p < 0.001), i.e., significantly more DS + FO were having examinations on their own initiative (21.6 vs 17.1%) rather than being supervised by a doctor (30.1 vs 32.1%).

Overall, the great majority of DS users (78.8%) take the RDA of the DS active ingredients into consideration, although significant dependence between the DS use status and the answers to this question was detected (p = 0.018). Specifically, RDA consideration was more prevalent among DS + FO (82.1 vs 78.5%), unlike the ignorance of it (6.9 vs 9.1%).

Even though 80.4% of DS users reported knowledge of the dangers and/or side-effects of DS overuse, 48.8% of DS users reported that confidently. Also, 19.7% of DS users declared not knowing the dangers and/or side effects of DS overuse. However, subjective knowledge of such dangers was reported with confidence significantly (p < 0.001) more frequently among DS + FO (53.6 vs 48.4%) who also displayed less frequently ignorance (15.1 vs 20.0%).

Despite having taken or currently taking DS, 70.0% of DS users believe confidently or with some doubt (33.4 and 36.6%, respectively) that a balanced diet without DS can ensure the desired results, while 9.6% reported not knowing whether that statement was true or not. Nonetheless, disagreement was significantly (p < 0.001) more prevalent among DS + FO (25.5 vs 20.1%).

Duration of DS use

Based on Fig. 1, there was a statistically significant dependence between the DS use categories and the duration of DS use (p < 0.05). Overall, most respondents reported using DS for 1 month (35.8%), yet, while most DS − FO reported 1 month as well (35.8%), most DS + FO reported higher periods of time, i.e., DS use systematically for more than 1 month (42.9 vs 30.8%). More details can be found at Supplementary table 5.

Fig. 1
figure 1

Relative frequencies (%) of the duration of dietary supplement (DS) use. Asterisks indicate significance of the chi-square test of independence (DS + FO, DS users who had used fish oils (FO) among other DS; DS − FO: DS users who had used DS other than FO; ***p < 0.001)

Full size image

Positive and negative outcomes from DS use

According to Table 3, the top 6 positive outcomes from DS use, overall and for both, DS + FO and DS − FO alike, were health (44.7%; 55.5 vs 44.0%) and fitness (41.5%; 52.6 vs 40.7%) improvements, along with restoration of nutrient deficiencies (23.6%; 27.0 vs 23.3%), increase in muscle mass (15.4%; 23.0 vs 14.8%), weight loss (14.2%; 17.1 vs 14.0%), and weight gain (6.6%; 12.1 vs 6.2%). However, there was a significant dependence between the DS use status and the majority of the reported positive outcomes (p < 0.05). Namely, every positive outcome was selected by a higher percentage of DS + FO, compared to DS − FO, except for the “other positive outcomes” where significance was absent (1.8 vs 1.6%; p > 0.05).

Table 3 Absolute and relative frequencies in parenthesis (%) of the positive and negative outcomes experienced during dietary supplement (DS) use. The p-values of chi-square tests are presented. Statistically significant dependence between DS use status and selected answers is indicated with asterisks (DS + FO, DS users who had used fish oils (FO) among other DS; DS − FO, DS users who had used DS other than FO; *p < 0.05; **p < 0.01; ***p < 0.001)
Full size table

Regarding negative outcomes from DS use, the majority of DS users overall reported having experienced no negative outcomes, without a significant difference between DS + FO and DS − FO (73.0%; 72.9 vs 73.1%, p = 0.876). Unlike the reported positive outcomes, for the majority of the reported negative outcomes, no significant dependence (p > 0.05) was found between them and the DS use categories, i.e., except for pressure disorders which was selected by more DS + FO (3.7%; 5.9 vs 3.5%, p < 0.001). Yet, the top 3 negative outcomes for both groups were tachycardia (8.5%; 9.2 vs 8.4%), diarrhea (5.4%; 5.7 vs 5.4%), and headache (5.3%; 5.1 vs 5.3%). Lastly, 55.7% of DS + FO vs 38.0% of DS − FO selected 2 or more positive outcomes (Supplementary table 3), while only 6.2% of DS + FO vs 3.7% of DS − FO selected 2 or more negative outcomes (Supplementary table 4).

Discussion

Prevalence, reasons, and outcomes from DS use

According to our 2018–2019 data, DS + FO and DS − FO consisted 6.5 vs 93.5% of the overall DS users, or 3.5 vs 51.3% of our overall sample, respectively.

Almost every reason for DS use was selected by a higher percentage of DS + FO compared to DS − FO, yet the top reasons for DS use for both groups were the same, i.e., improvement of physical condition, nutrient deficiencies, and treatment of pathological conditions. The latter includes noteworthy examples such as arthritis and cardiovascular disease (CVD). In fact, DS + FO were more likely to report a higher number of reasons for DS use, which may be correlated with the fact that they use or have used a higher number of DS [20]. Although the majority of DS + FO and DS − FO did not notice any side effects (\(\cong\) 73.0%), DS + FO had perceived a higher number of positive outcomes from DS use. Also, disagreement towards the statement “a balanced diet without DS would be adequate to ensure the desired results” was significantly higher among DS + FO compared to DS − FO.

In fact, such disagreement was confidently reported by about 20% DS users. A possible explanation for the respondents’ rationale might be that certain groups may face dietary limitations or conditions where nutrient intake from food alone would be difficult, hence, leading to nutrient deficiencies. Indeed, dietary limitations can come in different forms. For example, special dietary requirements may take place in certain conditions such as pregnancy and lactation (e.g., increased need for folic acid) [27]. Another case would be having dietary restrictions, such as in vegan/vegetarian diets that exclude animal products, leading to an increased risk for vitamin B12 deficiency [28]. Additionally, athletes or physically active consumers might seek to optimize their performance and recovery [29]. Lastly, since dietary needs dynamically change throughout life, individuals in certain life stages such as infancy, adolescence, and aging, have unique nutritional requirements [27] which might be difficult to achieve through diet alone. For example, it is less probable for seniors, compared to younger individuals, to get dietary vitamin D from their diet or to get sunlight exposure [27].

Studies exploring the reasons for FO use, specifically, have found general health or well-being to be a very commonly cited reason [22, 23] followed by cognitive performance [22] or joint health [23]. General well-being was the top reason, even in a clinical setting with cardiac patients where cardiovascular health came in second place [30]. Indeed, studies by Parmenter et al. have shown that most adult participants (regardless of whether they were FO users or not) believed that a specific FO omega-3 product would support their overall health and well-being, followed by cognitive improvement [31], yet, among the perceived benefits, regarding children, parents most commonly reported growth/development, which was then followed by cognitive improvement [32]. Unlike the previous studies, a US study showed heart health or the lowering of cholesterol to be the most common reasons for taking FO or omega-3 DS [21].

Interestingly, in a UK study, while few respondents reported having experienced any negative effects from DS use, a considerable number took them without knowing whether they were experiencing any positive outcomes [16]; similar doubt was also observed in a Dutch study [33]. In addition, Blendon et al. pointed out that more than 65.0% of their DS users would continue taking DS even if they were shown to be ineffective by the Food and Drug Administration or a government agency [34].

FO use has been severally associated with the presence of pathological conditions. For example, FO use has been shown to be more likely among individuals suffering from arthritis [18, 23] and musculoskeletal [31, 35] or cognitive conditions [31], but less likely among those reporting CVDs [18, 23], yet a UK study found that hypertension and high cholesterol were more prevalent among FO users compared to FO non-users [36].

Medical supervision and knowledge regarding proper DS use

DS + FO were more likely to display professional-independent behaviors. While approximately the same percentages of DS + FO and DS − FO undergo medical and/or hematological examinations before using DS (\(\cong\) 50.0%), DS + FO are more likely to self-monitor themselves (21.6 vs 17.1%) by having examinations (e.g., blood tests), rather than being supervised by a doctor (30.1 vs 32.1%). Unfortunately, around 50.0% of DS users remain medically unsupervised.

We previously found that DS + FO, compared to DS − FO, were informed or were recommended DS by more sources, with doctors and pharmacists being the main ones [20]. However, while almost every source was selected by a higher percentage of DS + FO, “the internet” seemed to be significantly more prevalent among DS + FO, especially as a source of information (41.3 vs 28.9%) as it closely contented with doctors (47.8 vs 46.6%) and pharmacists (49.1 vs 41.6%) [20]. Given the abundance of information on the internet, and its close ties with DS + FO, this might explain the more prevalent independent behavior.

Elsewhere, we have discussed the interplay between DS consumers (with and without differentiating between DS + FO and DS − FO) and their sources of information/recommendation in accordance with the literature [20]. Subsequently, one could see the need for a deeper involvement of healthcare professionals and the noticeable impact of other sources such as friends and family. Below, we provide some instances, showing the lack of communication of information regarding DS use between consumers and healthcare professionals.

In a sample of cardiac patients, the majority (86.0%) consumed FO in the absence of a recommendation by healthcare professionals [30], yet another study, on a general population, found that frequent visits to general practitioners and alternative therapists was more probable for FO users [23]. A New Zealand study found that about 40.0% of DS users and 50.0% of those who use both, DS and medications, did not inform their general practitioner about their DS use, with the top reasons being “it never came up” and “I was never asked” [22]. Additionally, a Serbian study, found that around 90.0% of their respondents viewed the provision of information regarding herbal DS use, or relevant remedies, to healthcare professionals as unimportant, with 50.0% consuming DS based on advice by non-healthcare professional sources [37].

We previously showed that higher percentages of DS + FO, compared to DS − FO, had used DS and were knowledgeable regarding DS label comprehension [20]. This is reinforced by our current findings that mark DS + FO as more confident and knowledgeable regarding DS use. Specifically, a higher percentage of DS + FO reports knowing and adhering to the RDA of the consumed DS active substances (82.1 vs 78.5%) while also declaring that they surely, or with some doubt, know the dangers of DS overuse (84.9 vs 80.0%). Lastly, DS + FO, compared with DS − FO, report DS usage for higher periods of time, i.e., systematically more than 1 month (42.9 vs 31.6%).

Hilleman, Teply, and Packard concluded that the intake of omega-3 fatty acids by most of their FO-consuming cardiac patients was rather relatively low [30]. Additionally, Backes et al., who analyzed a sample of patients using prescribed FO in a lipid-specialty clinic, found that, regarding the total daily dose of EPA and DHA, only 28.9% could confidently recall it, 33.0% could provide an estimate, and 38.2% had no knowledge of it; unfortunately, overall, the actually consumed dose was less than 50.0% compared with the prescribed one [38]. Again, Hill et al. found that most of their respondents, regardless of whether they were suffering from rheumatoid arthritis, were consuming FO at a suboptimal level relating to its symptomatic relief [23]. Without taking into account the influence of diet, Mengelberg et al. underlined that approximately 26.0% of their New Zealander respondents consumed a dose that met EPA and DHA recommendations, unlike most of the remaining respondents who consumed less than the recommended quantity (2.7% did not know the RDA and 11.0% did not give an answer) [22].

Now, in general, a UK study found that even though most consumers were adhering to the dosing recommendations, it was deemed unimportant to not strictly follow these recommendations [16], e.g., certain respondents would sometimes consume a higher amount of vitamin C due to having a cold [16]. Leung and Lum, regarding the “upper limit dosage,” pointed out that, in Hong Kong, among parents who gave their kindergarten aged children DS, 64.0% of parents reported knowing it, 26.0% acknowledged its existence but could not specify it, and 10.0% did not know if there was any limit [39].

Regarding usage patterns, Mengelberg et al. found that 53.4% had taken FO for more than 12 months, while another 53.4% reported taking them every day [22]. Among Norwegian women who used cod-liver oil DS, around 35.0% were daily, whole-year users [40]. Similarly, in the UK, around 2 out of 3 DS users had consumed DS for at least a year [16]. Even though (i) many, if not most, of our respondents reported systematic use for more than 1 month and (ii) the reported duration was not directed to a specific DS, our durations are significantly lower than those of the aforementioned studies (i.e., only around 10.0% had used DS for at least a year according to Supplemental table 5).

Conclusions

Before using DS, only 1 in 2 DS users go through medical/hematological examinations, and during DS use, only 1 out of 3 were medically supervised. Yet a significantly higher percentage of DS + FO, compared to DS − FO, self-assesses themselves during DS use (e.g., via blood tests). Also, DS + FO were more likely to report a longer duration, a higher number of reasons, and more perceived benefits regarding DS use, while also thinking of their goals as unachievable with the help of a balanced diet without DS. However, a higher percentage of DS + FO was subjectively cognizant of the active ingredients’ RDA and the dangers of overuse.