The present study clearly confirms previous observations stating that obese individuals are characterized by micronutrient deficiencies [4
]. The deficiencies are suggested by both low intake and low serum and intracellular levels as shown by our data.
The study demonstrates an insufficient dietary micronutrient supply of retinol, ß-carotene, vitamin D, vitamin E, vitamin C, folate, iron, and calcium in obese individuals. A highly significant difference compared to the reference population, both in men and women, was observed for ß-carotene, folate, vitamin C, and fiber - nutrients which are essentially found in fruits, vegetables and whole-grain products, thus pointing to an unbalanced diet leading to micronutrient deficiencies.
Evaluation of nutritional intake has some methodological weakness such as underreporting that limits the interpretation of dietary record data. The resulting error is possibly even more distinct in obese subjects [21
]. However, underreporting especially applies to carbohydrate-rich snacks [25
], which are low in micronutrient content by nature, and not to ‘healthy’ food. Therefore we assume that this error is negligible regarding micronutrient intake.
Before intervention, obese subjects showed deficiencies of several micronutrients in serum or BMC, namely 25(OH) vitamin D, vitamin C, selenium, iron, ß-carotene and lycopene. We did not find any statistically significant correlation between micronutrient intake and serum or BMC concentrations, which could be due to the limited subject number. However, the reported intakes of ß-carotene, vitamin D, vitamin C, and iron were clearly below DRI and point to a relation between micronutrient intake and body micronutrient status.
Based on our data, we conclude that a DRI-covering low-calorie formula diet does not meet the demands of obese individuals. The reasons can be manifold and could cover metabolic alterations during a period of major weight loss, unbalanced dispersal of lipophilic compounds and fat-tissue specific oxidative stress. Indeed, we observed even more subjects with deficiencies in some micronutrients after a three-month period of formula diet compared to the baseline status before intervention. In particular, vitamin C, selenium, iron, zinc, and lycopene deficiencies increased or could not be corrected by protein-rich formula diet containing vitamins and minerals according to DRI. The formula meal replacement products contained micronutrients in amounts often even higher than those recommended for the general population. We also observed an increase in subjects with calcium deficiency. However, serum calcium concentration is not an adequate measure for dietary calcium intake, but indicates an electrolyte imbalance induced by weight loss and accompanying fluid changes in the body.
Possible interaction effects between pharmaceuticals and micronutrients have to be considered, which may account for increased demands. Within this study, frequency and dosage of drug use did not change during the LCD period in most of the cases and thus most likely did not confound the results. In 36% of the subjects the dosage of at least one drug was slightly reduced, which would, if anything, have improved micronutrient bioavailability.
We did not find statistically significant correlations between vitamin E in serum and BMC, but the decrease in BMC concentrations could point to the fact that intracellular levels respond more sensitively to altered oxidative stress, which particularly occurs in a period of major weight loss [26
]. This is explained by an upregulation of the renin-angiotensin system and a reduction of glutathione and glutathione peroxidase in erythrocytes, resulting in higher concentrations of reactive oxygen species which again promote the metabolic syndrome [6
]. All three lipophilic vitamins negatively correlated with total body fat, assessed by bioelectric impedance analysis. This is most likely due to the storage capacity for lipophilic compounds, which is so far only suggested for 25(OH)D [27
]. It is tempting to speculate that this also applies to other lipophilic compounds. Thus, a higher amount of fat tissue could lead to an increase in accumulation of lipophilic vitamins, which in turn are lacking in the serum pool. The positive association between adipose tissue mass and systemic 25(OH)D concentrations, the almost 2fold reduction of 25(OH)D deficiency following LCD, and the significant increase in 25(OH)D serum concentrations observed in our study all suggest storage in adipose tissue and release during weight loss, also described elsewhere [29
McClung et al.
] hypothesized that obesity influences iron absorption by inflammatory mediated mechanisms. Proinflammatory cytokines promote hepcidin release in liver and fat tissue, which is involved in iron homeostasis, inhibiting absorption in enterocytes [31
]. This hypothesis is supported by a significantly negative correlation of iron concentration with CRP levels observed in our obese subjects.
Measurement of vitamins in blood samples might not reflect the amount of vitamins absorbed or the concentration in tissues [32
]. To gain information on body distribution we monitored both serum and intracellular levels of vitamins. BMC are easily available and serve as a model system for assessment of the nutritional and antioxidative status, as well as to control for the success of supplementation and effect of medicamentous therapies [33
]. With a cellular turnover of 5 to 25
days, BMC are supposed to reflect the current cellular supply [36
We evaluated the level of vitamin C, lycopene, α-tocopherol, and ß-carotene and thus the antioxidative capacity directly in BMC. These dietary-derived antioxidants are an important component to support the exogenous antioxidative system [33
]. Except for ß-carotene and α-tocopherol, which were only slightly reduced, we observed a significant decrease in vitamin C and lycopene during the three-month formula diet suggesting again an enhanced oxidative stress in obese individuals and thus a higher demand of antioxidative vitamins compared to healthy, normal weight subjects, especially in a period of major weight loss.
The current choices for functional markers of vitamin C status are vitamin C concentrations in plasma and leukocytes. Plasma or serum vitamin C levels are highly sensitive to recent dietary intakes, but may not reflect tissue content as well as leukocyte levels [37
]. Serum vitamin C concentrations remained unchanged during the vitamin C enriched formula diet, whereas leukocyte levels improved. Leukocyte concentrations more reliably display long-term supply and deficiencies [37
]. Bioavailability of vitamin C is a complex issue involving distribution to the tissues and utilization by the tissues [35
]. Before intervention, leukocyte levels were depleted and might replete preferentially due to their high metabolic priority [37
]. Contrary to our expectations intracellular vitamin C concentrations in BMC significantly decreased. To obtain a state of complete saturation, the repletion dynamics of vitamin C in certain tissues may be more specific than others when the vitamin intake during repletion is limited [37
]. However, the results should be interpreted carefully because of the small sample size, the variation of vitamin C content throughout cell types and the reliability of sampling and analysis procedure due to the unstable nature of vitamin C. In contrast to the fat-soluble antioxidants α-tocopherol, lycopene and ß-carotene, which were demonstrated to be reliable markers in BMC [32
], the capacity of BMC vitamin C concentrations and its reliability as a biomarker for vitamin C status has not been described so far.
Interestingly, we observed a significant increase in all antioxidants measured in BMC at the end of the program year. This subgroup only included nine subjects, but suggests a change in dietary habits following intensive weight reduction and diet counseling. These subjects were successful in losing weight and were able to maintain this weight after the formula diet, therefore most likely putting their nutritional knowledge into practice. However, this finding needs to be confirmed by other studies.
A potential limitation of the study is that the dietary assessment strategies differed in the two study populations. In the NNSII a comprehensive dietary history method including cross-check features was used. The dietary record approach in this study was chosen to provide quantitatively accurate information on food consumption as well as influence of food processing, but may be biased due to the estimation of the weight of food consumed in some cases instead of weighing, and also by affecting eating behavior. Moreover, for several micronutrients simple blood concentrations were measured, which might not always reflect a complete picture of the nutritional status. At this stage we only included a small sample size, which limits the explanatory power of the study results.