Effect of Changes in Food Groups Intake on Magnesium, Zinc, Copper, and Selenium Serum Levels During 2 Years of Dietary Intervention

Ofra Paz-Tal, Ayala Canfi, Rachel Marko, Esther Katorza, Zeev Karpas, Iris Shai, Dan Schwarzfuchs, Einat K. Sheiner

Research output: Contribution to journalArticlepeer-review

6 Scopus citations


No significant associations between changes in food groups intake and the 4 elements were found in the Mediterranean diet group.

Conclusions: During this 2-year intervention, serum concentrations of 4 essential elements were associated with a diversity of food group intake patterns. Comprehensive predictors for elevating zinc, copper, and selenium in serum included decreasing consumption of sweets and cakes while increasing consumption of bread, cereals, and pasta.

Background: Essential elements in serum are related to specific changes in food groups intake.Objective: To address the effect of 2-year food intake changes in an intervention study on serum concentrations of magnesium, zinc, copper, and selenium.Method: Two hundred thirty-one participants, a subgroup of the Dietary Intervention Randomized Control Trial (DIRECT) study (age = 52 years; body mass index = 32.8 kg/m2; 85% males) randomized to low-fat, Mediterranean, or low-carbohydrate diets in a 2-year dietary intervention trial were followed for serum concentrations determined using inductively coupled plasma–mass spectrometry. Changes in the intake of 11 food groups were evaluated by food frequency questionnaires.

Results: Using multivariate regression models, adjusted for age, sex, baseline body weight (kg), and changes in intakes of 11 food groups (g/d), at 12 months, serum element elevations were observed mainly in the low-carbohydrate group: selenium, by increasing consumption of fats and oils (β = 0.415, p = 0.009) and legumes (β = 0.183, p = 0.010) and decreasing fruit intake (β = −0.438, p = 0.030); copper, by increasing consumption of legumes (β = 0.453, p = 0.018) and dairy products (β = 0.320, p = 0.039); magnesium by increasing fish consumption (β = 0.374, p = 0.042) in the low-carbohydrate group and in the entire study population (β = 0.237, p = 0.016); and zinc exclusively in the low-fat group by decreasing consumption of fats and oils (β = −0.575, p = 0.022). At 24 months, serum elements were elevated mainly in the low-fat diet group, mostly by decreasing intake of snacks, sweets, and cakes: zinc (β = −0.570, p = 0.027), copper (β = −0.649, p = 0.012), and selenium (β = −0.943, p < 0.001). Also in this group, magnesium levels were elevated by increasing vegetable intake (β = 0.395, p = 0.041), copper by increasing fruit intake (β = 0.375, p = 0.025), and selenium by increasing consumption of bread, pasta, and cereals (β = 0.751, p = 0.011). The entire group, further adjusted to assigned diet type, increased selenium (β = 0.294, p = 0.004) and copper (β = 0.220, p = 0.038) by increasing consumption of bread, pasta, and cereals; selenium level was also predicted by decreasing consumption of snacks, sweets, and cakes (β = −0.256, p = 0.014).

Introducing energy expenditure, expressed in metabolic equivalents (MET = 1 kcal·kg−1·h−1), as an additional variable emphasized the negative effect of sweets and cakes on increasing serum concentrations of zinc, copper, and selenium after 24 months (β = −0.549, p = 0.021; β = −0.669, p = 0.012; β = −0.982, p < 0.001, respectively), especially in the low-fat diet group.

Original languageEnglish
Pages (from-to)1-14
Number of pages14
JournalJournal of the American College of Nutrition
Issue number1
StatePublished - 2 Jan 2015


  • ICP-MS
  • METs
  • Mediterranean diet
  • copper
  • essential elements in serum
  • food groups
  • low-carbohydrate diet
  • low-fat diet
  • magnesium
  • selenium
  • weight loss
  • zinc

ASJC Scopus subject areas

  • Medicine (miscellaneous)
  • Nutrition and Dietetics


Dive into the research topics of 'Effect of Changes in Food Groups Intake on Magnesium, Zinc, Copper, and Selenium Serum Levels During 2 Years of Dietary Intervention'. Together they form a unique fingerprint.

Cite this