Main

Mother's milk is optimal with respect to calcium absorption by the infant (1). Dietary PUFAs and LCPUFAs affect calcium absorption and balance in animals (2). AA and DHA are present in breast milk (3), but are not considered essential fatty acids. Whether AA and DHA are responsible for enhanced calcium absorption in mother's milk is unknown.

Until recently, formula in North America contained only the essential n-6 and n-3 PUFAs, linoleic acid (LA), and α-linolenic acid (ALA). In 2002, the U.S. Food and Drug Administration and Health Canada approved the supplementation of infant formula with AA and DHA. Today, most companies have marketed formula containing AA and DHA, yet the effect of formula with and without AA and DHA on bone mineralization and calcium balance remains unclear. In preterm infants, dietary AA as 0.5 g/100 g of fat and DHA as 0.3 g/100 g of fat in formula had no effect on mineral balance (calcium, phosphorus, magnesium, zinc) (4). Also in preterm infants, dietary AA (0.46 g/100 g of fat) and DHA (0.26 g/100 g until hospital discharge followed by 0.16 g/100 g of fat) in formula had no benefits to bone mass through to 1 y corrected age (5). This is in contrast to animal-based research whereby feeding formula with added AA and DHA (0.5:0.1 g/100 g fat) to piglets increases bone mass by 8% to 22% compared with formula with only LA and ALA (6–8). However, higher amounts of AA:DHA (1.0:0.2 and 2.0/0.4 g/100 g of fat) offered no additional benefit despite keeping the ratio of AA:DHA constant at 5:1 (8).

A proposed mechanism for elevations in bone mass in response to dietary n-6 and n-3 PUFAs and LCPUFAs is altered mineral metabolism (2). Bone contains many minerals, including calcium, phosphorus, magnesium, and zinc, each of which is essential to bone mineralization. Calcium and phosphorus provide structural integrity in the skeleton; 99% of calcium and 85% of phosphorus are in the skeleton as hydroxyapatite (9). Approximately 67% of the body's magnesium is located on the hydroxyapatite surface of bone (9). Magnesium plays a role in calcium metabolism as a cofactor of enzymes required for synthesis of calcitriol (10) and parathyroid hormone (10–13). Thus, magnesium is involved in bone mineral homeostasis and influences hydroxyapatite formation (14). Magnesium affects bone cell function; during deficiency, the number of osteoblasts is decreased and the number of osteoclasts is increased (15,16). Approximately 29% of zinc is found in bone (17). Zinc is essential in bone metabolism as a cofactor of enzymes (18), including alkaline phosphatase, an important enzyme in bone mineralization (19). Elevations in bone calcium, phosphorus, magnesium, or zinc content could enhance bone mineralization and may explain the higher bone mass seen with dietary LCPUFAs.

The affect of different PUFAs and LCPUFAs on mineral metabolism is unclear. In growing rats, Claassen et al. (2) found that dietary γ-linolenic acid (GLA) plus eicosapentaenoic acid (EPA) and DHA increased calcium absorption, calcium balance, and bone calcium content. Dietary AA:DHA as 0.5:0.1 g/100 g of fat had no affect on calcium absorption in piglets following 15 d (6) or on calcium or phosphorus in tibial diaphysis, plasma, or urine (6,7). It is possible that a higher amount of AA:DHA, as used by Claassen et al. (7.4 g/100 g of fat as GLA, 2.5 g/100 g of fat as EPA and 0.7 g/100 g as DHA) (2), might enhance mineral balance. Additionally, the effects of dietary LCPUFA on bone magnesium and zinc content in piglets have not been investigated. Thus, the objectives were to determine the effects of varying amounts of AA:DHA (0.5:0.1, 1.0:0.2, 2.0:0.4 g/100 g of fat) in formula versus control formula on 1) calcium absorption, 2) femur mineral composition, and 3) mineral excretion.

METHODS

Animals and diet.

Male piglets (n = 40) born at the Glenlea Swine Research Unit, University of Manitoba, were transported to the housing facility at the main campus of the university. The average birth weight (± SD) of piglets born at this institution is 1.6 ± 0.2 kg. Piglets were selected with a birth weight of ≥1.4 kg, from 10 litters consisting of 10–12 piglets each with at least four males. Piglets arrived on d 3 of life and were taught to lap liquid formula. Piglets were fed by gavage feeding plus lapping of formula to ensure enough formula was consumed to continue growth for 2 d of adaptation. The strength of control formula was progressively increased from half strength on d 3 to full strength by the end of d 4 of life, followed by experimental or control formula on d 5. Piglets were housed individually in stainless steel cages, and room temperature was maintained at 29–30°C. Based on 0900 h weight, the piglets were offered 350 mL/kg/d of liquid formula. This amount was chosen based on energy requirements (20) and has been shown to support growth in piglets (7,21). This amount was divided into three equal portions provided at 0900, 1500, and 2100 h for 15 d (from 5 to 21 d of life) per Weiler and Fitzpatrick-Wong (7). Animal care and procedures were examined by the University of Manitoba Committee on Animal Use and were within the guidelines of the Canadian Council of Animal Care (22). This study was a part of a large piglet study that also studied the effects of dietary AA:DHA on bone mass (8).

Piglets were randomized on arrival and within litters to receive one of four dietary treatments. Treatments were control formula or control supplemented with 0.5:0.1 g AA:DHA, 1:0.2 g AA:DHA, or 2:0.4 g AA:DHA per 100 g of fat. Supplementation was held at a constant n-6:n-3 ratio of 5:1. The AA was provided in the form of RBD-ARASCO (40.6 g/100 g of fatty acids as AA) and DHA in the form of RDB-DHASCO (40.0 g/100 g of fatty acids as DHA). AA was derived from a common soil fungi and DHA was derived from a marine microalgae (Martek Biosciences Corp.). These sources were chosen because they were previously used in our laboratory and because they are used in the manufacturing of many infant formula products. Formulas were isocaloric with equal amounts of fat. The formula was based on nutritional requirements for healthy growing piglets between 3 and 10 kg as set by the National Research Council (21) and currently proven to support growth in our laboratory (7). The dietary composition of the control formula and dietary composition of PUFAs in each treatment formula has been previously published (8). Formula contained 1050 kcal/L, 60 g/L fat, 50 g/L protein, 2.1 g/L calcium, 1.4 g/L phosphorus, and 6.3 mg/L zinc. Piglets were allowed approximately 1 h of exercise before each feed.

Sample collection.

Urine was collected over a 24-h period using metabolic cages starting on d 14 and ending on d 15 of study. Urine was collected from 2100 h to 0900 h, 0900 h to 1500 h, and 1500 h to 2100 h and then pooled for analysis. On d 16, the piglets were anesthetized by intraperitoneal injection of sodium pentobarbital (30 mg/kg, 65 mg/mL concentration) or isofluorane gas, calcium absorption was measured, and then animals were killed using sodium pentobarbital overdose (180 mg/kg), and femurs were excised.

Calcium absorption.

Calcium absorption was measured using an intestinal ligated loop technique as described previously by Weiler et al. (23). The intestinal ligated loop technique is designed to assess calcium absorption by testing the disappearance of a radiolabeled Ca isotope across the layers of the intestinal tract. This technique was conducted on the morning of the d 16 of the study. Food was removed 12 h before the procedure. Piglets were anesthetized and a 5- to 10-cm section of duodenum distal to the ligament of Treitz was ligated at both ends using suture, and the duodenal section was then filled with radioisotope buffer solution (pH of 7.4, mOsmol of 315) containing 45Ca (18 MBq/L; Amersham Ltd.), Poly R-478 (100 mg/L; Sigma Chemical Co.–Aldrich Ltd.), mannitol (20 mmol/L; Fisher Scientific), and CaCl2 (2 mmol/L; Fisher Scientific). Samples were taken every 5 min for a 30-min period. Samples were analyzed for 45Ca by scintillation counting (Model LS 6000TA; Beckman Instruments Inc.). Poly R-478 was measured by UV spectrometry as described (24). Calcium absorption was calculated according to the method of Ghishan et al. (25).

Bone and urine mineral analysis.

Once removed, femurs were cleaned and measurements were taken. Measurements included wet weight (g) and length (mm). Length was measured using calipers to the nearest 0.01 mm. Whole femurs were defatted in acetone (Optima, Fisher Scientific, Allentown, PA) over 5 d and then ashed in a muffle furnace at 600°C for 6 h. The femurs were then ground and weighed to determine ash weight. Femur ash samples (200 mg) were digested in 1.5 mL concentrated nitric acid and 0.25 mL of pooled 24-h urine samples were digested in 0.5 mL concentrated nitric acid for measurement of minerals. Both urine and femur digests were left at room temperature for 48 h and then diluted to obtain a final concentration of 5% (vol/vol) nitric acid in deionized water. Femur and urinary calcium, phosphorus, magnesium, and zinc determined by inductively coupled plasma optical emission spectroscopy (Varian Liberty 200, Varian Canada, Mississauga, ON, Canada). The coefficient of variance for each mineral was <10%. Minerals were expressed per gram of bone before ashing representing bone mineral relative to bone matrix and also per gram of bone ash to reflect the relative proportions of each mineral within bone.

Statistical analysis.

Main effects were detected using two-way ANOVA (diet and litter) using SAS statistical software (SAS software release 8.2; SAS Institute, Cary, NC). Litter was included in the statistical analysis as a main effect because previous research demonstrated that litter has a strong effect on bone parameters (21). Post hoc analysis was conducted using Duncan's Multiple Range Test when main effects were detected by the two-way ANOVA. Body weight was explored as a covariate in the analysis, but yielded the same interpretation as the two-way ANOVA; results of the unadjusted two-way ANOVA are presented for clarity of presentation. A P-value of less than 0.05 was accepted as significant. Data are expressed as mean ± SEM (SEM).

RESULTS

At the start of the study, there were no differences among the groups in weight (range 2.00–2.14 kg). By the end of the feeding trial, weight remained consistent among the groups and ranged on average from 5.48 to 5.63 kg among the dietary groups. Accordingly, there were also no differences among groups in length, ranging from 53.7 to 54.2 cm, at the end of the study. There were no differences among groups in feed intake because the piglets consumed all the formula they were given at each feed (350 mL/kg/d).

There were no dietary effects on calcium absorption corrected to intestinal weight (control 10.5 ± 1.1; 0.5:0.1 g, AA:DHA 9.0 ± 0.5; 1.0:0.2 g, AA:DHA 9.4 ± 1.1; 2.0:0.4 g, AA:DHA 9.2 ± 0.7 mmol/30 min/g intestine). The average calcium absorption for all animals was 84%. No dietary effects were observed on femur wet weight, length, or ash weight (Table 1). There was a significant main effect of diet on femur calcium when expressed as mg/g of femur ash (p = 0.0498) (Table 2). The 1.0:0.2 g AA:DHA group had significantly higher femur calcium than control and the 2.0:0.4 g AA:DHA group, while the 0.5:0.1 g AA:DHA had intermediate values. There were no diet effects on femur calcium when expressed per whole femur (g) (Table 2) or per length (control 25.3 ± 1.7; 0.5:0.1 g, AA:DHA 25.8 ± 1.1; 1.0:0.2 g, AA:DHA 25.1 ± 1.3; 2.0:0.4 g, AA:DHA 24.1 ± 1.0). There was a significant main effect of diet on femur zinc when expressed as μg/g of femur ash (p = 0.0227) (Table 2). The 1.0:0.2 g AA:DHA group had significantly lower values compared with control, while the 0.5:0.1 g AA:DHA and 2.0:0.4 g AA:DHA groups had intermediate values. There was no effect of diet on femur zinc when expressed per whole femur (μg/g) (Table 2) or per femur length (control 9.5 ± 0.8; 0.5:0.1 g, AA:DHA 9.3 ± 0.5; 1.0:0.2 g, AA:DHA 8.5 ± 0.5; 2.0:0.4 g, AA:DHA 8.5 ± 0.4 μg/mm). There were no main effects of diet on femur phosphorus expressed per g, whole femur (Table 2) or per femur length (control 15.8 ± 1.1; 0.5:0.1 g AA:DHA 16.0 ± 0.7; 1.0:0.2 g AA:DHA 15.6 ± 0.8; 2.0:0.4 g AA:DHA 15.1 ± 0.6 mg/mm). There were no main effects of diet on femur magnesium expressed per g, whole femur (Table 2) or per femur length (control 273 ± 18; 0.5:0.1 g, AA:DHA 273 ± 18; 1.0:0.2 g, AA:DHA 262 ± 12; 2.0:0.4 g, AA:DHA 262 ± 13 mg/mm).

Table 1 Femur wet weight, length, and ash weight of piglets fed formula with AA:DHA for 15 d versus control
Full size table
Table 2 Femur mineral content of piglets fed formula with AA:DHA for 15 d versus control
Full size table

Diet did not affect urinary calcium, phosphorus, magnesium, or zinc when expressed as over a 24-h period (mmol) or when corrected to the weight of the animal (mmol/kg) (Table 3).

Table 3 Urinary mineral excretion of piglets fed formula supplemented with AA:DHA for 15 d versus control
Full size table

DISCUSSION

This is the first study to report the effects of dietary AA and DHA on the composition of bone mineral including calcium, phosphorus, magnesium, and zinc during infancy. Dietary AA:DHA did not alter femur growth or mineral content when added in amounts (0.5:0.1 g/100 g of fat) similar to that observed in human milk of Canadians (26). Higher amounts of AA:DHA at 1.0:0.2 g/100 g of fat, but not 2.0:0.4 g/100 g of fat elevated calcium and reduced zinc, but had no effect on magnesium or phosphorus in the femur. In agreement with previous piglet studies on calcium and phosphorus content in tibial diaphysis sections, without periosteum (6,7), whole femur calcium and phosphorus contents were unchanged with dietary AA:DHA as 0.5:0.1 g/100 g of fat. The alterations in femur calcium and zinc in the piglets fed 1.0:0.2 g/100 g of fat as AA:DHA may be related to reduced bone resorption (8) and possibly higher proportions of less mature woven bone that are known to mineralize more quickly (27). Regardless, dietary AA:DHA did not increase femur total ash weight or total calcium, in line with no change to bone mineral content or bone mineral density, as previously reported (8).

The effects of dietary PUFAs and LCPUFAs on femur mineral content and growth have been studied in rats. In growing rats, dietary GLA plus EPA/DHA in a total n-6:n-3 ratio of 3:1 (GLA 7.4 g/100 g of fat and EPA/DHA 2.5 g/100 g of fat) and 1:1 (GLA 6.6 g/100 of fat and EPA/DHA 5.0 g/100 g of fat) resulted in higher amounts of calcium in bone compared with control, but a ratio of 1:3 (GLA 5.2 g/100 g of fat and EPA/DHA 12.4 g/100 g of fat) had no effect (2). GLA and EPA/DHA did not alter femoral ash weight or length (2). Other studies in which only n-3 LCPUFAs were supplemented through use of fish oil (28,29) did not yield any change to femur weight, length, ash weight, calcium, magnesium, or phosphorus in growing rats (28). Thus, it appears that both n-6 and n-3 LCPUFAs are required to elevate femur calcium content. The rats used by Claassen et al. (2) ate 20–22 g food/d and had a final weight of 350 g. Given the proportion of LCPUFAs in the diets, the rats consumed 0.3g GLA/kg, 0.07 g EPA/kg, and 0.02 g DHA/kg of body weight in the 3:1 n-6/n-3 group. In our piglet study, the amounts of LCPUFAs for the group receiving 1.0:0.2 g/100 g of fat as AA:DHA were 0.22 g AA/kg and 0.04 g DHA/kg of body weight. These studies suggest that dietary n-6 (GLA or AA) and n-3 (EPA and/or DHA) PUFAs that bypass the rate-limiting step of Δ-6 desaturase are needed in a ratio from 3:1 to 5:1 to observe elevations in bone calcium. None of these studies investigated the effects of dietary n-6 and n-3 PUFAs and LCPUFAs on bone zinc content.

Dietary AA:DHA at 1.0:0.2 g/100 g of fat reduced femur zinc. During zinc deficiency, the amount of zinc in the bone decreases because bone is an endogenous source of zinc when the dietary supply is low (30). Dietary zinc intake was not different among the groups of piglets, which suggests that zinc metabolism is altered in response to LCPUFA. Retention of zinc was not affected by diet as indicated by urinary zinc values. A possible explanation for lower femur zinc is that intestinal absorption is reduced or bone uptake is reduced because it is being accumulated or used in other tissues. In rats, bone zinc content stabilizes after weanling (31) and with reductions in the zinc content of bone ash, there are significant reductions in bone mineral mass (32). In a normal adult, approximately 29% of body zinc is found in bone (17) and those who suffer from bone fracture have lower bone zinc content (33). Thus, the longer term implications of lower bone zinc herein require investigation.

An important objective of this study was to confirm that dietary AA:DHA does not alter calcium absorption during rapid growth (6). Dietary AA:DHA at 0.5:0.1 g/100 g of fat, and up to 4-fold higher amounts, in formula containing a total n-6:n-3 ratio of 5:1 does not alter calcium absorption. In growing rats, dietary GLA and EPA plus DHA with a total n-6:n-3 ratio of 3:1 and 1:1 increased calcium absorption by 45% and 21.4%, respectively using mass balance, compared with control rats fed LA and ALA (total n-6:n-3 ratio of 3:1) (2). The lack of increase in calcium absorption in piglets may be the result of already optimal calcium transport (84% absorption) compared with the older growing rats; the different animal models (rat versus piglet); and the in situ ligated loop did not represent whole intestinal absorption. In piglets, calcium absorption was measured at 84%, whereas in rats, although the percentage of calcium absorbed was not stated, calcium absorption was increased by up to 45% (2). It is most likely that the differences between these two studies were due to the animal's developmental ages rather than the specific amounts of PUFAs because in human preterm infants dietary AA:DHA also did not affect mineral balance (4).

The final objective of this study was to determine whether dietary AA:DHA alters urinary mineral excretion; no effects of AA:DHA were observed. In growing rats, dietary GLA plus EPA/DHA supplemented to achieve a total n-6:n-3 ratio of 1:3 (GLA 5.2 g/100 g of fat and EPA/DHA 12.4 g/100 g of fat) reduced urinary calcium excretion, whereas a total n-6:n-3 ratio of 3:1 increased urinary calcium (2), suggesting that the dietary n-6:n-3 ratio is related to calcium excretion rather than specific amounts of PUFAs. The current study supports this thesis because the total n-6:n-3 PUFA ratio was constant. However, diets made with menhaden oil versus those made with soybean oil had no effect on urinary calcium in healthy (28,29) or diabetic (29) rats. A menhaden oil diet has a much lower total n-6:n-3 ratio compared with a soybean oil diet. AA status resulting from a high consumption of n-6 PUFA increases calcium excretion (34). Whether it is the total n-6:n-3 ratio or the amount of specific PUFAs/LCPUFAs that influence calcium excretion requires further investigation.

In conclusion, higher whole body or lumbar spine BMD found in piglets supplemented with 0.5:0.1g/100 g of fat as AA:DHA was not explained by alterations in calcium absorption, bone mineral composition or mineral excretion. The specific minerals in the whole body and lumbar spine were not measured, therefore, alterations in mineral composition therein cannot be ruled out. Dietary AA:DHA as 1.0:0.2 g/100 g of fat increased femur calcium (per g of femoral ash) and reduced bone resorption (8). A longer term study may have shown improved bone mass through reduced bone resorption and higher bone calcium. In addition, mineral balance was not measured in this study and could potentially explain why dietary AA:DHA at 1.0:0.2 g/100 g of fat resulted in altered bone calcium and zinc content. Research is needed to determine the mechanism for increased femoral calcium and reduced femoral zinc content. Whether the decrease in femur zinc seen here is detrimental to bone is unknown and requires further investigation.