Bioavailability and safety of a high dose of docosahexaenoic acid triacylglycerol of algal origin in cystic fibrosis patients: a randomized, controlled study

Introduction 

Several reports on patients with cystic fibrosis (CF) have described blood fatty acid imbalances or deficiencies characterized by low levels of the ω-6 and ω-3 essential fatty acids linoleic acid (18:2ω-6) and α-linolenic acid (18:3ω-3) and their corresponding long-chain (LC) derivatives, arachidonic acid (ARA; 20:4ω-6) and docosahexaenoic acid (DHA, 22:6ω-3) [1], [2], [3], [4], [5], [6], [7], [8]. Recently, Freedman et al. [35] reported low DHA levels and high ARA levels in nasal submucosa and low DHA levels in rectal biopsy tissue from patients who had CF compared with healthy controls, indicating that fatty acid imbalances are found not only in blood but also in tissues expressing the cystic fibrosis transmembrane regulator (CFTR). Although malabsorption secondary to pancreatic insufficiency in CF may contribute to this imbalance, Strandvik et al. [9] suggested that faulty metabolism of these polyunsaturated fatty acids (PUFAs) may also be a factor, possibly leading to higher dietary requirements for specific fatty acids. Kang et al. [10] reported that a chemical inhibitor of chloride channel conductance resulted in decreased accretion of PUFAs in cultured airway epithelial cells. This suggests that ion imbalances in CFTR+ tissues may affect PUFA balances in membranes.

Fatty acid imbalances may contribute to lung and gastrointestinal symptoms associated with CF, although the extent of this association is unclear. Animal studies have shown that essential fatty acid deficiency results in pulmonary disease similar to CF [11], [12]. Others have proposed that high ω-6/ω-3 ratios may contribute to the chronic inflammation observed in CF [13], [14]. The cftr−/− transgenic mouse model expresses abnormally low DHA and high ARA levels in the lung, ileum, and pancreas [13], although this was not subsequently confirmed [15]. The CFTRtm1HGU/tm1HGU mouse has low levels of ARA and normal levels of DHA in all phospholipid classes of pancreas and lung [16]. Freedman et al. [13] reported that correcting the high ARA/DHA ratios in the cftr−/− mouse using large doses of dietary DHA reversed pancreatic histopathology [13]. Supplementation with another LC ω-3 fatty acid found in fish oil, eicosapentaenoic acid (EPA; 20:5ω-3), which is thought to have antiinflammatory properties, was not effective in this mouse model. In a number of small clinical studies, patients with CF have been supplemented with fish oil at levels up to 5.4 g of LC ω-3 per day. Effects on inflammatory markers and lung function in these studies have been variable and inconclusive [4], [17], [18], [19], [20], [21], [22], [23].

To investigate the role of DHA and fatty acid imbalances in CF, we conducted a double-blind, placebo-controlled, clinical study in patients with CF who had mild to moderate lung disease. The objective of the study was to determine whether an algal triacylglycerol source of DHA could be safely and effectively absorbed and to assess accretion in CFTR+ tissue during 6 mo of dietary supplementation. In addition, patients were evaluated for effects on lung function.

Results 

Subjects 

Nineteen of the 20 subjects completed the 6-mo study. One subject in the DHA treatment group withdrew consent after randomization but before supplementation. This subject was excluded from all analyses. Blood was not obtained from one subject in the placebo group at the 1-mo visit. Compliance with taking the study supplement, as assessed by capsule count, was greater than 90% in all but one subject (later identified as being in the placebo group). Dietary DHA intakes, as assessed by food frequency questionnaire, were similar in both groups and were stable in each supplementation group throughout the study. Mean dietary DHA intakes (excluding supplements) at baseline and at 3 and 6 mo were 51 ± 13, 56 ± 25, and 77 ± 22 mg/d in the algal DHA group and 49 ± 11, 60 ± 22, and 66 ± 24 mg/d in the placebo group.

The baseline characteristics of the study subjects were similar between groups, with the exception of height-for-age z scores, which were lower in the DHA supplementation group (Table 2). The distribution of genotypes was similar between groups, with all but one subject in the placebo group being homo- or heterozygous for the ΔF508 allele, the most common of the CF genotypes associated with clinically more significant disease.

Table 2. Baseline demographics of study cohort⁎

	Characteristic	Placebo (n = 10)	Algal DHA (n = 9)
	Age (y)	13.3±1.1	11.7±0.81
	Male/female	8/2	6/3
	ΔF508 genotype (homo-/heterozygous/other)	3/6/1	3/6/0
	Weight (kg)	45.1±5.7	32.6±2.7
	Weight-for-age (z score)	−0.187±0.37	−0.996±0.19
	Height (cm)	151±5.7	138±3.2
	Height-for-age (z score)	−0.418±0.22a	−1.106±0.18b
	%Ideal body weight	104.9±6.5	100.1±5.4
	BMI	19.03±1.4	16.78±0.82
	Dietary DHA intake (mg/d)	49±11	51±13
	Vitamin E supplement (IU/d)	260±32	283±24

BMI, body mass index; DHA, docosahexaenoic acid

⁎ Mean ± standard error of the mean; superscript letters indicate that means are significantly different from each other (Student’s t test, P ≤ 0.05).

Fatty acid absorption and accretion 

Docosahexaenoic acid absorption from the algal DHA triacylglycerol supplements was monitored by following DHA levels in plasma phospholipids. DHA levels in the placebo group remained stable throughout the study, whereas all subjects in the algal DHA treatment group had rapid increases in their plasma phospholipid DHA levels, demonstrating a uniform level of absorption in this group. Mean plasma phospholipid DHA levels in the DHA group increased five-fold in a time-dependent, saturable manner (Table 3). As expected, as the DHA levels increased, ARA levels decreased by nearly 40%, resulting in an 88% decrease in the plasma phospholipid ARA/DHA ratio. There were also decreases in other ω-6 fatty acids, including dihomo-γ-linolenic acid (20:3ω-6), docosapentaenoic acid of the ω-6 series (DPAω-6; 22:5ω-6), 22:4ω-6, and total ω-6 PUFA. EPA (20:5ω-6) and total ω-3 PUFAs increased, and DPA of the ω-3 series (DPAω-3; 22:5ω-3) decreased, which is a typical response after DHASCO supplementation [29], [30], [31]. Plasma phospholipids levels of saturated and monounsaturated fatty acids did not change after supplementation.

Table 3. Plasma phospholipid fatty acid levels⁎

	Fatty acids	Placebo (n = 10)				Algal DHA (n = 9)
		Baseline	1 mo	3 mo	6 mo	Baseline	1 mo	3 mo	6 mo
	Saturated†	41.06±0.47	41.27±0.37	40.39±0.34	41.10±0.35	39.93±1.17	40.39±0.97	40.93±0.93	40.59±0.71
	Monounsaturated‡	17.23±0.63	15.82±0.49	16.98±0.33	17.07±0.67	15.75±0.56	14.86±0.20	14.55±0.20	14.83±0.49
	18:2ω-6	19.96±0.45	21.78±0.62	20.76±0.31	21.10±0.38	20.48±1.12	19.10±0.73	19.04±0.73	19.33±0.84
	18:3ω-6	0.20±0.02	0.20±0.03	0.18±0.03b	0.19±0.02	0.27±0.03	0.15±0.02	0.17±0.02	0.15±0.01
	20:2ω-6	0.36±0.04	0.33±0.04	0.33±0.04	0.35±0.04	0.33±0.04	0.32±0.04	0.33±0.04	0.32±0.04
	20:3ω-6	4.56±0.15a	4.39±0.26a	4.37±0.23a	4.44±0.16a	4.43±0.21a	3.22±0.18b	3.17±0.18b	2.94±0.20b
	20:4ω-6	10.86±0.42a,b	10.74±0.32a,b	10.92±0.23a	10.05±0.29b	12.15±0.72a,c	9.01±0.51b,d	8.05±0.50d	7.58±0.40d
	22:4ω-6	0.68±0.03a	0.68±0.05a	0.63±0.04a	0.65±0.02a	0.69±0.07a	0.29±0.02b	0.25±0.02b	0.23±0.02b
	22:5ω-6	0.68±0.03a	0.62±0.05a	0.60±0.04a	0.61±0.03a	0.66±0.06a	0.24±0.02b	0.18±0.02b	0.15±0.01b
	Total ω-6	37.32±0.55a	38.78±0.39a	37.84±0.38a	37.43±0.41a	39.05±1.24a	32.37±1.17b	31.21±1.13b	30.73±1.05b
	18:3ω-3	0.18±0.02a	0.21±0.01a,b	0.19±0.01a	0.24±0.02a,b	0.30±0.05b	0.23±0.02a,b	0.25±0.02a,b	0.21±0.02a
	18:4ω-3	0.10±0.02	0.12±0.03	0.12±0.03	0.11±0.02	0.13±0.04	0.16±0.04	0.13±0.04	0.13±0.04
	20:5ω-3	0.62±0.08a	0.58±0.07a	0.72±0.10a	0.70±0.07a	1.09±0.22a,b	1.12±0.12a,b	1.44±0.11b	1.43±0.18b
	22:5ω-3	1.02±0.06a	0.98±0.08a	1.10±0.07a	1.00±0.06a	1.11±0.10a	0.47±0.07b	0.53±0.07b	0.46±0.03b
	22:6ω-3	2.10±0.24a	1.93±0.18a	2.43±0.38a	2.05±0.29a	2.29±0.26a	10.21±0.61b	10.78±0.61b	11.46±0.74b
	Total ω-3	3.72±0.22a	3.58±0.17a	4.10±0.42a	3.76±0.22a	4.51±0.41a	12.02±0.66b	12.86±0.66b	13.46±0.90b
	ARA/DHA	5.65±0.50a	5.79±0.47a	5.28±0.60a	5.50±0.49a	5.70±0.56a	0.92±0.18b	0.78±0.08b	0.69±0.07b
	PUFA ω-6/ω-3	10.30±0.52a	11.01±0.48a	9.88±0.72a	10.41±0.63a	9.08±0.62a	2.78±0.21b	2.50±0.18b	2.40±0.24b

ARA, arachidonic acid; DHA, docosahexaenoic acid; PUFA, polyunsaturated fatty acid

⁎ Mean (g/100 g total fatty acid) ± standard error of the mean. Repeated measures analysis of variance was used to model main effects of treatment and time allowing for interactions between them. The level of significance was set at P < 0.05 with Bonferroni’s adjustment for multiple comparisons (P < 0.003). Values in the same row with different letter superscripts are significantly different (Tukey’s pairwise comparisons). † Includes 14:0, 16:0, 17:0, 18:0, 20:0, 22:0, and 24:0. ‡ Includes 14:1, 16:1, 18:1, 20:1, 22:1, and 24:1.

Similarly, erythrocytic DHA levels were stable in the placebo group throughout the treatment period, whereas DHA levels in the algal DHA supplementation group increased four-fold in a time-dependent manner, albeit with characteristically slower kinetics than in the plasma. Erythrocytic ARA levels decreased 29% and the ARA:DHA ratio decreased 80% (Table 4). These changes were accompanied by concomitant decreases in the ω-6 fatty acids dihomo-γ-linolenic acid, ARA, DPAω-6 and total ω-6 PUFAs, and DPAω-3. Similar to plasma, EPA and total ω-3 PUFA levels in erythrocytes increased after DHA supplementation.

Table 4. Erythrocyte fatty acid levels⁎

	Fatty acids	Placebo (n = 10)				Algal DHA (n = 9)
		Baseline	1 mo	3 mo	6 mo	Baseline	1 mo	3 mo	6 mo
	Saturated†	38.48±0.22	38.27±0.21	38.55±0.23	39.86±0.59	38.71±0.32	39.19±0.30	39.34±0.35	40.21±0.26
	Monounsaturated‡	21.32±0.41	21.47±0.40	21.11±0.37	21.4±0.44	20.98±0.38	20.55±0.19	19.95±0.22	20.53±0.25
	18:2ω-6	11.33±0.41	11.51±0.20	11.89±0.21	11.75±0.25	10.74±0.39	10.02±0.32	10.60±0.32	10.44±0.35
	18:3ω-6	0.13±0.01	0.10±0.02	0.10±0.01	0.12±0.01	0.13±0.01	0.06±0.01	0.07±0.02	0.09±0.01
	20:2ω-6	0.26±0.02	0.30±0.01	0.24±0.02	0.29±0.01	0.28±0.02	0.23±0.01	0.24±0.01	0.26±0.01
	20:3ω-6	2.33±0.15a	2.38±0.17a	2.26±0.16a	2.28±0.12a	2.02±0.09a,b	1.58±0.19b	1.59±0.01b	1.50±0.04b
	20:4ω-6	14.90±0.24a	14.78±0.27a	14.47±0.21a	13.34±0.91a,b	15.41±0.42a	14.32±0.41a	11.87±0.36b,c	10.98±0.51c
	22:4ω-6	4.39±0.17a	4.40±0.17a	4.42±0.18a	4.22±0.16a	4.42±0.21a	3.79±0.19a	2.67±0.14b	2.04±0.26a,b
	22:5ω-6	1.04±0.05a	1.04±0.04a	0.98±0.04a	0.95±0.5a	1.01±0.07a	0.78±0.06b,a	0.41±0.02b	0.26±0.07b
	Total ω-6	34.43±0.32a	34.61±0.38a	34.45±0.35a	33.04±0.81a,b	34.10±0.44a	30.88±0.39b	27.49±0.48c	25.68±1.02c
	18:3ω-3	0.16±0.03	0.18±0.01	0.13±0.01	0.14±0.01	0.13±0.01	0.13±0.02	0.13±0.01	0.13±0.01
	18:4ω-3	0.09±0.01	0.06±0.01	0.12±0.01	0.14±0.01	0.12±0.02	0.05±0.01	0.14±0.02	0.15±0.01
	20:3ω-3	0.07±0.04	0.06±0.01	0.05±0.02	2.08±0.01	0.04±0.01	0.03±0.02	0.03±0.02	0.11±0.02
	20:5ω-3	0.47±0.04a	0.43±0.03a	0.52±0.05a	0.48±0.04a	0.61±0.05a	0.64±0.03a	0.90±0.06b	0.99±0.11b
	22:5ω-3	2.27±0.12a	2.21±0.08a	2.35±0.10a	2.25±0.12a	2.41±0.13a	1.95±0.11a	1.29±0.07b	1.08±0.14b
	22:6ω-3	2.58±0.21a	2.62±0.23a	2.66±0.24a	2.51±0.22a	2.76±0.23a	6.53±0.21b	10.66±0.25c	11.08±0.75c
	Total ω-3	7.76±0.18a	7.75±0.18a	7.89±0.22a	7.57±0.22a	8.07±0.23a	11.19±0.27b	14.53±0.26c	14.49±0.75c
	ARA/DHA	6.09±0.45a	5.95±0.48a	5.83±0.49a	5.73±0.64a	5.88±0.49a	2.22±0.12b	1.12±0.05c	1.16±0.27c
	PUFA n-6/n-3	4.46±0.11a	4.49±0.11a	4.40±0.12a	4.41±0.19a	4.25±0.11a	2.77±0.09b	1.90±0.06c	1.86±0.22c

ARA, arachidonic acid; DHA, docosahexaenoic acid; PUFA, polyunsaturated fatty acid

⁎ Mean (g/100 g total fatty acid) ± standard error of the mean. Repeated measures analysis of variance was used to model main effects of treatment and time allowing for interactions between them. The level of significance was set at P < 0.05 with Bonferroni’s adjustment for multiple comparisons (P < 0.003). Values in the same row with different letter superscripts are significantly different (Tukey’s pairwise comparisons). † Includes 14:0, 16:0, 17:0, 18:0, 20:0, 22:0, and 24:0. ‡ Includes 14:1, 16:1, 18:1, 20:1, 22:1, and 24:1.

DHA uptake into CFTR-expressing tissue was assessed by measuring DHA levels in rectal samples obtained at baseline and after 3 mo of supplementation. Rectal DHA levels did not change in the placebo group, whereas DHA levels in the algal DHA treatment group increased five-fold after 3 mo of supplementation (Table 5), demonstrating uptake and accretion of this dietary fatty acid into this CFTR+ tissue. EPA levels were also higher, as were total ω-3 PUFAs in the rectal tissue. Although none of the ω-6 fatty acids, including ARA, decreased significantly in rectal tissue, the ARA/DHA and total ω-6/ω-3 LC-PUFA ratios decreased by 84% and 54%, respectively. There were significant correlations between DHA levels in rectal tissue and plasma phospholipids (r2 = 0.86, P < 0.001) and in rectal tissue and erythrocytes (r2 = 0.84, P < 0.001).

Table 5. Fatty acid levels in rectal tissue⁎

	Fatty acids	Placebo (n = 10)		Algal DHA (n = 9)
		Baseline	3 mo	Baseline	3 mo
	Saturated†	34.20±1.69	38.68±3.15	38.90±1.91	35.83±1.20
	Monounsaturated‡	40.66±2.69	36.42±3.00	33.53±1.20	33.17±1.34
	18:2ω-6	9.53±1.15	10.04±1.03	10.13±0.55	11.64±0.47
	18:3ω-6	0.52±0.10	0.34±0.08	0.31±0.05	0.27±0.04
	20:2ω-6	0.68±0.15	0.55±0.10	0.55±0.09	0.60±0.10
	20:3ω-6	2.19±0.33	1.59±0.31	1.65±0.45	1.79±0.23
	20:4ω-6§	5.86±0.47	6.66±1.13	8.27±0.76	7.11±0.38
	22:2ω-6	0.67±0.22	0.24±0.05	0.36±0.10	0.26±0.08
	22:4ω-6	0.95±0.07	1.12±0.19	1.20±0.14	0.56±0.04
	22:5ω-6	0.31±0.05	0.34±0.08	0.35±0.07	0.13±0.01
	Total ω-6	20.72±1.35	20.88±2.34	22.81±1.22	22.38±0.69
	18:3ω-3	0.97±0.24	0.75±0.17	0.89±0.27	0.54±0.10
	18:4ω-3	0.23±0.07	0.28±0.05	0.43±0.09	0.59±0.16
	20:3ω-3	1.27±0.65	1.00±0.40	1.28±0.26	1.04±0.30
	20:5ω-3	0.29±0.04a	0.29±0.05a	0.42±0.05a	1.12±0.09b
	22:5ω-3	0.55±0.06	0.62±0.12	0.76±0.11	0.52±0.04
	22:6ω-3∥	0.97±0.19a	0.70±0.16a	0.85±0.11a	4.36±0.31b
	Total ω-3	4.70±0.84a	4.16±0.67a	5.13±0.28a	8.29±0.49b
	ARA/DHA#	7.42±1.17a,b	11.46±2.95a	10.61±1.36a	1.69±0.14b
	PUFA ω-6/ω-3	5.19±0.72a	5.82±0.67a	4.59±0.41a,b	2.79±0.20b

ARA, arachidonic acid; DHA, docosahexaenoic acid; PUFA, polyunsaturated fatty acid

⁎ Mean (g/100 g total fatty acid) ± standard error of the mean. Means were analyzed by analysis of variance with the level of significance set at P < 0.05 with Bonferroni’s adjustment for multiple comparisons (P < 0.003). Values in the same row with different letter superscripts are significantly different (Tukey’s pairwise comparisons). † Includes 14:0, 16:0, 17:0, 18:0, 20:0, 22:0, and 24:0. ‡ Includes 14:1, 16:1, 18:1, 20:1, 22:1, and 24:1. § Baseline ARA level is equivalent to 6.5 mol % of total fatty acids. ∥ Baseline DHA level is equivalent to 0.86 mol% of total fatty acids. # Baseline ARA/DHA ratio is equivalent to 9.12 when compared on a mole percent basis.

Safety and tolerability 

Overall safety and adverse events were assessed by interviews, physical examinations, liver function tests, and measurements of antioxidant levels at regular intervals throughout the study. The supplements were well tolerated. None of the subjects required increased enzyme dosage to accommodate the additional dietary fat contributed by the supplements. Gastrointestinal complaints (stomach ache, cramps, and heartburn) in the DHA group (n = 1) did not differ from those in the placebo group (n = 5). Respiratory complaints and other non-supplement–related adverse experiences were similar between groups.

Liver assessments 

Liver involvement is common in CF, and to ensure that DHA did not contribute to or exacerbate liver symptoms, liver function tests were closely monitored throughout the study. None of the subjects showed evidence of liver damage during the study. There were no differences in mean serum liver enzymes (γ-glutamyltransferase, alanine aminotransferase, aspartate aminotransferase, or alkaline phosphatase) or albumin concentrations between treatment groups at any period in the study and no time-dependent changes in these enzymes within groups (data not shown). One subject in each group had an episode of increased alkaline phosphatase concentration. Both occurrences spontaneously corrected upon repeat testing 1 wk later without alteration of the study supplement, suggesting that the episodes were not related to the regimens. No subjects experienced an increase in liver size during the study.

Anthropometrics 

Both groups had below normal weight-for-age and height-for-age z scores and relatively low body mass indexes (<20 kg/m2; Table 2). Subjects in the DHA treatment group tended to be younger and, in general, smaller than those in the placebo group, with lower height-for-age z scores at baseline and at all subsequent time points throughout the study (data not shown). However, subjects in both groups maintained their height and weight (z scores) throughout the 6-mo treatment period, suggesting that the DHA supplementation did not adversely affect growth.

Lipid-soluble antioxidants 

A panel of major LSAs was monitored during the treatment because LC-PUFA might have increased oxidative stress in these patients. All study participants were taking supplemental vitamin E, and most subjects (15 of 19) were taking ADEK supplement (Scandipharm Birmingham, AL, USA; Table 2), a common nutritional supplement that contains β-carotene, retinyl palmitate, α-tocopherol, and other vitamins and minerals. Subjects in both groups had lower than normal concentrations of lutein, zeaxanthin, and α-carotene at baseline and throughout the study (Table 6). Concentrations of cryptoxanthin and lycopene were in the low normal range. β-Carotene, retinol, retinyl palmitate, α-tocopherol, and γ-tocopherol concentrations were in the normal range. There were no time or treatment effects of algal DHA on any of the LSA concentrations measured in this study, suggesting that DHA was not depleting these antioxidant protection systems. Plasma β-carotene (r2 = 0.39, P = 0.006) and α-tocopherol (r2 = 0.23, P = 0.038) concentrations were positively related and γ-tocopherol (r2 = 0.53, P < 0.001) concentrations were negatively related to percentage of predicted forced vital capacity at baseline. There were significant positive correlations (P < 0.005) between plasma phospholipid or erythrocytic DHA and plasma α-carotenoids (r2 = 0.37 and 0.56, respectively) at baseline.

Table 6. Plasma antioxidant concentrations in patients who had cystic fibrosis and were treated with placebo or algal DHA for 6 mo⁎

	Antioxidant	Normal range	Placebo (n = 10)				Algal DHA (n = 9)
			Baseline	1 mo	3 mo	6 mo	Baseline	1 mo	3 mo	6 mo
	Lutein (ng/mL)	50–570	33±7	35±8	37±7	35±7	31±2	34±3	35±3	32±2
	Zeaxanthin (ng/mL)	20–130	15±4	16±6	18±4	16±6	14±2	15±2	16±3	8±1
	Cryptoxanthin (ng/mL)	7–180	16±9	22±11	18±8	15±9	13±3	10±3	12±4	10±2
	Lycopene (ng/mL)	10–330	32±14	48±33	48±21	54±27	26±8	28±10	28±7	21±7
	α-Carotene (ng/mL)	10–370	10±6	11±7	9±4	11±6	9±3	8±5	10±3	9±3
	β-Carotene (ng/mL)	70–680	110±34	140±40	120±42	180±58	120±23	130±28	140±22	200±55
	Retinol (ng/mL)	350–1250	370±20	370±33	360±47	400±24	320±18	370±38	430±52	370±33
	Retinyl-palmitate (ng/mL)	10–170	27±4	34±4	29±5	20±4	38±8	32±4	27±6	18±4
	α-Tocopherol (μg/mL)	6.5–17.2	9.2±1.10	9.6±1.10	9.4±0.73	9.4±1.30	11.0±0.93	11.2±1.20	11.2±0.68	10.7±0.78
	γ-Tocopherol (μg/mL)	0.7–4.6	1.2±0.18	1.3±0.13	1.2±0.21	1.1±0.12	0.9±0.23	1.1±0.1	0.9±0.22	0.8±0.13

DHA, docosahexaenoic acid

⁎ Means ± standard error of the mean.

Lung function 

The study subjects represented a relatively healthy cohort of patients with CF, with mean forced expiratory volumes in 1 s greater than 70% predicted. As expected in CF, lung function tended to decrease slightly during the study period in both groups, but there were no significant time or treatment effects (Table 7). DHA supplementation had no measurable effect on lung function followed through 6 mo. We found no correlation between plasma phospholipid DHA, ARA, EPA, ω-6 PUFA, ω-3 PUFA, ω-6/ω-3 ratio, or ARA/DHA ratios and lung function at any time point.

Table 7. Lung function (percentage predicted) in subjects who had cystic fibrosis and were treated with placebo or algal DHA for 6 mo⁎

	Measurement	Placebo (n = 10)			Algal DHA (n = 9)
		Baseline	3 mo	6 mo	Baseline	3 mo	6 mo
	Forced vital capacity	86±4.6	85±5.3	83±4.1	90±6.2	88±4.4	84±4.9
	Forced expiratory volume in 1 s	74±5.8	73±7.4	72±5.7	83±7.2	79±5.8	77±6.5
	Forced expiratory flow at 25–75%	62±9.3	66±10	62±8.3	79±11	71±8.5	74±11

DHA, docosahexaenoic acid

⁎ Means ± standard error of the mean. There were no treatment effects or time by treatment interactions as assessed by repeated measure analysis of variance at a significance level of P < 0.05.

Discussion 

This study demonstrates that an algal triacylglycerol source of DHA can be effectively absorbed by patients with CF after dietary supplementation with 50 mg of DHA per kilogram of body weight per day over a 6-mo period. Plasma DHA levels increased rapidly, reflecting the normally rapid turnover of phospholipid fatty acids in plasma, and was compatible with normal absorption kinetics [30]. The new equilibrium level of plasma phospholipid DHA was five-fold higher than in the placebo group and, at 11.5% of total fatty acids, is among the highest plasma DHA levels reported in humans, exceeding even those reported in healthy adults receiving a 50% larger dose of DHA [32]. Because plasma phospholipids are the main delivery form of fatty acids to tissues, such supplementation would be expected to increase tissue DHA levels. Erythrocytic and rectal DHA levels increased four- to five-fold after DHA supplementation. The magnitude of the changes in rectal fatty acid levels suggests that altered CFTR function does not prevent accretion of DHA into CFTR+ tissue, although without comparison data in normal individuals, accretion efficiency could not be assessed. Nonetheless, strong linear correlations between DHA levels in the rectum and in plasma and erythrocytes suggest predictable DHA uptake into CF rectal tissue.

Although several studies have suggested that blood levels of DHA and ARA may be low in patients with CF compared with normal individuals [3], [4], [5], [6], [7], [33], [34], comparison of the patients with CF in this study with a similarly aged cohort of healthy individuals [31] (Hoffman, unpublished data) suggests that DHA, but not ARA, levels were depressed in the plasma of these subjects. Freedman et al. [35] reported high ARA and low DHA levels in CFTR+ tissue in patients who had CF with pancreatic insufficiency. The mean baseline DHA level and the ARA/DHA ratio in rectal tissue of patients in this study fell between those reported for healthy controls and individuals with CF and may represent slight deviations from normal. The mean baseline rectal ARA level in this study population appeared to be lower than that in normal individuals and those with CF [35]. Although large-dose DHA supplementation increased rectal tissue DHA and decreased the ARA/DHA ratio, tissue ARA levels were not suppressed.

Different investigators have suggested that the imbalance in the ω-6/ω-3 ratio, or more specifically the ARA/DHA ratio, may contribute to the inflammatory characteristic of CF lung disease [19], [36]. This algal DHA treatment increased DHA, resulting in dramatically lower ω-6/ω-3 ratios in blood and tissue. Keicker et al. [36] found decreased production of stimulant-induced 5-lipoxygenase and leukotriene B4 production by granulocytes isolated from patients with CF after in vitro treatment with fish oil containing ω-3 LC-PUFA, and in a clinical study of patients with CF, fish oil lowered leukotriene B4 levels in sputum [22]. Kelley et al. [37] reported decreased production of prostaglandin E2, leukotriene B4, interleukin-1, and tumor necrosis factor-α in granulocytes isolated from normal healthy adults treated with the same algal DHA triacylglycerol. In cftr−/− mice, Freedman et al. [38] found decreased production of inflammatory cytokines and decreased neutrophil infiltration in the lung after a Pseudomonas lipopolysaccharide challenge when DHA was administered, suggesting an antiinflammatory protective mechanism.

The patients in this study had pancreatic insufficiency and were being treated with pancreatic enzymes, including lipases, before entry into the study. Although it has been suggested that a triacylglycerol, which requires lipase activity before absorption, may not be highly bioavailable in CF [39], this study demonstrates the absorption of this algal triacylglycerol source of DHA, without the need to increase pancreatic enzyme doses. This relatively concentrated source of DHA (approximately 40% by weight) resulted in a total daily fat load of 125 mg/kg of body weight or approximately 4 to 5 g of fat per day, which is less than 10% of the typical fat intake in an American child’s diet [40]. The supplement consumption was also spread across meals, which likely enhanced absorption. The low incidence of gastrointestinal side effects further suggests that absorption was not an issue. As a result, tolerability and compliance were high among the study participants. In contrast, large doses of fish oil as a source of ω-3 fatty acids have been associated with limited compliance, gastrointestinal side effects, and increased enzyme requirements in patients with CF [21], [22], [41]. This study therefore demonstrates that a natural algal triacylglycerol is well tolerated and can effectively deliver a large dose of DHA to patients with CF and may be preferred for future evaluation of ω-3 PUFA supplementation. Alternate delivery forms, such as monoacylglycerol, ethyl ester, or synthetically structured triacylglycerol sources, may not be necessary for efficient absorption in CF and might be prohibitively costly due to extra processing steps.

Arachidonic acid is important in the growth of infants [42], but little is known about its role in the growth of children and adolescents. Blood ARA levels may be lower than normal in CF [34] and this DHA treatment further decreased blood but not tissue ARA levels. Anthropometrics were carefully monitored throughout this study due to concerns about potential growth suppression. There was no evidence to suggest that growth was compromised by DHA supplementation. Despite the differences in baseline weight, comparison between placebo and DHA supplementation using sex- and age-corrected z scores indicated that the DHA treatment group had stable growth and body mass index during the course of the study.

Patients with CF have compromised LSA status, presumably related to increased oxidative stress caused by inflammation, and low liver glutathione production [43]. Concerns over increased oxidative stress related to increased PUFA intake have been raised [44] because large-dose PUFA could theoretically activate a cycle of increased lipid peroxidation and stimulation of tissue damage [45]. An array of LSA was monitored during this study to determine whether DHA had any adverse effect on antioxidant status. At baseline, the patients in this study had normal plasma levels of α- and γ-tocopherol, retinol, retinyl palmitate, and β-carotene, probably attributable to the ADEK supplementation. In contrast, others have reported decreased levels of β-carotene and retinol in CF, although α-tocopherol levels are usually normal [46], [47], [48], [49]. Similar to other reports [47], [50], subjects were deficient in lutein, zeaxanthin, and α-carotene, which are not present in the ADEK supplement. In addition, lycopene and cryptoxanthin were in the low normal range. Based on this observation, supplementation with lutein and zeaxanthin may be indicated because these antioxidants, which are highly concentrated in the macula, are thought to be ophthalmologically protective and may have selective antioxidant activity in the retina [51]. There was no observable effect of the DHA supplementation on the concentration of any LSA measured in this study, suggesting that the increased DHA level did not cause excessive oxidative stress or depletion of LSA.

Liver disease, manifested by cholestasis, fibrosis, steatosis, and/or cirrhosis, can lead to morbidity and mortality in CF [52]. Liver function was carefully monitored in this study. DHA produced no untoward effects on liver size or biochemical parameters such as serum liver enzymes or albumin.

Lung function was also evaluated in this study population. The anticipated trend toward decreased lung function was observed in both groups and is consistent with normal progression of the disease. Despite large alterations in fatty acid levels, there was no identifiable effect of the DHA supplementation on forced expiratory volume or flow in these patients with CF. This study, however, was not powered to detect more subtle effects on lung function and its duration was not adequate to address whether or not DHA can slow or decrease the progression of lung disease in patients with CF. Larger studies of longer duration will be required to identify potential benefits of DHA supplementation on lung and gastrointestinal CF pathology and symptomatology.