A one-tailed Wilcoxon Signed Ranks test did not reveal a significant difference between Week 0 and Week 12 acne grade, Z (Week 12-Week 0) = −1.044, p > .05. A total of 8 (62%) participants improved at least one-point acne grade over the 12 weeks of supplementation. Acne grade worsened for 4 (31%) participants and did not change for 1 (7%) individual. Table shows the number (and percentage) of individuals classified in each acne grade before and after supplementation. A one-tailed t-test for dependent means did not reveal a significant difference in mean acne counts for Week 0 and Week 12, t(12) = 0.338, p > .05. Mean acne count in Week 0 was 30 (SE = 5.06) and 31 (SE = 5.67) in Week 12. As expected, the unbiased effect size was small, d' = 0.08 (SEd' = 0.26). One-way repeated measures analysis of variance did not reveal significant difference in mean Δa* or ΔL* across weeks F(2,24) < 1.0, Huynh-Feldt ε = 1.000, p > 0.05 (Figure ).
Changes in Δa* and ΔL* throughout 12 weeks of fish-oil supplementation.
Even though statistical significance was not obtained for acne severity and lesion counts, the results of this study should be interpreted with caution and are likely inconclusive, for we clearly observed that the change in overall acne ratings over the 12-week supplementation period was in the predicted direction. After 12 weeks of supplementation, 8 out of 13 participants received a lower acne grade (i.e., indicating improvement) than at baseline. Five of these 8 individuals improved one grade and the rest improved at least two grades. One participant’s acne grade shifted from the most severe (grade 8) to the least severe grade (grade 0). Prior to supplementation, nearly one-third of the participants (n = 4) had an acne grade of 8, which is the most severe grade based on the scale used. All of these individuals had a less severe acne rating, as indicated by a lower grade at the end of the study. Of the 5 individuals who did not show improvement, acne severity worsened for 4 individuals and did not change for one individual. Notably, of the 4 individuals whose acne worsened, 3 individuals had mild acne (grades 0 or 2) at the start of the study (Figure ). The individual whose acne did not change also exhibited mild acne (rating of 2) at the start of the study. On the other hand, 7 of the 8 individuals who entered the study with moderate to severe acne (grades 4, 6, and 8) saw improvement in acne severity by the end of the study. This may be the reason why significant results were not obtained in this study, at least for the acne severity ratings. The sample population included a relatively large proportion (38.5%) of individuals with mild acne. Since their acne grades were already near the low end of the scale at the start of the study, there was little room for obvious improvement and conversely greater opportunity for worsening. Similarly, a lack of significant difference in acne lesion counts between Weeks 0 and 12 may be partially due to a floor effect; that is, the inclusion of several individuals who had mild acne at the start of the study. These findings raise the question of whether the efficacy of fish oil supplementation is dependent on the initial acne severity. In other words, perhaps only individuals with moderate to severe acne can benefit from fish oil supplementation.
Difference in acne severity changes among individuals with mild vs. moderate-severe acne at baseline.
In addition, it should be noted that in counting the inflammatory lesions, we did not differentiate between the types of inflammatory lesions for our data analysis. In other words, we did not distinguish between papules, pustules, and cysts in the lesion count, but rather categorized them all as inflammatory lesions. This may be why we did not achieve high correlation between a* and actual inflammatory lesion counts, 0.42, 0.62, and 0.44 at baseline, week 6, and week 12, respectively. Thus, given the fact that acne is a dynamic disease, any improvement in the type of inflammatory lesion might have not been accounted for in our study. For the same reasons, more accurate and reliable results could be obtained if a larger measuring head were used to obtain the redness and lightness measurements.
Although non-significant, there was a visible decrease in mean Δa* at week 6 (Figure ) which suggest that fish oil supplementation for 6 weeks is sufficient to reduce inflammation, because a* is well-understood to be correlated with erythema (increased blood flow to the skin) [20
]. The non-significant linear increase in ΔL* from baseline to week 12 is also consistent with existing studies that found an inverse relationship between a* and L* in inflammatory skin diseases [22
]. However, there should be more investigation into whether L* correlates with subjective assessment of different stages of an acne lesion.
Although we cannot draw any firm conclusions from our study with a small sample size and no placebo group, there is some promising evidence that fish oil supplementation is associated with an improvement in ratings of overall acne severity, especially for individuals with moderate and severe acne. It is possible that increasing the dose of EPA from 930 mg to 3–6 grams daily, as recommended for arthritis patients, would reveal more significant results [24
]. In addition, effects of n-3 fatty acids should be examined in cohorts of subjects with the same acne severity grades or lesion counts in order to isolate the potential effect on different types of acne severity.