Krill assessment off base

I read with interest the column in the March issue by Anthony Almada entitled ?Icy Atlantic waters may hold the secret to super omega-3s? in which he raised concerns regarding the studies on krill oil. I would like to address these concerns:

  1. ?This study?s findings need to be interpreted with caution.? I agree scrutiny makes scientifically bound medical research prevail.
  2. ?Diagnostic methods are archaic.? The criteria used are established by the American College of Obstetricians and Gynaecologists (Practice Bulletin No. 15, 2000 April) and are currently in use.
  3. ?...absence of a prospective daily symptom diary invites misdiagnosis of PMS.? We used the ACOG-PMS Questionnaire in lieu of a diary because of reduced time for completion, standardisation of timing and environment and prevention of confounding due to stressful life events that bias a diary.
    We removed the possibility of ascertainment bias by using a standardised measurement tool. In fact, the use of a diary is subject to higher risk for ascertainment bias due to lack of standardisation and validated scoring systems. There is no possibility of differential misclassification bias since this was a randomised, placebo-controlled, double- blind, prospective clinical trial.
  4. ?...did not exclude subjects who display a high placebo response.? The statistical comparisons between active and control groups allow estimates of treatment effects adjusted for the placebo effect. Excluding these patients overestimates the treatment effect, is highly unethical, seriously scientifically flawed and should never be practiced.

With respect to conflict of interest:

  1. Industry-sponsored clinical trials are routinely conducted with close sponsor involvement. This does not mean they are biased or invalid.
  2. The independent ethics review evaluated all issues of potential conflicts of interest.
  3. Data analyses were conducted by an independent CRO affiliated with McGill University and reviewed by two independent biostatisticians.
  4. The study was reviewed by a committee of peers that decided it was publication worthy.
  5. As a scientist with an extensive publication record and affiliations with North American universities, I never jeopardise the credibility and peer respect that I have earned.

Tina Sampalis, MD, PhD VP of R&D and Business Development Neptune Technologies and Bioressources

Anthony Almada, MSc, replies
Dr. Sampalis confuses diagnostic criteria in the context of diagnosis with assessment/outcome measures. In their paper,1 Sampalis et al defined eligibility for the study as ?Women of reproductive age who fulfilled the DSM-III-R diagnostic criteria for PMS.? DSM-III-R (Diagnostic and Statistical Manual of Mental Disorders, Third Edition, Revised) was published in 1987 and was superseded by DSM-IV in 1994, with the 2000 ACOG Practice Bulletin criteria for diagnosis2 now being widely used, as Sampalis states.

The ACOG diagnostic criteria were not used to confirm a diagnosis of PMS prior to randomisation but only to design a nonvalidated questionnaire for the primary outcome measure. Additionally, the absence of the enrolled subjects completing prospective daily diaries for at least two symptomatic menstrual cycles, as described in DSM-III-R, and in Table 1,1 to confirm a diagnosis, could indeed lead to a misdiagnosis of PMS.

We would like to hear your views on anything you have read in FFN or on any industry issues. Please write to Peter Sofroniou, editorial director, at [email protected]
In a paper I cited, 3 Rapkin and colleagues from UCLA have shown that retrospective recall (the method apparently used by Sampalis et al) correlates with prospective daily diaries only 50 per cent, with the former method leading to higher symptom scores. Moreover, because a placebo run-in phase prior to randomisation was not employed (which would have excluded placebo responders from enrollment) and a large number of symptoms of mild to moderate severity were assessed, wide variations in measured outcomes may have ensued. 5 The lack of a true placebo in the study, the finding that 64 per cent of the patients receiving fish oil ?complained of unpleasant reflux? (vs. 0.0 per cent in the krill group), and the failure to utilise an objective measure of compliance (eg, fatty acid profile of serum phospholipids) all raise additional concern regarding blinding and compliance.

Industrial-academic collaborations are not uncommon. However, this study is the only evidence in the public domain describing the effects of krill oil in PMS, with the lead author also being the recipient of stock options from Neptune.4 Also, the journal in which this article was published (Alternative Medicine Review) is owned by a marketer of dietary supplements to health practitioners, pharmacies and veterinarians (Thorne Research), which also markets its own krill oil product.

If this column resulted in the performance of another randomised controlled trial , conducted by truly independent and financially disinterested researchers with a surfeit of experience in this area, I would be wholly gratified.

1. Sampalis F, et al. Evaluation of the effects of Neptune Krill Oil? on the management of premenstrual syndrome and dysmenorrhea. Altern Med Rev 2003; 8:171-9.
2. ACOG (American College of Obstetricians and Gynecologists). Premenstrual Syndrome. ACOG Practice Bulletin No. 15, April 2000.
3. Rapkin AJ, et al. Comparison of retrospective and prospective assessment of premenstrual symptoms. Psychol Rep 1988; 62:55-60.
4. Neptune Technologies & Bioressources Inc Quarterly Report for the Period ending November 30, 2003. BC Form 51-901F.
5. Rapkin AJ. A review of treatment of premenstrual syndrome & premenstrual dysphoric disorder. Psychoneuroendocrinology 2003 28:39-53.

FFN reserves the right to shorten letters.

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.