Referee 2's Review (Melissa Andrew):

Summary

This is manuscript describes a study addressing an important question: Is mild anemia clinically important in terms of association with relevant constructs such as cognition, functional status, and quality of life (QOL) in elderly people? In this cross-sectional study with a case-control design, mild anemia was associated with worse selective attention, and worse disease-specific (anemia-specific) QOL. The authors found that mild anemia, as defined in their study, was not associated with other domains of cognitive function, mood, or general QOL.

The definition and measures used appear to be well validated, many of them with existing Italian versions. The findings are discussed in the context of relevant existing literature.

In general, this paper is well written and may be of interest to clinicians caring for older adults and to public health services considering anemia-related interventions. It makes a valid contribution to the scientific literature, and merits publication.

Comments

1. In general, inclusion and exclusion criteria are well described. Individuals over age 84 were not invited to participate in the study. Why was this? It would be beneficial to justify this choice in the paper.

2. The exclusion criteria were quite stringent, essentially excluding the sickest and frailest older people. What effect might this have had for the generalizability of the study findings, and their applicability to the population of older adults encountered in health care settings?

3. The authors state that investigating the association between anemia and cancer was a secondary aim of the study. Is this going to be addressed in another manuscript? If not, perhaps more discussion on this topic is warranted.

4. Page 5, 1st full paragraph:
a. Who collected the medical history, current drug use, use of health services...? The wording suggests that it was the psychologists, but the context suggests it was the nurses. Perhaps this point could be clarified.
b. What is meant by "trustworthiness of the interview" - is this a comment on validity, or reliability of the information provided by the subject in the psychologist's view? This could be clarified.

5. Page 6, paragraph 4 (function). It would be interesting to know how the IADL scales differed for men and women - perhaps the authors could list the 8 activities used for women and the 5 for men. Presumably this difference is felt to be culturally appropriate in terms of task division by sex in this population.

6. In general the statistical analyses seem robust. I would caution against the use of the word "outcomes" in cross-sectional analyses (p.8, paragraph 2). Instead, "dependent variable" or an equivalent term that doesn't imply causation could be used. The authors do clearly and appropriately state in the discussion section that this cross-sectional observational design does not allow for determination of causation.

7. The "sensitivity analysis" approach to investigating the influence of chosen anemia cutoffs is a nice feature of this paper.

8. P. 9, last paragraph: With a 1.5 hour interview, was there any missing data (e.g. due to participant fatigue)? If so, how was missing data dealt with in the analysis?

9. Tables 1, 2, and 3: I find these unclear, where they have combined differences and odds ratios. Particularly, for example, where a difference of 1.0 (0.5 to 2.0) would be important but the same OR is not, it is confusing to have the two mixed into the same column. This might be improved by having 2 separate columns (one for differences, one for ORs), reporting them in 2 separate parts in the table, or at least by indicating the statistically significant ones somehow (with symbols, bold text, etc...) so that the results are clear at a glance.

10. Worse performance in selective attention was the only cognitive finding associated with mild anemia in this study. Do the authors have any hypotheses that might account for this, and which might inspire further research in the area?

11. The use of the term "the elderly" may be interpreted by some as having a derogatory, or ageist, connotation. Perhaps the authors might consider the use of alternatives such as "older adults", "elders", "older people". Even the use of "elderly" as an adjective (as in elderly people, or the elderly person) may be less problematic than its use as a noun.

**********
N.B. These are the comments made by the referee when reviewing an earlier version of this paper. Prior to publication the manuscript has been revised in light of these comments and to address other editorial requirements.