Perceived caregiver stress in Alzheimer's disease and mild cognitive impairment: A case control study.

Sir,

We thank the reader for having gone through the article[1] keenly and the editor for asking us to respond. We agree that choosing the right controls for the study is an important step in a case–control study.[2]

Matching of cases and controls addresses the issues of confounding in the design stage of a study.[2] Controls are matched to cases on one or more attributes (i.e., age, gender, smoking status, etc.).[3] Alternatives to matching are to do frequency matching, use multivariate analyses to control confounding, etc.[23] One disadvantage of matching is that we cannot always find an exact match, a fact that is also agreed by the reader of this paper. We explained in the current paper[1] and also in our previous papers[4] of well-known sex bias in reporting dementia patients in India; hence, finding sex matching is difficult. Also, too much matching can make cases and controls too similar.[23] We used younger controls as the age has been causally linked with perceived stress in early onset Alzheimer's disease (AD) compared to late onset AD.[5] It has been said that the matching variable should not be associated with causality.[5] This may lead to unnecessary matching. Furthermore, it has been said that if the matching factor is associated with the disease but not with the exposure, matching will be less efficient. In addition, two age groups in this study were not statistically different (P > 0.05).

In addition, it has been pointed that gender and social status have not been taken into account. We humbly feel that the reader has confused “caregiver stress” with “caregiver burden” here, which may have some interrelation, but are not entirely same. We have studied perceived caregiver stress and not caregiver burden in the current study. While it is generally agreed and we too have mentioned that females are a major source of informal caregiving among Indian patients with dementia,[1] this question about gender is assumedly of little significance. In addition, the level of cognitive dysfunction and hours of caregiving have been linked with caregiver burden if readers mean the same rather than social and gender issues as per the available literature.[6] A large study of >6000 adults has shown that perceived stress does not vary as per the factors suspected by readers. The results of this large study[1] were similar after adjusting for demographic variables, smoking, systolic blood pressure, body mass index, chronic medical conditions, and psychosocial factors and did not vary by race, sex, age, or education.[7]

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.