Depression and Anxiety Symptoms in UK Thalidomide Survivors: A Brief Survey

Reviewers Comments

Response

P 3/111-119: Two German Studies were cited with the prevalence rate of depression in the general population of this country. There are no citations of prevalence rates in UK, but in line 119 there is a reference to the "general UK population". The levels of depression are quite different in both countries. They are higher in Germany (9,2%) than in UK (7,4%), in addition the patterns of depressive symptoms in different age groups are quite different, too. (see data from EHIS: Journal of Affective Disorders 279 (2021) 164–172 and Journal of Health Monitoring · 2019 4(4)DOI 10.25646/6227 Robert Koch Institute, Berlin). Please revise.

Thank you very much for directing us to these more recent studies. The Arias de la Torre (2021) paper with prevalence rates for the UK was particularly helpful. We have re-written section 3.1 incorporating the UK figures and we think this has strengthened it considerably.

P4/141: Analysis of GAD-7 scores. It would be of interest to add data from Thalidomide survivors who had either both diagnoses, or only one diagnose.

We do not have data on confirmed clinical diagnoses for the sample, so we can only comment on the presence or absence of depressive/anxiety symptoms. In section 3.2 we have added further information about the number of respondents who had no symptoms of anxiety or depression and those who had anxiety and/or depression.

P4/152: Please check your reference numbers: e.g. Meltzer et al is cited under Number 16, not 14.

This has been corrected as part of revising and up-dating all the references.

I like this article a lot, because of the very insightful and coherent discussion, the authors know TE affected people and their special problems very well. The topic is very important concerning prevention and healthcare.

Thank you. We also feel that is an important and under-reported topic, which has practical implications for thalidomide survivors lives.

Reviewers Comments

Response

This brief manuscript explores an intriguing question: are the now-older adults who were “Thalidomide babies” in the late 1950s and 1960s vulnerable to symptoms of depression and anxiety? The clearly written paper, based on a small non-representative cross-sectional sample, finds that the prevalence rates of depression and anxiety are higher among the survivors, relative to rates documented in other studies of the general population. Despite these many strengths, the manuscript has a number of limitations that require the authors’ attention.

We very much appreciate your comments and agree that the paper had limitations that needed addressing. Both reviewers’ comments have greatly helped us to do this.

1. In the introduction, I would suggest providing information on how widespread use of Thalidomide was, to give some sense of the magnitude of the problem and the proportion of births in the late 1950s and early 1960s that were exposed.

We have revised the first paragraph of the introduction to provide this contextual information.

2. Seeing as the sample is now in their upper 50s and lower 60s, I would suggest providing some conceptual framework that is relevant to their life course stage. For instance, cumulative disadvantage perspectives would suggest that disadvantages experienced earlier in life widen over time, and thus there may be quite large gaps in mental health among those with versus without exposure. Alternatively, literature on disability and stress more generally suggests that individuals with early onset disability or impairment accommodate and adapt to their symptoms over time, and thus are better adjusted psychologically than those with later onset conditions. This kind of framing might help to situate the analysis in a larger literature and set a framework for interpreting the results.

Thank you for these very helpful suggestions. We have added a new paragraph to the introduction to reflect the points you highlight and picked them up again in the changes to the discussion. 

3. The sample focuses on Thalidomide survivors only, thus it is difficult to interpret the results. I would suggest benchmarking the study’s descriptive statistics against population-level data both against survivors overall and of the general population, to get a sense of potential sources of sample bias. The fact that these sample members both survived until late middle-age and participated in a survey suggests that the may be particularly robust and well-adjusted relative to those who did not survive nor participate in the study. More generally, the authors should discuss the possible impact on the results of sample bias.

In the Materials and Methods section we have added information about the age range of Thalidomide survivors.

At the beginning of the Results section we have added information about the gender split of the UK thalidomide survivor population as a whole and noted that the sample population does not fully reflect this, with more women than men responding. We have also noted this in the Limitations section.

In the first paragraph of the Introduction we explain that those Thalidomide survivors who lived beyond childhood are expected to have a near normal life expectancy. From our own work and the wider literature there is no evidence that those Thalidomide survivors who reached late middle age are more robust or well adjusted. However, it is possible that those with more vulnerable mental health may have been less likely to respond to the survey and so we have noted this in the Limitations section.

4. When describing the methods, it is not necessary to report on those variables collected in the survey that are not used in the analysis.

Thank you for this guidance. We have removed the descriptions of the measures not included in analysis (though we do note that two sections of the questionnaire were not included in our analysis) (p.2).  

5. Why transform the continuous mental health measures into categorical measures? What rationale is used for the selection of cut points – are these clinical cut points? Were sensitivity analyses conducted, using continuous scores?

We transformed the continuous mental health variables into clinical categories that are commonly used for these measures. This enables us to compare levels of symptom severity in our sample to other groups. We have added the references for these cut-offs (Kroenke et al, 2001; Spitzer et al., 2006). 

We do not have information regarding confirmed diagnoses of depression and or anxiety for our sample, and so we are not able to do a sensitivity analysis. The PHQ9 has been found to have good sensitivity (see Kroenke et al, 2001). 

We do use the continuous PHQ-9 variable for means and SD for males/females, and we have added the overall mean for the sample for context (p3 PHQ-9, p4 GAD7)

6. How were missing data handled in the multivariate analyses? Were values imputed? Were cases dropped? Please discuss the implications of the missing data imputation for interpreting the study results.

We had a small amount of missing data (see table 1 and table 2). These cases were excluded from the relevant analyses. The number of participants excluded from the relevant analyses is now stated.

7. When discussing the demographic composition of the sample, please give some sense of how this compares to the more general population of Thalidomide survivors, and the general population of persons roughly ages 55-64.

We agree that it is important to compare our sample population with the UK thalidomide survivor population as a whole. We explain in Comment 3 above, that in the Materials and Methods section we have added information about the age range of thalidomide survivors, and in the Results section we have added information about the gender split of the UK thalidomide survivor population as a whole.

Apologies but we are unclear what further information about the general population is needed.

8. The discussion is generally good, although I question some claims. A quite vast literature shows that persons with disability report higher levels of depressive symptoms, with these effects typically found at the most intense levels of impairment (see work by Warner, Carr, Freedman, Taylor, and others). This stands in contrast to the results cited at the top of p. 5, suggesting that disability is unrelated to mental health. I would suggest that the authors discuss more fully different potential patterns among those with earlier-versus later-acquired disabilities with attention to processes of adaptation.

Thank you very much for these comments and guidance. They really helped our thinking. We have revised the Discussion section to reflect the points you make and to incorporated a number of additional references, both in this section and the Introduction (in particular referencing the work of Carr, Freedman and Taylor).

9. More generally, given the sources of bias in the non-representative sample, a more productive analytic path may be to identify the sources of heterogeneity in the mental health symptoms of Thalidomide survivors. It still would not be possible to discern whether the patterns detected (e.g., higher rates of depression of women) are unique to the survivor population, but would help to minimize concerns about focusing on prevalence rates in a non-representative sample.

Although our response rate was good (44%) we do accept that our sample may not be representative of the whole UK population of Thalidomide survivors. However, the survey was conducted as part of wider but still modest study and so we were limited as to the number of variable we could collect. We focussed on being able to work, as previous research about the health related quality of life of Thalidomide survivors had shown this to be important. We very much agree that a number of psychosocial factors, not explored in this survey, may be important in accounting for the heterogeneity in the mental health symptoms amongst Thalidomide survivors, and we hope to explore this in future studies.

With attention to these concerns, the manuscript will make an interesting contribution to the literature.

Thank you. We know this study has a focus on thalidomide survivors but we believe their experiences can contribute to our understanding of disability and depression more widely.