Epi in the News

Hi Sujit,

These are my initial thoughts:
1. This is an incidence study because it measures the number of new cases of ADHD and dementia.
2. People with an existing diagnosis of dementia or ADHD were excluded from the study. So not having these conditions is a criteria for inclusion.
3. Incidence of dementia in newly diagnosed ADHD patients = New cases / Population x Time frame = 96 / 730 x 17 = 0.0077 Or 77 new dementia cases per 10,000 people-years within the adult ADHD population.

I hope that's correct - although I'm not sure...

I was also wondering if we could discuss this study together at some point: https://www.fredhutch.org/en/news/center-news/2018/07/endometriosis-linked-to-childhood-abuse.html

I'm interested in the mind-body connection. Thanks.

Hello Sujit,

Please let me know if my calculation is correct/wrong (sorry as tried to paste 2x2 table)

For new students
1. In this study, is dementia measured with prevalence or incidence? – Answer: Dementia is measured with incidence (prospective cohort study)
2. What are the eligibility criteria to be in this study? – Answer: The eligible sample consisted of Israeli citizens who were nationwide Meuhedet members born between 1933 and 1952
3. Calculate and name a crude (unadjusted) relative risk.
Answer:

Dementia + Dementia - Total
ADHD + 96 634 730
ADHD - 7630 100 858 108 488
Total

RR=(96/730)/(7630/108488) = Risk ratio: 1.9

Hi Sujit, Feifei and Gulnaz,

4. What are key advantages of using existing medical records to conduct a study? And the key disadvantages?

Advantages are cost and speed as the data has already been collected. Selection bias should be low as this should be a population-based sample. Medical records from countries with universal centrally-funded healthcare are particularly useful in this regard.

Medical record studies suffer from misclassification. Not every individual is tested for ADHD or dementia. There are likely to be many individuals with these conditions undiagnosed in the population. Conversely, those with diagnoses may not have the condition as there is no gold standard universal test being applied.

After taking into account factors including age, sex, socioeconomic status, smoking and various health conditions...

5. What does it mean to take smoking 'into account'? Was this a justifiable decision?

Taking into account means treating smoking as a confounder. This is justifiable if smoking is not thought to be on the causal pathway between ADHD and a subsequent diagnosis of dementia. It is possible that smoking might be on the causal pathway, and if so it would be a mediator of the association, and adjusting for it would artificially reduce the true association between the two.

6. What was the effect modification the investigators found?

Prescriptions of psycho-stimulant medication. The association between ADHD and dementia was only seen for participants not taking psycho-stimulant medication, and not seen in participants prescribed this medication.

7. What mediating (causal pathway) mechanisms did the investigators hypothesise?

However, Prof Chris Hollis, of the University of Nottingham, said there could be a number of factors muddying the waters. “Those adults who seek and receive an ADHD diagnosis are also more likely to be assessed for other cognitive/neuropsychiatric conditions including dementia,”

8. What epidemiologic concept is Prof Hollis implying here? Explain the implication.

Professor Hollis is implying ascertainment bias. If some patients are more likely to request investigations then they will be more likely to acquire diagnoses, both of ADHD and of dementia. The association seen could therefore be spurious and caused by differential patient behaviour, and not by any true association.

9. This study was done with medical records from one country. Does that mean the findings are not generalisable to other countries?

The findings are only not generalisable to other countries if there is a reason to suppose that the diagnoses of ADHD and dementia or the association between the two would be different in Israel from other countries.

10. Does this hypothesis require an RCT to confirm causation?

It would not be possible to design an RCT to confirm causation of this hypothesis. A Mendelian randomisation study might be helpful, if genetic variants associated with ADHD are known, to reduce the problem of selection bias in classification of the exposure ADHD. The outcome, dementia would be better classified if all patients in the study underwent screening for dementia, but this would be expensive and also many patients might decline screening. There would also be an ethical question about screening for dementia as there is limited effective treatment if this is diagnosed early.

Best wishes

Judith

Hi Sujit and fellow students,

The following are my answers:

1. In this study, dementia measured with the incidence.

2. The eligibility criteria were: i) adult people who were included in electronic health records of the nonprofit health maintenance organisation (HMO) Meuhedet Healthcare Services in Israel; and ii) those who were diagnosed with ADHD over the study period. 

People with an existing diagnosis of ADHD or dementia were excluded from the study.

3. Relative risk = 96/730 ÷ 7630/(109218-730) = 0.1315 ÷ (7630/108488) = 0.1315 / 0.0703
= 1.87

4. Key advantages of using existing medical records to conduct a study:
i. Easy availability of data
ii. Cost-effective
iii. Quicker

Key disadvantages:
i. Quality of data might be poor
ii. Data might be incomplete
iii. Chance of missing data, under-reporting
iv. data might not be up-to-date
v. Chance of misclassification of exposure or outcome

5. Taking smoking ‘into account’ means the authors considered smoking as a confounder.
It is a justifiable decision as after searching literature, I found that smoking might be a mediator in this relationship. People with ADHD are more likely to smoke, and smoking also increases the risk of dementia. So, adjusting for smoking would have reduced the true association between ADHD and dementia. So, there is need of further research to establish this association.

6. ADHD treatment might be an effect modifier in this relationship as the investigators did not find a clear association between ADHD and dementia risk among those treated with psychostimulant medication. It needs to be studied further.

7. Mediating (causal pathway mechanisms) that the investigators hypothesise:
the processes involved in adult ADHD may reduce the ability of the brain to compensate for the effects of processes that can happen later in life including neurodegeneration or blood flow in the brain, which might be linked to a higher risk in developing dementia.

8. There is chance of misclassification and ascertainment of outcome. According to Prof Hollis, brain imaging to confirm the dementia diagnosis could have been more reassuring. It would help to exclude the presence of other cognitive/neuropsychiatric conditions, and would reduce the chances of misclassification.
Moreover, the study results cannot be generalised to all adult population in Israel as there might be a large number of people in the community who has poor health-seeking behaviour and therefore not getting diagnosed with ADHD and dementia.

9. The findings might not be generalizable to other countries. The quality of data obtained from medical records might vary from country to country. Moreover, patient characteristics (socio-demographic background, behaviour, culture and underlying health conditions) and diagnostic accuracy might vary across different countries.

10. RCT of a treatment intervention for ADHD might help confirm the causal association, as the experts highlighted the need to examine whether ADHD medications mitigate any potential dementia risk. However, conducting RCT might not be possible due to ethical issues and resource constraints.

Thank you!
Madhutandra