Dr Caroline Leaf – credit where credit’s due

screen-shot-2016-10-27-at-10-24-44-pm

It’s not often I see something positive in Dr Leaf’s work, but today was one such occasion.

I’m often (legitimately) critical of Dr Leaf’s paucity of references and citations for her Facebook posts and social media memes.  Today was different – Dr Leaf made a statement and backed it up with an easily obtainable peer-reviewed journal article.  It’s a shame it wasn’t backed up by an accurate interpretation, but it’s a positive step none-the-less.

Dr Leaf claimed that “People who served others experienced a 68% increase in healing compared to those who only got treatment for themselves.”

Since the article was so easy to find, I decided to look it up.  The article was by Poulin et al, “Giving to others and the association between stress and mortality”, in the American Journal of Public Health [1].  Actually, the article was familiar, because Dr Leaf has written about the same article before, but her social media post that time was more nebulous.

So does the study by Poulin and his colleagues show that people who served others experienced a 68% increase in healing compared to those who only got treatment for themselves?  In a word … no.

First of all, the study wasn’t looking at healing, it was looking at mortality.  They may seem similar, but getting better from something (“healing”) is not the same as not dying from something (“mortality”).

Second, no one in the study was being “treated”.  I’m not sure where Dr Leaf got the idea that the control group was getting “treatment”.  The study compared those who self-reported “helping behavior directed toward close others … in any of 4 unpaid helping activities directed toward friends, neighbors, or relatives who did not live with them” versus those that did not.

Thirdly, there’s no mention of a 68% improvement anywhere in the article.  The article gives its results as hazard ratios.  For the non-statisticians, the hazard ratio is “the ratio of the particular event taking place in treatment group compared to control group.”  The simplest (probably over-simplified way) way of thinking about hazard ratios is to do a simple sum – the hazard ratio minus 1 is the percentage increase or decrease in risk, where a positive number is an increased risk and a negative number is a decreased risk.  So a hazard ratio of 1.13 means that a person in the exposure group has a 13% increased risk compared to the control group (=1.13 – 1).  And a hazard ratio of 0.7 means a 30% decreased risk (0.7 – 1 = -0.3).  So for the helping group to have a 68% decreased risk of dying, the hazard ratio would be 0.32 (0.32 – 1 = -0.68).

If you’re lost in the numbers, don’t stress.  The point is that Dr Leaf was very specific about the helping group increasing in healing by 68%, but there’s nothing in the results to suggest this.  The study authors wrote, “When we adjusted for age, baseline health and functioning, and key psychosocial variables, Cox proportional hazard models for mortality revealed a significant interaction between helping behavior and stressful events (hazard ratio [HR] = 0.58; P < .05; 95% confidence interval [CI] = 0.35, 0.98). Specifically, stress did not predict mortality risk among individuals who provided help to others in the past year (HR = 0.96; 95% CI = 0.79, 1.18), but stress did predict mortality among those who did not provide help to others (HR = 1.30; P < .05; 95% CI = 1.05, 1.62).”  Unless I’m missing something, there’s nothing in the results that remotely suggests a 68% improvement in anything.

And for what it’s worth, the study shows very weak associations anyway (in statistical terms, the confidence intervals are broad, and almost cross 1), so even if the study really did say something about a “68% increase in healing”, it’s something that is only slightly more likely to occur than by chance alone.  Then there’s other evidence that contradicts this particular study’s findings, so in all fairness, this study shouldn’t be used to base social media memes on in the first place.

Overall, it’s good that Dr Leaf cited an article in her social media meme, but her interpretation of the study was poor, something more at the level of a university freshman than a supposed expert in her field.  And it reflects badly on the Christian church that this is the level of ‘expertise’ that the church accepts and then promotes.

I would encourage Dr Leaf to continue to cite references for her memes, but she really needs to learn how to interpret clinical studies if she and the church are going to continue to promote her as some sort of expert.

References

[1]        Poulin MJ, Brown SL, Dillard AJ, Smith DM. Giving to others and the association between stress and mortality. Am J Public Health 2013 Sep;103(9):1649-55.

Does helping others help you?

John Holmes wrote “There is no exercise better for the heart than reaching down and lifting people up.”

We all know that exercise is good for us, but is the exercise of the heart, “reaching down and lifting people up” just as good for us?

Dr Caroline Leaf is a communication pathologist and self-titled cognitive neuroscientist.  Her meme of the day today was a claim that “Helping others can increase your lifespan.”  She explained that “Researchers found a link between serving others, improved health and decreased mortality! See more at: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3780662/pdf/AJPH.2012.300876.pdf”.

Screen Shot 2016-01-16 at 6.28.07 PM

The journal she referenced was a 2013 article by Poulin et al in the American Journal of Public Health [1].  Poulin and his colleagues examined data from nearly 850 people in the Detroit area.  At the start of their study, they asked their participants about stressful life events in the last year and whether they provided tangible assistance to friends or family members.  They then followed their participants for five years and analysed the characteristics of who died in that time.

According to the study by Poulin, those who helped others were younger, healthier, more likely to be White, of higher socioeconomic status, and higher in social support and social contact than those who didn’t help, all factors that have been shown to influence mortality.  They also noted that 70% of their cohort didn’t experience any stressful life events.  While they adjusted for these variables, their statistics would still be affected by them.  As it turns out, while their results were significant, their numbers had broad confidence intervals, so the effect they found is very weak.

What about other studies looking at the same question but in a different way?  Well, there are mixed findings.  Roth and colleagues published a study in 2013 in the American Journal of Epidemiology which also showed that care-givers had better life expectancy than matched controls [1] but then a number of other studies show the opposite.  The Caregiver Health Effects Study found that those who were providing care to a disabled spouse and who reported some strain associated with that care had a 63% elevated risk of death compared with non-caregiving spouses [2]. Other studies suggest that caregivers have poorer mental and physical health status than non-caregivers [3], and caregiving has been widely portrayed as a serious public health problem in the professional literature [4, 5].

So while Poulin found a loose association between helping others and decreased mortality, Dr Leaf has taken that a step too far:

> Firstly, correlation does not equal causation.  Just because a study found those who helped others had a decreased mortality doesn’t mean that the reverse, helping others increases your lifespan, necessarily holds.  There may be other explanations.
> Secondly, other studies show conflicting results, so Poulin’s study may be a statistical hiccough.

It’s not clear that helping others is actually good for our health.  That doesn’t mean to say we shouldn’t help others. I think we should, if for no other reason than the golden rule, “Do unto others as you would have them do unto you.”  But we can’t definitively say that helping others will help us directly by making us live longer.  That’s scientifically still up in the air.

References

[1]        Poulin MJ, Brown SL, Dillard AJ, Smith DM. Giving to others and the association between stress and mortality. Am J Public Health 2013 Sep;103(9):1649-55.
[2]        Schulz R, Beach SR. Caregiving as a risk factor for mortality: the Caregiver Health Effects Study. JAMA : the journal of the American Medical Association 1999 Dec 15;282(23):2215-9.
[3]        Pinquart M, Sorensen S. Differences between caregivers and noncaregivers in psychological health and physical health: a meta-analysis. Psychol Aging 2003 Jun;18(2):250-67.
[4]        Talley RC, Crews JE. Framing the public health of caregiving. Am J Public Health 2007 Feb;97(2):224-8.
[5]        Centre for Disease Control and Prevention. Caregiving, A Public Health Priority.  2010, 7 Dec 2010 [cited 2016 Jan 16]; Available from: http://www.cdc.gov/aging/caregiving/index.htm