7 ms·

I hate studies like this. The researchers controlled for age, sex, body mass index, race, deprivation, family history of hypertension, education, smoking status

by rcme 3y ago

I hate studies like this. The researchers controlled for age, sex, body mass index, race, deprivation, family history of hypertension, education, smoking status, blood pressure, blood lipids, inflammation, blood glucose, kidney function and use of medications to lower cholesterol or blood glucose levels. But there are a million other potential causal factors that could be correlated with talking on the phone. For instance, stressors in your life can be a reason to talk on the phone, and stress is linked with hypertension. I think these types of studies can be fine if it's clear that this is just a correlation, but the researchers seem to be presenting this as if they've established a causal relationship. > Professor Qin said: “Our findings suggest that talking on a mobile may not affect the risk of developing high blood pressure as long as weekly call time is kept below half an hour. More research is required to replicate the results, but until then it seems prudent to keep mobile phone calls to a minimum to preserve heart health.” I really wish more work was done to establish a causal relationship before publishing something like this.

freitzkriesler2 3y ago
Pretty much this. Maybe the sample size were all wall street traders. Regardless, it just sounds like someone publishing to not perish.
toxik 3y ago
> Years of use or employing a hands-free set-up had no influence on the likelihood of developing high blood pressure So… radio waves cannot be the cause.
- davemp 3y ago
  Yeah I imagine the average dosage of RF across populations has been increasing enough overtime along with rollout of new technologies that you’d be able to correlate some function of adoption of these technologies to population level effects if there was any real significance.
- krunck 3y ago
  Hands free being what? Keeping using the speaker and setting the phone on a table? Using Bluetooth ear plugs with the phone in a breast pocket? So much we don't know...
mrj 3y ago
Yikes. Yeah I'm of two minds about things like this. Usually it is meant to provide some interesting data for follow-up studies, so it has to be published. But until the actual science is performed by making this recommendation and tracking the results, there's no advice or conclusion that can be drawn from stuff like this. We need more science literacy --in journalism--.
htag 3y ago
Anything that is less than double-blind is essentially a well documented hunch. In some situations this enough to merit action. This study is interesting enough to merit more science to build stronger evidence. Merits of this study: * It was large, twelve years and over 200,000 participants. * The impact was dose responsive. People that reported talking longer on the phone had greater chances of hypertension. * The impact is large enough to cause concern. * In the published article, a mechanism of action was purposed (RF-EMF), and other studies related to this mechanism were cited. * It fits the expected amount of 'controlling for' third variables you would expect, even if we cannot control for everything. Weaknesses of this study: * Mobile usage was self reported. Using empirical data such as cellphone bills or data from cellphone would have been much more accurate. There could be a bias where those with higher blood pressure perceive themselves as talking on mobile more frequently. * There was no determination of the nature of the calls. Personally, my blood pressure feels very different when I talk to family for 30 minutes than if I'm talking to Comcast for 30 minutes. * About 3/5 of the study participants dropped out. We'd expect lots of dropouts to occur with such a long study, but it could skew the data if dropouts had some statistical significance on measured values and weren't random. Dropouts are almost never random. * The article admits those with mobile phone usage seem to be at a higher natural disposition for hypertension. More likely to smoke, many other high variables. It seems plausible that some other risk factors were associated with that cluster that did not get accounted for. * No direct measurement of the purposed mechanism of action was performed.
- Jeff_Brown 3y ago
  > Anything that is less than double-blind is essentially a well documented hunch. Not true! Economists have devoted person-millenia developing the theory of instrumental variables, which lets you extract causal relationships from historical non-experimental data. That said, it's damn hard to find a good instrument. Many economists use it badly, and most doctors don't even know the theory.
  - htag 3y ago
    I have my doubts about IVs in practice, but I agree they theoretically are an acceptable substitute. The problem is we conduct science on systems we don't fully understand. Because we don't fully understand it, it's easy to miss important information we didn't know is important. The double blind with random selection allows us to control for all important information, determine causation, and is the best known method for controlling for bias. We don't have to worry about missing important variables because the nature of random selection means those variables are randomly distributed.
ravenstine 3y ago
The frequency of shark attacks at the beach is correlates with the number of ice cream sales, therefore, ice cream must cause shark attacks! /s
Aunche 3y ago
>But there are a million other potential causal factors that could be correlated with talking on the phone. This is basically the logic that cigarette companies used to deny that they were causing lung cancer. A longitudinal study is about as good as you can get for testing these sort of hypotheses.
- JackFr 3y ago
  No. That was the mechanism the cigarette companies used to deny they were causing heart disease. It turns out there’s really no such thing as a Type A personality. Those were just smokers.
- rcme 3y ago
  True, but in the case of tobacco establishing a causal relationship was studied, e.g. https://pubmed.ncbi.nlm.nih.gov/13116124/ https://pubmed.ncbi.nlm.nih.gov/13116124/ (1953). Trying the same with RF-EMF would be an obvious next step. Just a personal opinion, but I don't think the results will hold in direct study.
clircle 3y ago
Yes and no. I understand that a causal relationship is the goal, but short of an experimental design, anything done is going to have similar holes punched through it. My view, as others have noted, is that this is a just one correlation. I'll feel that it's more causal once we see it pop up again and again in different studies, with different populations, controls, analyses, timescales, etc. It's ok to document correlations, but researchers need to temper their language when discussing correlation studies in general.
dehrmann 3y ago
It's 2023. Now that we've got SMS, we don't call anyone on our phones anymore.
OJFord 3y ago
> For instance, stressors in your life can be a reason to talk on the phone, and stress is linked with hypertension. Nevermind that, maybe the % correlation they found is simply roughly the proportion that find phone calls themselves stressful.

by rcme 3y ago

freitzkriesler2 3y ago

Pretty much this. Maybe the sample size were all wall street traders. Regardless, it just sounds like someone publishing to not perish.

toxik 3y ago

> Years of use or employing a hands-free set-up had no influence on the likelihood of developing high blood pressure So… radio waves cannot be the cause.

davemp 3y ago
Yeah I imagine the average dosage of RF across populations has been increasing enough overtime along with rollout of new technologies that you’d be able to correlate some function of adoption of these technologies to population level effects if there was any real significance.
krunck 3y ago
Hands free being what? Keeping using the speaker and setting the phone on a table? Using Bluetooth ear plugs with the phone in a breast pocket? So much we don't know...

mrj 3y ago

Yikes. Yeah I'm of two minds about things like this. Usually it is meant to provide some interesting data for follow-up studies, so it has to be published. But until the actual science is performed by making this recommendation and tracking the results, there's no advice or conclusion that can be drawn from stuff like this. We need more science literacy --in journalism--.

htag 3y ago

Anything that is less than double-blind is essentially a well documented hunch. In some situations this enough to merit action. This study is interesting enough to merit more science to build stronger evidence. Merits of this study: * It was large, twelve years and over 200,000 participants. * The impact was dose responsive. People that reported talking longer on the phone had greater chances of hypertension. * The impact is large enough to cause concern. * In the published article, a mechanism of action was purposed (RF-EMF), and other studies related to this mechanism were cited. * It fits the expected amount of 'controlling for' third variables you would expect, even if we cannot control for everything. Weaknesses of this study: * Mobile usage was self reported. Using empirical data such as cellphone bills or data from cellphone would have been much more accurate. There could be a bias where those with higher blood pressure perceive themselves as talking on mobile more frequently. * There was no determination of the nature of the calls. Personally, my blood pressure feels very different when I talk to family for 30 minutes than if I'm talking to Comcast for 30 minutes. * About 3/5 of the study participants dropped out. We'd expect lots of dropouts to occur with such a long study, but it could skew the data if dropouts had some statistical significance on measured values and weren't random. Dropouts are almost never random. * The article admits those with mobile phone usage seem to be at a higher natural disposition for hypertension. More likely to smoke, many other high variables. It seems plausible that some other risk factors were associated with that cluster that did not get accounted for. * No direct measurement of the purposed mechanism of action was performed.

Jeff_Brown 3y ago
> Anything that is less than double-blind is essentially a well documented hunch. Not true! Economists have devoted person-millenia developing the theory of instrumental variables, which lets you extract causal relationships from historical non-experimental data. That said, it's damn hard to find a good instrument. Many economists use it badly, and most doctors don't even know the theory.
- htag 3y ago
  I have my doubts about IVs in practice, but I agree they theoretically are an acceptable substitute. The problem is we conduct science on systems we don't fully understand. Because we don't fully understand it, it's easy to miss important information we didn't know is important. The double blind with random selection allows us to control for all important information, determine causation, and is the best known method for controlling for bias. We don't have to worry about missing important variables because the nature of random selection means those variables are randomly distributed.

ravenstine 3y ago

The frequency of shark attacks at the beach is correlates with the number of ice cream sales, therefore, ice cream must cause shark attacks! /s

Aunche 3y ago

>But there are a million other potential causal factors that could be correlated with talking on the phone. This is basically the logic that cigarette companies used to deny that they were causing lung cancer. A longitudinal study is about as good as you can get for testing these sort of hypotheses.

JackFr 3y ago
No. That was the mechanism the cigarette companies used to deny they were causing heart disease. It turns out there’s really no such thing as a Type A personality. Those were just smokers.
rcme 3y ago
True, but in the case of tobacco establishing a causal relationship was studied, e.g. https://pubmed.ncbi.nlm.nih.gov/13116124/ https://pubmed.ncbi.nlm.nih.gov/13116124/ (1953). Trying the same with RF-EMF would be an obvious next step. Just a personal opinion, but I don't think the results will hold in direct study.

clircle 3y ago

Yes and no. I understand that a causal relationship is the goal, but short of an experimental design, anything done is going to have similar holes punched through it. My view, as others have noted, is that this is a just one correlation. I'll feel that it's more causal once we see it pop up again and again in different studies, with different populations, controls, analyses, timescales, etc. It's ok to document correlations, but researchers need to temper their language when discussing correlation studies in general.

dehrmann 3y ago

It's 2023. Now that we've got SMS, we don't call anyone on our phones anymore.

OJFord 3y ago

> For instance, stressors in your life can be a reason to talk on the phone, and stress is linked with hypertension. Nevermind that, maybe the % correlation they found is simply roughly the proportion that find phone calls themselves stressful.