Not all scientific studies are created equal - David H. Schwartz

並非所有的科學研究皆如出一轍 － 戴維．施瓦茨

譯者: Isabel Lin 審譯者: Neoh Yi Yuan

Studies have shown that

研究顯示

taking vitamins is good for your health

服用維他命對你的健康有益

and bad for your health.

但也對你的健康有害。

That newly discovered herb can improve your memory

最新發現的草藥能增強你的記憶力，

or destroy your liver.

也能損壞你的肝臟功能。

Headlines proclaim a promising new cancer treatment

頭版新聞刊登了一項 充滿希望的癌症新療法後

and never mention it again.

卻不再提起。

On a daily basis,

日復一日地，

we are bombarded with attention-grabbing news,

我們被這些顯著且有科學研究 為佐證的新聞轟炸著，

backed up by scientific studies,

but what are these studies?

但是，這些研究到底是些什麼呢？

How are they performed?

它們是如何進行的？

And how do we know whether they're reliable?

還有，我們如何知道它們是否可靠？

When it comes to dietary or medical information,

當涉及營養或醫療資訊時，

the first thing to remember

需謹記在心的是

is that while studies on animals or individual cells

縱使以動物或個體細胞的研究實驗

can point the way towards further research,

可引導未來的研究，

the only way to know how something will affect humans

但唯一能了解事物如何影響人類的方法，

is through a study involving human subjects.

便是經由有人類參試者的研究。

And when it comes to human studies,

當其涉及人類的研究時，

the scientific gold standard is

科學的黃金法則

the randomized clinical trial, or RCT.

便是隨機臨床試驗 或 RCT（隨機對照試驗）。

The key to RCTs is that the subjects are randomly assigned

RCT 的關鍵是，

受試者將隨機地 被分配到各研究小組。

to their study groups.

They are often blinded to make them more rigorous.

這些受試者通常會被蒙眼 使其更加嚴謹。

This process attempts to ensure

這過程意欲確保

that the only difference between the groups

各小組間唯一的差別 僅是研究員的研究標的。

is the one the researchers are attempting to study.

For example,

例如，測試一項新的頭痛藥物時，

when testing a new headache medication,

a large pool of people with headaches

會集合一群有頭痛毛病的人， 將他們隨機的分成兩組，

would be randomly divided into two groups,

one receiving the medication

一組接受藥物治療， 而另一組只接受安慰劑。

and another receiving a placebo.

With proper randomization,

在正確的隨機分配下，

the only significant overall difference

在這兩組間 唯一顯著差別

between the two groups

是他們是否接受了藥物治療，

will be whether or not they received the medication,

rather than other differences that could affect results.

而不是其他可能會影響結果的差異。

Randomized clinical trials are incredible tools,

隨機臨床試驗是個不可思議的工具，

and, in fact, the US Food and Drug Administration

其實，美國食品及藥物管理局

often requires at least two to be conducted

經常性地規定一種新藥上市前 必須至少經過了兩次測試。

before a new drug can be marketed.

But the problem is that an RCT is not possible

然而，問題是 RCT 在許多案例上是不可行的。

in many cases,

either because it's not practical

若不是因為不切實際，

or would require too many volunteers.

便是因為需要太多自願者。

In such cases,

這種狀況下，

scientists use an epidemiological study,

科學家們便運用流行病學研究法。

which simply observes people going about their usual behavior,

它單純地觀察人們的日常行為，

rather than randomly assigning active participants

而不隨機分配現行參與者 到不變量控制群組。

to control invariable groups.

Let's say we wanted to study

假設，我們要研究市面上的 某種藥草成分是否會引起反胃。

whether an herbal ingredient on the market

causes nausea.

Rather than deliberately giving people something

我們並不刻意給人們服用 會引起他們反胃的東西，

that might make them nauseated,

we would find those who already take the ingredient

而是去尋找那些在日常生活中 已在服用這種成份的人。

in their everyday lives.

This group is called the cohort.

這個群組稱之為隊列。

We would also need a comparison group

我們也需要 沒有接觸這種成分的人們 來作為對照組。

of people who do not have exposure to the ingredient.

And we would then compare statistics.

然後比較統計數據。

If the rate of nausea is higher in the herbal cohort,

如果藥草隊列中，反胃的比例較高，

it suggests an association

那麼它暗示此藥草成份與反胃有所關聯。

between the herbal supplement and nausea.

Epidemiological studies are great tools

流行病學研究是個很棒的工具 能用於幾乎所有與健康有關的事物。

to study the health effects of almost anything,

without directly interfering in people's lives

它不直接干擾到人們的生活 也不分配他們到有潛在危險的環境裡。

or assigning them to potentially dangerous exposures.

So, why can't we rely on these studies

那麼，為什麼我們無法根據這些研究

to establish causal relationships

建立此內容物與其影響健康的因果關係？

between substances and their effects on health?

The problem is

問題出在 即使是最好的流行病學研究案例

that even the best conducted epidemiological studies

have inherent flaws.

也有它固有的缺陷。

Precisely because the test subjects

那便是因為其研究對象 不是隨機分類的群組。

are not randomly assigned to their groups.

For example, if the cohort in our herbal study

例如，我們藥草測試中的隊列

consisted of people who took the supplement

是由因醫療因素 服用這補給品的人群組成，

for health reasons,

they may have already had higher rates of nausea

他們可能已比其他取樣對象 有較高的反胃頻率。

than the other people in the sample.

Or the cohort group could've been composed of

或者，隊列群組成員可能是

people who shop at health food stores

健康食品店的常客， 或有特殊飲食習慣， 或較易取得醫療照護的人。

and have different diets

or better access to healthcare.

而這些實驗標的外 卻足以影響結果的因素，

These factors that can affect results,

in addition to the factor being studied,

are known as confounding variables.

被稱為混雜變量。

These two major pitfalls,

這兩大缺陷， 結合較一般的危險，

combined with more general dangers,

such as conflicts of interest or selective use of data,

例如：利益衝突 或是數據上的選用，

can make the findings of any particular epidemiological study suspect,

能使得該流行病學研究的結果令人懷疑，

and a good study must go out of its way

另外，一項優良的研究 必須貫徹到底

to prove that its authors have taken steps

以證明其作者執行了 排除這類誤差的必要步驟。

to eliminate these types of errors.

But even when this has been done,

但是就算是這樣做了，

流行病學研究法的本質 是在現有的群組間 檢驗其相異性，

the very nature of epidemiological studies,

which examine differences between preexisting groups,

rather than deliberately inducing changes within the same individuals,

而不是刻意的 在同一個體上導入變數。

means that a single study

這代表一項研究

can only demonstrate a correlation

只能在單一物質與單一醫療結論間 論證一種關聯性，

between a substance and a health outcome,

rather than a true cause and effect relationship.

而不是一項實質的因果關係。

At the end of the day,

總之！ 流行病學研究已經成為公眾健康的優良指標，

epidemiological studies have served as excellent guides to public health,

alerting us to critical health hazards,

為我們提出嚴重危害健康的警告，

such as smoking, asbestos, lead, and many more.

如吸煙、石棉、鉛、等等。

But these were demonstrated through

但這些是經由

眾多且執行完善的流行病學研究， 其皆指向同一論證結果。

multiple, well-conducted epidemiological studies,

all pointing in the same direction.

So, the next time you see a headline

所以， 當你再次看到新聞標題 報導著一種新的靈丹妙藥，

about a new miracle cure

or the terrible danger posed by an everyday substance,

或是關於某種日常物品帶來的驚悚危害，

try to learn more about the original study

請試著在妄下斷語前，

and the limitations inherent in any epidemiological study or clinical trial

對其原始研究、流行病學、 或任何臨床實驗固有的局限性 先做更深一層的了解。

before jumping to conclusions.