I would recommend you to read this publication; Clin Chem. 2009 May;55(5):1011-8. Epub 2009 Mar 26.

It's published by the insulin standardization work group and they have compared a lot of different brands and methods including ELISA and RIA. They found the ELISA from Mercodia and the closed system from Roche to be the best methods.

Ourlab does it with elisa's luminex based assay.

Both have some problems, for example in RIA you have non specific binding problem and some contamination and in ELISA the main problem is the volume of solution in the wells, because based on beer 's law the path of ray light here is not a certain amount, specially where you work personally.

It is better that you use ELISA or Luminescence methods machinery systems, to reduce the problems.

Let me inform you that since you have to standardized all test before utilized them for samples, it will cover the personal faults.

We use the ELISA kits from Mercodia (Sweden) and are very satisfied with the them.

The effectivity of both is almost the same it only change the principleI of detection, although I think ELISA is the best choice, with radioimmunoassay you have to deal with radioactive stuff and you have to take extra safety considerations in the procedure and the disposal of wastes

I would use ELISA if you can. RIA may be a little more sensitive depending on the ELISA; however, dealing with radioactivity can be a bear.

ELISA is the best choice, every time i have relatively accurcy result

say no to radiation if you can!! ELISA goes right. Can you reach some luminex? if you have problems with sample volume, or want to do another parameters (same well), is the best chose.

RIA: relatively cost-effective. You can dilute and assay more sample as far as the radioactive count is enough for detection. However the handling of waste is messy.

ELISA: relatively expensive. Easy and enough sensitivity (depending on the kit). However bad correlation in low insulin concentration is often observed. You would be aware this problem especially when you want to measure low insulin concentration sample such as the plasma from fasted mice/rats.....

Both ELISA and RIA are of equal ranking for quantification of any protein. The most important thing is the OPTIMIZATION of each reagent and parameter that you used. Both have several advantages and disadvantages. In ELISA the end product is measured using a spectrophotometer revealing results as OD. End product of RIA is measured in a counter with results as counts per minute. With RIA one has to be more disciplined and ensure that the unwanted materials are disposed off according to the SOP, even with Iodine 135 where the half life is 40 days. When using ELISA one has to choose reagents with care. When HRPO is employed, all other reagents used should not contain sodium azide. The substrate TMB is light sensitive and should be stored in an appropriate container. Please read both methodology carefully. Choose one where equipment and reagents are easily acquired.

I would go for RIA every time. Elisa is thwart with interferrence from matrix differences between samples and lacks sensitivity which precision is dependent on.

The ultimate signal is interferred with by color, temperature, pH, inhibitors, cofactors, quenching etc. Nothing can influence the rate of radioactive decay. Few countries if any require licensing or special disposal for such low levels as in RIA kits and the risks are grossly over rated. Elisa kits are a little more expensive but if you have large numbers of samples you can process thousands of sample for the cost of an iodination ($600) and a few uL of antiserum using RIA. This is my specialty.

Will it be human insulin? other specie? The matrix is also important; serum? plasma? tissue?...

If you only think in accuracy, precision and linearlty, and forget that you have to deal with radioisotopes, RIA is the method that will raise the lowest measurement uncertainty, best accuracy and better precision.

ELISA is easier, it will give you acceptable results and uncertainty, but you have to take care about the analytical trazeability of the calibrator. Many ELISA kits do not attend to this question, which is very important. There is a WHO reference material for INSULIN (human).

José Rioja

Interlabcenter

We used ELISA kit from Alpco and it worked perfectly and it is sensitive and relieble

According to theoretical and Practical possibilities, estimation of Serum insulin concentration usually i prefer Elisa.Both having equal preference,But RIA having cumbersome , hectic process and handling radio active material it is mess.Best of my knowledge i suggest to use Elisa method.

Go for ELISA. Watch the sample stability for insulin. Could be tricky.

Elisa

Thank you all for your help. I intend to measure insulin in mice serum. I used to use a RIA assay from Linco research, but I have moved to a new lab and here they have no license to handle radio active material. I was wondering if I would loose accuracy using a Elisa assay.

I would recommend you to read this publication; Clin Chem. 2009 May;55(5):1011-8. Epub 2009 Mar 26.

It's published by the insulin standardization work group and they have compared a lot of different brands and methods including ELISA and RIA. They found the ELISA from Mercodia and the closed system from Roche to be the best methods.

I think that the ELISA is better than the RIA. Especially when we're using biotin and avidin ELISA sensitivity in insulin absorption is highly pico grams. so ELISA is more safety than RIA.

As my experiences I will say ELISA, because its more sensitive than RIA and also their is no chance of any harm during the assay like RIA has little of chance of radio activity.

In general there is no difference in the sensitivities between ELISA and RIA. The detection limit depends on the affinity of the primary antibody. ... and the test manufacturers are using different antibodies ...

@Ravinder: If you are taking the same antibodies in ELISA and RIA, using the same test principle, same coating conditions, optimal labelling for the nucleotide and the enzyme, then you will get the same detection limit. The base for the reaction is the Mass-action law. If you are going to calculate the data, you will find that as well on a theoretical level. You can inlude in your considerations also all kinds of other labels (Biotin, luminescenze markers, other enzymes). The second most important parameter influencing the senstivity (but very much less than the affinity constant) is the surface you are able to use. Therefore microparticle based immunoassasy are mor sensitive than microtiterplate based ...

By the way: ELISA are able to detect down to fg/mL (pg/L). We are using them in blood services to detect such small amounts of anti-HIV- or anti-HCV-antibodies or HBV (HbsAg).

Commercially, there are kits RIA and ELISA assays for insulin. these kits are validated and optimized by our suppliers. I think the choice is dictated by the equipment available in your laboratory (ELISA reader or scintillation counter) and technical staff.

khelili