OpenMP - create threads only once

I am trying to write a simple application using OpenMP. Unfortunately, I have problems with acceleration. In this application, I have one while loop. The body of this cycle consists of some instructions that must be executed sequentially and one for the cycle. I use #pragma omp parallel forto make this for the loop parallel. This cycle does not have much work, but it is called very often.

I am preparing two versions of the for loop and running the application on 1, 2 and 4cores.
version 1 (4 iterations in a loop): 22 sec, 23 s, 26 sec.
version 2 (100,000 iterations in a loop): 20 s, 10 s, 6 sec.

As you can see, when there is not much work for the cycle, the time on 2 and 4 cores is higher than on 1core. I think the reason is because it #pragma omp parallel forcreates new threads in each iteration of the while loop. So, I would like to ask you - is it possible to create threads once (before the while loop) and guarantee that some execution of the while loop will be executed sequentially?

#include <omp.h>
#include <iostream>
#include <math.h>
#include <stdlib.h>
#include <stdio.h>
#include <time.h>
int main(int argc, char* argv[])
{
    double sum = 0;
    while (true)
    {
        // ...
        // some work which should be done sequentially
        // ...

        #pragma omp parallel for num_threads(atoi(argv[1])) reduction(+:sum)
        for(int j=0; j<4; ++j)  // version 2: for(int j=0; j<100000; ++j)
        {
            double x = pow(j, 3.0);
            x = sqrt(x);
            x = sin(x);
            x = cos(x);
            x = tan(x);
            sum += x;

            double y = pow(j, 3.0);
            y = sqrt(y);
            y = sin(y);
            y = cos(y);
            y = tan(y);
            sum += y;

            double z = pow(j, 3.0);
            z = sqrt(z);
            z = sin(z);
            z = cos(z);
            z = tan(z);
            sum += z;
        }

        if (sum > 100000000)
        {
            break;
        }
    }
    return 0;
}