The data source I used is Composition of the human body.
This clearly shows that proteins, water and lipids make up 97% of the mass of a human body and I will assume that we can get a rather precise estimate if we just account for those to not have to deal with all the other material within the human body. After all this is about getting a good estimate, not a precise calculation.
Water
As stated before we can assume that H2O is readily available already so we can say there is no energy required to form this. That already accounts for 65% of the mass. So we have a total energy of $$E_{\text{H}_2\text{O}} = 0 \text{kJ}$$
Proteins
I am having serious trouble finding data on enthalpy of proteins. The only thing I found was regarding enthalpy change in protein folding and binding with a value of $$H_\text{proteins}=4 \frac{\text{kcal}}{\text{mol}} \ \ (16.7\frac{\text{kJ}}{\text{mol}}) $$ It is most certainly not referring to the right thing, but to at least have something to calculate with I will use this value, assuming the real value is of a similar magnitude. I put no certainty in that.
The average molecular length of a protein is 375 amino acids for humans. And an average amino acid weighs 110Da (so $110\frac{\text{g}}{\text{mol}}$) This results in an average molar weight of $$M_\text{protein}=41,250 \frac{\text{g}}{\text{mol}}$$. This assumes that the proteins are equally distributed in the body - which is probably false, but might give us a decent enough estimate.
$$m_{\text{arm}} = 5 \text{kg}$$
$$m_{\text{arm;proteins}} = 0.12 \times m_{\text{arm}} = 0.6 \text{kg}$$
$$n_\text{proteins} = m_{\text{arm}} / M_\text{arm;proteins} = 0.01455 \text{mol}$$
$$E = H_\text{proteins} \times n_\text{proteins} - \sum_i {H_\text{material i} \times n_\text{material i}} \ \ \text{(if we assume no byproducts)}$$
$$E_\text{proteins} = 243 \text{kJ} - \sum_i {H_\text{material i} \times n_\text{material i}}$$
I am uncertain what states of the starting materials to assume. Water ($−285.83 \frac{\text{kJ}}{\text{mol}}$), graphite ($0 \frac{\text{kJ}}{\text{mol}}$), nitrogen gas ($0 \frac{\text{kJ}}{\text{mol}}$) might work with possibly negligible additions or taking into account byproducts.
If we were to assume water as a base material we can see the energy needed would increase as the enthalpy of O2 and H2 are $0 \frac{\text{kJ}}{\text{mol}}$ of Oxygen would likely be a byproduct in formation of amino acids given this scenario and the enthalpy of formation of water is negative.
Since amino acid structures have about 2-4 Oxygen atoms it is fair to assume that you would need on average 3 H2O for a singular amino acid. With an average of 375 amino acids per protein $$n_{\text{H}_2\text{O}} = 375 \times 3 \times n_\text{proteins} = 16.369 \text{mol}$$. This gives us with the equation above $$E = H_\text{proteins} \times n_\text{proteins} - {H_{\text{H}_2\text{O}} \times n_{\text{H}_2\text{O}}} \ \ \text{(if we assume no byproducts with H $\neq 0$)} = 243 \text{kJ} - (-285.83 \frac{\text{kJ}}{\text{mol})} \times 16.369 \text{mol} = 4,922 \text{kJ}$$
If we take calorimetric values instead of all I tried to come up with reasonable data, we have 4 kcal per gram of proteins. This results in $$E_\text{proteins;calorimetric} = m_\text{arm;proteins} \times 4 \frac{\text{kcal}}{\text{g}} = 600g \times 4 \frac{\text{kcal}}{\text{g}} = 2400 \text{kcal} = 10,041kJ$$
So about twice as much as I had with the other data.
Lipids
With lipids I am also having troubles finding good data, so I decided to use the calorimetric data which appearantly is between $38,702 \frac{\text{kJ}}{\text{kg}}$ and $39,748 \frac{\text{kJ}}{\text{kg}}$.
I will use $39,748 \frac{\text{kJ}}{\text{kg}}$ as an estimate of the energy used to build lipids, which is probably not correct entirely but might not be off by that much.
$$m_{\text{arm;lipids}} = 0.2 \times m_{\text{arm}} = 1 \text{kg}$$
$$E_\text{lipids} = m_{\text{arm;lipids}} \times 39,748 \frac{\text{kJ}}{\text{kg}} = 39,748 \text{kJ}$$
Preliminary energy sum
Putting all the results together we reach an estiamated total of $$E_\text{total} = \sum_i{E_i} = 44.670 MJ$$ or if we take the calorimetric value for the proteins $$E_\text{total;cal.proteins} = 49.789 MJ$$
This may be a very basic answer but I found this link with some calorific values :
Bone and meat calorific values
This is an approximation but for this
mixture of meat and bone (the composition percentage is not specified)
they measured a gross calorific value of 19.69 MJ.kg-1.
This leads to approximatively 112 MJ for your 5.7 kg arm.
From what I recall, the calorific value is the energy gained by
burning the material which is pretty similar to separate the molecules
into single atoms or at least into some very smaller molecules. If so,
this is the opposite of the energy needed to creates molecules from
single atoms.
Now I was talking about a 5kg arm so taking that into account that would make 98MJ.
So we got estimates ranging from 44MJ to 98MJ.
According to a calorie calculator my recommended daily intake is 8140kJ (8.14MJ). This means the energy required would be 5.4-12 days worth of calorie intake.