The U-Shaped Curve, Polyamines, and Those Apparently Quite Demanding Gastrointestinal Bacteria
In the children’s story of Goldilocks and the three bears, Goldilocks broke into the bears’ house and found (among other things) three bowls of soup. She determined that one bowl of soup was too hot, one bowl was too cold, and one bowl was just right. The same thing is very common in human (and other) biochemical systems. For example, Figure 1 shows the case for total cholesterol.

According to the data, a total cholesterol value of about 220mg/dL is optimal for avoiding death by any means – no matter your age or your sex. Too low a cholesterol value can increase your chances for premature death, and too high a cholesterol value will also foster the same end.
Similarly, there’s also such a thing as being too skinny or being too fat.1

It turns out that the same u-shaped curve relationship probably also applies to polyamine consumption; i.e., not only may too low a consumption of certain polyamines be deleterious to your health, but too much of this otherwise good thing could also be dangerous. After all, recall Kiechl et al.’s important cautionary comment with regard to dietary spermidine at the end of this passage:
“To our knowledge, our study is the first to show an inverse relation between the amount of dietary spermidine intake and all-cause mortality in the general community. The association emerged as independent of other determinants of longevity and lifestyle. The survival advantage was driven by a reduced risk of death from all major causes. The key association was highly consistent in subgroups and successfully replicated in an independent cohort from the same geographical region. It exhibited a linear dose-response type and particular strength. Spermidine showed the strongest inverse relation with mortality among 146 nutrients studied. The reduction in mortality risk related to a diet rich in spermidine (top compared with bottom third of spermidine intake) was comparable to that associated with a 5.7-y younger age. All of the findings apply to spermidine from dietary sources and to amounts characteristically found in the Western diet and cannot readily be extrapolated to high-dose spermidine supplementation or extreme diets.” [Emphases added.]
There are isolated reports in the biological and medical literature suggesting that spermidine can, in fact, be slightly to seriously toxic at especially high rates of consumption. See, for example,
1) Serum spermidine in relation to risk of stroke: a multilevel study;
2) Acute and subacute toxicity of tyramine, spermidine, spermine, putrescine and cadaverine in rats; and,
The collective academic polyamine literature provides the following average daily polyamine consumption data for the countries shown in Table I, data that permit epidemiological examination of national polyamine consumption habits for evidence that too much – not to mention too little -- of one or other polyamine in the diet can be deleterious to human health and longevity.
Using the published estimated national average daily dietary polyamine intakes for each of the countries shown in Table I, and regressing2 those against national average life expectancies produces a multivariable linear equation predicting average national life expectancy with about 90.3% accuracy3 as shown in the following Table II ordinary least squares regression analysis output.

Putting the Table variable coefficients into equation form yields: log(life expectancy in years) = 1.738343-0.251801*log(mg putrescine/day) + 0.866728*log(mg spermidine/day) - 0.405104*log(mg spermine/day) + 0.5*0.645324*log(mg putrescine)^2 -0.5*0.592945*log(mg spermidine/day)^2 - 0.437871*log(mg putrescine/day)*log(mg spermidine/day) + 0.362375*log(mg spermidine/day)*log(mg spermine/day).
The adjusted R-squared value (Table II) for the linear regression of average national life expectancy against average national polyamine consumption rates indicates that the average national life expectancy at birth can be predicted with about 90.3% accuracy, just by knowing the average daily dietary consumption of the polyamines putrescine, spermidine, and spermine of a given nation’s human inhabitants.
So, it appears possible that the effects of these dietary polyamine consumption data on average national life expectancy are externally “telegraphing” what is going on internally in the human body with regard to ingested polyamines. For example, note that the negative sign of the squared spermidine factor (LSPD*LSPD) indicates that as the amount of spermidine goes up in a diet, the effect of that spermidine on life expectancy becomes more and more negative – just as seen on the right hand sides of the U-shaped curves shown in Figures 1 and 2 above. Conversely, the positive sign of the squared putrescine factor (LPUT*LPUT) indicates that as the amount of putrescine goes up in human daily food consumption, the effect of that putrescine on human life expectancy becomes increasingly positive – just as seen on the left hand sides of those same two Figures.
Judging from the characteristics and predictive strength of the equation, then, striking some kind of an approximate balance among putrescine, spermidine, and spermine in the human diet is important to the matters of health and longevity.
These predicted contrasting effects on life expectancy of increasing the amounts of spermidine and putrescine on people eating the average American diet can be more tangibly illustrated by numerically titrating the average American with servings of foods that are especially rich in spermidine (e.g., wheat germ) or foods that are particularly rich in putrescine (e.g., citrus fruits or sharp cheddar cheese) utilizing the basic, internationally-based equation provided above. Table III shows the results of some of this ‘table top’ mathematical experimentation.

As indicated in Table III, after the first 25 grams of the stuff, only gradual and slight progress in equation-predicted life expectancy is achieved by simply stepping up the amounts of spermidine and spermine-rich (0.16 mg/g and 0.15 mg/g, respectively) wheat germ in, say, the average American diet – and this predicted average longevity very, very gradually decreases as this hypothetical dietary component is further increased. On the other hand, adding in only small amounts of putrescine-rich sharp cheddar cheese (0.65 mg/g) or orange (0.12 mg/g) to the average American diet (or the average Japanese, Italian, or strict Paleo diet) is predicted to lead to quite large increases in average human life expectancy.4
All in all, these Table III results suggest that modern human diets (especially that of Americans) may be more starved for putrescine than they are for spermidine and/or spermine.
In this last regard, it is known that gastrointestinal bacteria that dwell in the colon utilize putrescine to locally produce the spermidine and spermine required to maintain and grow the tissues of the lowermost, most removed reaches of the human digestive system. This bacterial autonomy is important because little or no food-sourced spermidine or spermine makes it past the duodenum and small intestine, so the gut bacteria living within the colon must make their own — for themselves and our lower gut tissues. It is further recognized that during fasting, when no bacteria-nourishing, spermine and spermidine-poor semi-digested foodstuffs are descending down the mammalian digestive tract, these same bacteria are forced to directly extract putrescine from the mammalian circulatory system to obtain the basic energy supply necessary for their continued survival.

Perhaps, then, as humans age and gradually lose much of their cellular capacity for producing their own supplies of polyamines, their gut microbiota shift from beneficially acting mutualistically with us to acting as amensal parasites instead, strongly competing with us, their host, for dietary polyamines and thus foreshortening our lifespans. If this is so, human diets particularly rich in bacteria-feeding putrescine might serve to relieve some of the competitive pressure for systemic and dietary polyamine supply otherwise developing over time between humans and their gut micro-residents.
Finally (almost), for those who would like to mess around with the polyamine-dependent life expectancy prediction equation described here for the sake of personal curiosity, here’s the Excel file that was used to generate Table III.
For those same messing around purposes, here is also a sample of polyamine-rich foodstuff portions and their corresponding putrescine, spermidine, and spermine contents in milligrams. Different quantities of these polyamine-rich foods can be input into any of the national base cases provided in Table I in order to see what sort of effect such dietary polyamine additions would likely have on that nation’s average national life expectancy and average national health level.
PS: this post will very likely be the last you’ll be seeing from me until at least mid-summer. I’ve got a lot of seasonal chores to take care of over the next several months.
It’s odd, isn’t it, that having a body resting exactly on the “overweight” or “preobesity” cusp of being officially “fat” has the smallest possible chance of dying prematurely?
Which suggests that all other longevity-determining factors only account for about 10% of nation-to-nation variation in life expectancy.
The strict paleodiet seems an especially dangerous dietary habit, but by adding some daily cheese and an orange it appears that, on average, you might still be able to be around to chew on meat until you’re past a hundred or so.
Given how insanely strict my diet is as a result of autoimmune issues picked up overseas, this is interesting data. It makes me grateful for one of the foods I can eat—the orange. Godspeed with your seasonal chores.