Of particular concern is the idea that mammalian health can be affected by epigenetic tags received from parents or grandparents. For example, one group reported that pre-diabetic mice have different epigenetic tag patterns in their sperm and that their offspring have a higher chance of contracting diabetes. (Virginia Hughes has written an excellent article summarizing this and other related epigenetic studies.) A flurry of other biomedical and epidemiological research has strongly hinted that a susceptibility to obesity, diabetes, and heart disease can be passed on through epigenetic tags.
However, Heard & Martienssen are not convinced. In their Cell review, they admit that epigenetic inheritance has been demonstrated in plants and worms. But, mammals are completely different beasts, so to speak. Mammals go through two rounds of epigenetic "reprogramming" -- once after fertilization and again during the formation of gametes (sex cells) -- in which most of the chemical tags are wiped clean.
They insist that characteristics many researchers assume to be the result of epigenetic inheritance are actually caused by something else. The authors list four possibilities: Undetected mutations in the letters of the DNA sequence, behavioral changes (which themselves can trigger epigenetic tags), alterations in the microbiome, or transmission of metabolites from one generation to the next. The authors claim that most epigenetic research, particularly when it involves human health, fails to eliminate these possibilities.
It is true that environmental factors can influence epigenetic tags in children and developing fetuses in utero. What is far less clear, however, is whether or not these modifications truly are passed on to multiple generations. Even if we assume that epigenetic tags can be transmitted to children or even grandchildren, it is very unlikely that they are passed on to great-grandchildren and subsequent generations. The mammalian epigenetic "reprogramming" mechanisms are simply too robust.
This is an oft discussed topic in books about the paleolithic model, first discussed I suspect by Weston A. Price. Price noticed that populations that have the best health on the planet (at the time that he prowled the world to find them) had several things in common: very low incidence of dental carries, very good dental structure, very good bone structure overall, and a very low incidence of chronic disease (heart disease, cancer and their precursors) - In other words, as Denise Minger put it in her excellent book "Death By Food Pyramid", you are what you eat, what your parents ate, what their parents ate, and what you food ate/was fed.